"Fossies" - the Fresh Open Source Software Archive

Member "automake-1.16.3/ChangeLog" (19 Nov 2020, 961776 Bytes) of package /linux/misc/automake-1.16.3.tar.xz:


As a special service "Fossies" has tried to format the requested text file into HTML format (style: standard) with prefixed line numbers. Alternatively you can here view or download the uninterpreted source code file. See also the latest Fossies "Diffs" side-by-side code changes report for "ChangeLog": 1.16.2_vs_1.16.3.

    1 2020-11-18  Jim Meyering  <meyering@fb.com>
    2 
    3 	version 1.16.3
    4 
    5 	* configure.ac (AC_INIT): Bump version number to 1.16.3.
    6 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
    7 	* NEWS: Record release version.
    8 
    9 2020-11-16  Jim Meyering  <meyering@fb.com>
   10 
   11 	tests: correct preceding change.
   12 
   13 	* t/auxdir-pr19311.sh: Fix error in case stmt and match more
   14 	upcoming versions.
   15 
   16 2020-11-16  Jim Meyering  <meyering@fb.com>
   17 
   18 	tests: auxdir-pr19311.sh no longer fails with latest autoconf
   19 
   20 	* t/list-of-tests.mk (XFAIL_TESTS): Remove from this list.
   21 	* t/auxdir-pr19311.sh: Instead, run this test only when autoconf
   22 	is 2.69d or newer.  Otherwise, skip it.
   23 
   24 2020-11-16  Jim Meyering  <meyering@fb.com>
   25 
   26 	maint: placate maintainer-check's rm -f check
   27 
   28 	* t/vala-recursive-setup.sh: Add an unnecessary -f option
   29 	to an rm invocation to avoid "make maintainer-check" failure.
   30 
   31 2020-11-15  Jim Meyering  <meyering@fb.com>
   32 
   33 	maint: Update files from upstream with 'make fetch'
   34 
   35 	* lib/config.guess: Update.
   36 	* lib/config.sub: Likewise.
   37 	* lib/texinfo.tex: Likewise.
   38 
   39 2020-11-15  Jim Meyering  <meyering@fb.com>
   40 
   41 	tests: avoid missing .dvi failure with parallel tests
   42 
   43 	* t/txinfo-no-clutter.sh: Tests of texinfo-related rules
   44 	had overlap that made them fail often when some rules were
   45 	run in parallel, so inhibit parallelism in that one directory.
   46 	See discussion starting at
   47 	https://lists.gnu.org/r/automake-patches/2020-11/msg00011.html
   48 
   49 2020-11-15  Jim Meyering  <meyering@fb.com>
   50 
   51 	tests: protect against parallel false failure
   52 
   53 	* t/parallel-tests-console-output.sh: Do not depend on the order
   54 	of items in test summary. With a parallel test run, they may
   55 	appear in a different order, e.g., when running tests like this:
   56 	  make check AM_TESTSUITE_MAKE='make -j14'
   57 	Sort the expected output and the actual output before comparing.
   58 
   59 2020-11-15  Jim Meyering  <meyering@fb.com>
   60 
   61 	doc: fix quoting in suggested parallel test invocation
   62 
   63 	* t/README: Fix reversed single/double quotes.
   64 
   65 2020-11-14  Jim Meyering  <meyering@fb.com>
   66 
   67 	tests: accommodate an $ac_aux_dir of "." or "./"
   68 
   69 	* t/auxdir-pr15981.sh: This test would fail when run with
   70 	autoconf-2.69d because $ac_aux_dir would be "./" rather than
   71 	the expected ".".  Accept both.
   72 
   73 2020-11-14  Jim Meyering  <meyering@fb.com>
   74 
   75 	tests: avoid failures due to missing ar-lib
   76 
   77 	* t/ar4.sh: Create dummy ar-lib, as done in other tests,
   78 	to avoid failure like this:
   79 	configure: error: cannot find required auxiliary files: ar-lib
   80 	* t/ar5.sh: Likewise.
   81 
   82 2020-11-13  Karl Berry  <karl@freefriends.org>
   83 
   84 	install-sh: trailing whitespace.
   85 
   86 	* lib/install-sh: remove trailing whitespace. Sigh.
   87 	(scriptversion): 2020-11-14.01
   88 
   89 2020-11-13  Robert Menteer  <reetnem@mac.com>
   90 
   91 	dejagnu: quote `pwd` when writing "set objdir" line to site.exp.
   92 
   93 	This change fixes https://bugs.gnu.org/44600.
   94 
   95 	* lib/am/dejagnu.am (site.exp): quote set objdir line.
   96 	* NEWS: mention this.
   97 
   98 2020-11-12  Karl Berry  <karl@freefriends.org>
   99 
  100 	install-sh: new option -S SUFFIX for simple file backups.
  101 
  102 	* lib/install-sh: implement and document -S.
  103 	Patch sent by Julien Elie:
  104 	        https://lists.gnu.org/archive/html/automake-patches/2018-03/msg00004.html
  105 	(scriptversion): 2020-11-13.01
  106 	* t/install-sh-option-S.sh: new test.
  107 	* t/list-of-tests.mk (handwritten_tests): add it.
  108 	* NEWS: mention it.
  109 
  110 2020-11-10  Karl Berry  <karl@freefriends.org>
  111 
  112 	install-sh: --help tweaks.
  113 
  114 	* lib/install-sh (usage): avoid too-long line, mention
  115 	bug reporting address (bug-automake) and automake home page.
  116 
  117 2020-11-10  Karl Berry  <karl@freefriends.org>
  118 
  119 	install-sh: new option -p to preserve mtime.
  120 
  121 	* lib/install-sh: new option -p to call cp -p.
  122 	Idea from patch sent by Julien Elie:
  123 	https://lists.gnu.org/archive/html/automake-patches/2018-03/msg00002.html
  124 	(scriptversion): 2020-11-11.02
  125 	* NEWS: mention this.
  126 
  127 	2020-11-10  Karl Berry  <karl@freefriends.org>
  128 
  129 2020-11-10  Karl Berry  <karl@freefriends.org>
  130 
  131 	vala: forgot to update lists-of-tests.mk.
  132 
  133 	* t/list-of-tests.mk (handwritten_tests): include
  134 	t/vala-libs-distcheck.sh and t/vala-libs-vpath.sh.
  135 	Should have been committed with 2020-10-29 vala change,
  136 	but somehow missed.
  137 
  138 	2020-11-10  Karl Berry  <karl@freefriends.org>
  139 
  140 2020-11-10  Karl Berry  <karl@freefriends.org>
  141 
  142 	install-sh: do not chown existing directories.
  143 
  144 	* lib/install-sh: do not chown existing directories.
  145 	Original patch sent by Julien Elie:
  146 	https://lists.gnu.org/archive/html/automake-patches/2018-03/msg00003.html
  147 	(scriptversion): 2020-11-11.01
  148 	* NEWS: mention this.
  149 
  150 2020-11-10  Karl Berry  <karl@freefriends.org>
  151 
  152 	install-sh: do not redundantly specify -f to rm.
  153 
  154 	* lib/install-sh: do not redundantly specify -f to rm.
  155 	Mention implication for RMPROG in the --help message.
  156 	Original patch sent by Julien Elie:
  157 	https://lists.gnu.org/archive/html/automake-patches/2018-03/msg00005.html
  158 	* NEWS: mention this.
  159 
  160 2020-11-07  Reuben Thomas  <rrt@sc3d.org>
  161 
  162 	vala: improve support, especially builddir vs. srcdir.
  163 
  164 	This change fixes https://bugs.gnu.org/13002.
  165 
  166 	* NEWS: mention these changes.
  167 	* bin/automake.in: generated C files go in builddir, not srcdir.
  168 	Distribute the header files generated from VAPI files.
  169 	* t/vala-libs-distcheck.sh: new test for `make distcheck' of a
  170 	Vala library.
  171 	* t/vala-libs-vpath.sh: new test for a VPATH build of a Vala library.
  172 	* t/vala-libs.sh: add local VAPIs used as external --package to test.
  173 
  174 	* t/vala-recursive-setup.sh: we need to make
  175 	maintainer-clean at one point to remove stamp files to avoid
  176 	confusing a VPATH build performed after a non-VPATH build.
  177 	* t/vala-non-recursive-setup.sh: likewise.
  178 
  179 	* t/vala-parallel.sh: some test paths need changing to take into
  180 	account that generated C files now go in builddir.
  181 	* t/vala-per-target-flags.sh: likewise.
  182 	* t/vala-recursive-setup.sh: likewise.
  183 	* t/vala-vpath.sh: likewise.
  184 
  185 2020-11-07  Karl Berry  <karl@freefriends.org>
  186 
  187 	tests: recompute dependencies when lists-of-tests.mk changes.
  188 
  189 	This change fixes https://bugs.gnu.org/44458
  190 	and updates https://bugs.gnu.org/11347.
  191 
  192 	* t/local.mk ($(srcdir)/%D/testsuite-part.am): restore
  193 	dependency on '%D/list-of-tests.mk' (i.e., t/list-of-tests.mk),
  194 	partially reverting the change of 26 Apr 2012 for bug#11347.
  195 	Otherwise, new tests that have dependencies will not cause an
  196 	update of testsuite-part.am, leading to strange failures
  197 	(bug#44458). The original problem being fixed in #11347 was
  198 	unnecessary rebuilding when modifying tests; that should not be
  199 	affected here, but when new tests are added, it seems reasonable,
  200 	as well as necessary, to ensure dependencies are updated.
  201 
  202 2020-10-27  Miro Hron\v{c}ok  <miro@hroncok.cz>
  203 
  204 	python: determine Python (3.10) version number correctly.
  205 
  206 	This change fixes https://bugs.gnu.org/44239
  207 	(and https://bugzilla.redhat.com/show_bug.cgi?id=1889732).
  208 
  209 	* m4/python.m4: use print('%u.%u' % sys.version_info[:2]) for
  210 	the version number instead of merely sys.version[:3], so the
  211 	numbers are treated as numbers.
  212 	* t/python-vars.sh (PYTHON_VERSION): Likewise.
  213 	* doc/automake.texi: Document it.
  214 	* NEWS: mention it. (Minor tweaks from Karl Berry.)
  215 
  216 2020-10-23  Jim Meyering  <meyering@fb.com>
  217 
  218 	doc: correct "moved in", to "moved to"
  219 
  220 	* NEWS: Correct wording.
  221 	* contrib/README: Likewise.
  222 	* doc/automake.texi: Likewise.
  223 
  224 2020-10-23  Reuben Thomas  <rrt@sc3d.org>
  225 
  226 	contrib/README: fix and clarify the English
  227 
  228 2020-10-23  Reuben Thomas  <rrt@sc3d.org>
  229 
  230 	Improve Vala compiler detection: use API version, not compiler version
  231 
  232 	* m4/vala.m4: check `valac --api-version', not `valac --version'.
  233 	* doc/automake.texi: update documentation.
  234 
  235 2020-10-07  Zack Weinberg  <zackw@panix.com>
  236 
  237 	Use complete configure.ac’s in testsuite.
  238 
  239 	Autoconf 2.70 will issue warnings if it encounters a configure.ac that doesn’t
  240 	call both AC_INIT and AC_OUTPUT.
  241 
  242 	Automake already issues warnings if it encounters a configure.ac that uses an
  243 	AM_ macro but doesn’t call AM_INIT_AUTOMAKE or AC_CONFIG_FILES([Makefile]).
  244 
  245 	In two places, the testsuite was tripping these warnings, leading to spurious
  246 	failures with Autoconf 2.70 betas.
  247 
  248 	* t/aminit-moreargs-deprecation.sh: Add AC_OUTPUT to test configure.ac.
  249 	* t/mkdirp-deprecation.sh: Use a complete test configure.ac, not a stub
  250 	  containing only a use of AM_PROG_MKDIR_P.
  251 
  252 2020-10-06  Karl Berry  <karl@freefriends.org>
  253 
  254 	automake: install-exec did not depend on $(BUILT_SOURCES).
  255 
  256 	This change fixes https://bugs.gnu.org/43683.
  257 
  258 	* lib/am/install.am (install-exec): %maybe_BUILT_SOURCES% dependency,
  259 	twice.  Basic patch from madmurphy (tiny change), message#8.
  260 	(.MAKE) [maybe_BUILT_SOURCES]: depend on install-exec.
  261 	* NEWS: mention it.
  262 	* doc/automake.texi (Sources): mention this (also that make dist
  263 	depends on $(BUILT_SOURCES)).
  264 	* t/built-sources-install-exec.sh: new test.
  265 	* t/list-of-tests.mk (handwritten_TESTS): add it.
  266 	* t/built-sources-install.sh: typo.
  267 	* t/built-sources-check.sh: typo.
  268 
  269 2020-09-21  Zack Weinberg  <zackw@panix.com>
  270 
  271 	Update documentation of warnings options and strictness levels.
  272 
  273 	The warning categories ‘cross’ and ‘portability-recursive’ were not mentioned
  274 	in the manual.
  275 
  276 	Also clarify the relationship between warnings categories and strictness
  277 	levels, and streamline the description of strictness levels by merging the
  278 	“Gnits” section into the “Strictness” section.
  279 
  280 	* doc/automake.texi (Gnits, Strictness): Combine these sections.
  281 	  Minor revisions to explanation of strictness levels.
  282 	  (automake Invocation): Add documentation of all the warnings
  283 	  categories that have been added since the last time this section
  284 	  was updated.  Minor clarifications.
  285 
  286 2020-09-21  Zack Weinberg  <zackw@panix.com>
  287 
  288 	Use WARNINGS=none to suppress warnings from autom4te runs.
  289 
  290 	aclocal uses autom4te in trace mode to scan configure.ac for macros whose
  291 	definition is not yet available.  It has a kludge to prevent this from
  292 	producing spurious warnings, but a cleaner, fully backward compatible, way to
  293 	get the same effect is to set WARNINGS=none in the environment and not pass
  294 	down any -W options.  (This is better than passing -Wnone on the command line
  295 	because it automatically inherits to any subprocesses started by autom4te.)
  296 
  297 	Perl’s ‘local’ feature can be used to make the enviironment variable setting
  298 	temporary, reverting to the previous value when we exit the function.
  299 
  300 	automake also runs autom4te (technically autoconf) in trace mode; warnings
  301 	from this invocation will not be *spurious*, but in the common case where
  302 	the person running automake is going to run autoconf next, they will be
  303 	duplicated.  Therefore, make the same change to automake.
  304 
  305 	* bin/aclocal.in (trace_used_macros)
  306 	* bin/automake.in (scan_autoconf_traces):
  307 	  Use “local $ENV{WARNINGS}='none'” to suppress warnings from autom4te.
  308 
  309 2020-09-18  Zack Weinberg  <zackw@panix.com>
  310 
  311 	New utility function Automake::ChannelDefs::merge_WARNINGS.
  312 
  313 	This function merges a list of warnings categories into the environment
  314 	variable WARNINGS, returning a new value to set it to.  The intended use
  315 	is in code of the form
  316 
  317 	{
  318 	  local $ENV{WARNINGS} = merge_WARNINGS ("this", "that");
  319 
  320 	  # run a command here with WARNINGS=this,that,etc
  321 	}
  322 
  323 	This is not actually used in automake, but will be in autoconf.
  324 
  325 	* lib/Automake/ChannelDefs.pm (merge_WARNINGS): New function.
  326 
  327 2020-09-12  Zack Weinberg  <zackw@panix.com>
  328 
  329 	t/python-virtualenv.sh: Skip when versions don’t match
  330 
  331 	On some operating systems ‘python’ is Python 2.x but ‘virtualenv -ppython’
  332 	will create a virtualenv that uses Python 3.x.  This is a bug, but it’s
  333 	not *automake’s* bug, and should not cause t/python-virtualenv.sh to fail.
  334 	Skip the test, instead of failing it, when the inner=outer version check
  335 	fails.
  336 
  337 	(This also has nothing to do with the main goal of this patchset, it just
  338 	annoyed me while I was testing.)
  339 
  340 	* t/python-virtualenv.sh: Skip test, rather than failing it, when
  341 	  $py_version_pre != $py_version_post.
  342 
  343 2020-09-12  Zack Weinberg  <zackw@panix.com>
  344 
  345 	Consistently use ‘our’ instead of ‘use vars’.
  346 
  347 	At file scope of a file containing at most one ‘package’ declaration,
  348 	‘use vars’ is exactly equivalent to ‘our’, and the latter is preferred
  349 	starting with Perl 5.6.0, which happens to be the oldest version we
  350 	support.
  351 
  352 	(This change has nothing to do with the previous two, but I want to make the
  353 	same change in Autoconf and that means doing it here for all the files synced
  354 	from Automake.)
  355 
  356 	(I don’t know why, but this change exposed a latent bug in FileUtils.pm where
  357 	the last pod block in the file didn’t have a ‘=cut’ delimiter, so the code
  358 	after it was considered documentation, causing ‘require FileUtils’ to fail.)
  359 
  360 	* lib/Automake/ChannelDefs.pm
  361 	* lib/Automake/Channels.pm
  362 	* lib/Automake/Condition.pm
  363 	* lib/Automake/Configure_ac.pm
  364 	* lib/Automake/DisjConditions.pm
  365 	* lib/Automake/FileUtils.pm
  366 	* lib/Automake/General.pm
  367 	* lib/Automake/Getopt.pm
  368 	* lib/Automake/Options.pm
  369 	* lib/Automake/Rule.pm
  370 	* lib/Automake/RuleDef.pm
  371 	* lib/Automake/VarDef.pm
  372 	* lib/Automake/Variable.pm
  373 	* lib/Automake/Wrap.pm
  374 	* lib/Automake/XFile.pm:
  375 	  Replace all uses of ‘use vars’ with ‘our’.
  376 
  377 	* lib/Automake/FileUtils.pm:
  378 	  Add missing ‘=cut’ to a pod block near the end of the file.
  379 
  380 2020-09-12  Zack Weinberg  <zackw@panix.com>
  381 
  382 	Consistently process -W(no-)error after all other warning options.
  383 
  384 	automake and aclocal were processing ‘-W(no-)error’ whenever it
  385 	appeared on the command line, which means that
  386 	‘-Werror,something-strange’ would issue a hard error, but
  387 	‘-Wsomething-strange,error’ would only issue a warning.
  388 
  389 	It is not desirable for warnings about unknown warning categories ever to be
  390 	treated as a hard error; that leads to problems for driver scripts like
  391 	autoreconf, which would like to pass whatever -W options it got on its own
  392 	command line down to all the tools and not worry about which tools understand
  393 	which warning categories.  Also, this sort of order dependence is confusing
  394 	for humans.
  395 
  396 	Change parse_warnings to take just one option, the _complete_ list of warning
  397 	categories seen on the command line, and to process -Werror / -Wno-error after
  398 	processing all other warnings options.  Thus, unknown warnings categories will
  399 	always just be a plain warning.  This does mean aclocal has to stop using
  400 	parse_warnings as a Getopt::Long callback, but that’s not a big deal.
  401 
  402 	Similarly, change parse_WARNINGS to record whether ‘error’ appeared in the
  403 	environment variable, but not activate warnings-are-errors mode itself.
  404 	parse_warnings picks up the record and honors it, unless it’s overridden by
  405 	the command line.
  406 
  407 	* lib/Automake/ChannelDefs.pm ($werror): New package global (not exported).
  408 	  (parse_WARNINGS): Do not call switch_warning for ‘error’ / ‘no-error’;
  409 	  just toggle the value of $werror.
  410 	  (parse_warnings): Do not call switch_warning immediately for
  411 	  ‘error’ / ‘no-error’; toggle $werror instead.  Call switch_warning ‘error’
  412 	  at the very end if $werror is true.  Remove unused $OPTION argument.
  413 	* bin/automake.in: parse_warnings now takes only one argument.
  414 	* bin/aclocal.in: Call parse_warnings after parse_options instead of
  415 	  using it as a parse_options callback.
  416 
  417 2020-09-11  Zack Weinberg  <zackw@panix.com>
  418 
  419 	Sync ChannelDefs.pm from autoconf.
  420 
  421 	ChannelDefs.pm *ought* to be kept in sync between automake and autoconf,
  422 	because it defines the set of valid -W options, and autoreconf assumes
  423 	that it can pass arbitrary -W options to all of the tools it invokes.
  424 	However, it isn’t covered by either project’s ‘make fetch’ and it hasn’t
  425 	actually *been* in sync for more than 17 years.
  426 
  427 	This patch manually brings over all of the changes made on the autoconf side.
  428 	Most importantly, there is a new warnings channel ‘cross’, for warnings
  429 	related to cross-compilation.  Also, the ‘usage’ function now *returns*
  430 	the text to be put into a usage message, instead of printing it itself.
  431 	(This is necessary on autoconf’s side.)
  432 
  433 	* lib/Automake/ChannelDefs.pm: Sync from autoconf.
  434 	  (cross): New warnings channel.
  435 	  (portability-recursive): Document.
  436 	  (usage): Now returns the text to be printed, instead of printing it.
  437 	  (parse_warnings): Second argument may now be a list.
  438 
  439 2020-09-05  Zack Weinberg  <zackw@panix.com>
  440 
  441 	automake: be robust against directories containing ().
  442 
  443 	This change fixes https://bugs.gnu.org/14196.
  444 
  445 	* m4/missing.m4 (AM_MISSING_HAS_RUN): always quote the
  446 	invocation (not just if $am_aux_dir contains space or tab), in
  447 	case $am_aux_dir contains () or other metachars not rejected by
  448 	AM_SANITY_CHECK; quoting with '...' suggested by Jim Meyering.
  449 	* t/man6.sh (HELP2MAN): adjust grep since missing value
  450 	is quoted now.
  451 	* t/am-missing-prog.sh: likewise.
  452 
  453 2020-09-04  Issam E. Maghni  <issam.e.maghni@mailbox.org>
  454 
  455 	maint: Update files from upstream with 'make fetch'
  456 
  457 	* lib/config.guess: Update.
  458 	* lib/config.sub: Likewise.
  459 	* lib/gendocs_template: Likewise.
  460 	* lib/gitlog-to-changelog: Likewise.
  461 	* lib/texinfo.tex: Likewise.
  462 	* lib/update-copyright: Likewise.
  463 
  464 2020-08-31  Zack Weinberg  <zackw@panix.com>
  465 
  466 	perl: use warnings instead of -w; consistent ordering of use, etc.
  467 
  468 	Per thread at:
  469 	https://lists.gnu.org/archive/html/automake-patches/2020-08/msg00009.html
  470 
  471 	* bin/aclocal.in: use warnings instead of #!...-w;
  472 	consistent ordering of basic "use" directives,
  473 	then BEGIN block,
  474 	then standard modules in ASCII order,
  475 	then Automake:: modules (not sort),
  476 	finally use vars.
  477 	Also sort @ISA lists and use qw(...) in ASCII order.
  478 	* bin/automake.in: likewise.
  479 	* lib/Automake/ChannelDefs.pm: likewise.
  480 	* lib/Automake/Channels.pm: likewise.
  481 	* lib/Automake/Condition.pm: likewise.
  482 	* lib/Automake/Config.in: likewise.
  483 	* lib/Automake/Configure_ac.pm: likewise.
  484 	* lib/Automake/DisjConditions.pm: likewise.
  485 	* lib/Automake/FileUtils.pm: likewise.
  486 	* lib/Automake/General.pm: likewise.
  487 	* lib/Automake/Getopt.pm: likewise.
  488 	* lib/Automake/Item.pm: likewise.
  489 	* lib/Automake/ItemDef.pm: likewise.
  490 	* lib/Automake/Language.pm: likewise.
  491 	* lib/Automake/Location.pm: likewise.
  492 	* lib/Automake/Options.pm: likewise.
  493 	* lib/Automake/Rule.pm: likewise.
  494 	* lib/Automake/RuleDef.pm: likewise.
  495 	* lib/Automake/VarDef.pm: likewise.
  496 	* lib/Automake/Variable.pm: likewise.
  497 	* lib/Automake/Version.pm: likewise.
  498 	* lib/Automake/Wrap.pm: likewise.
  499 	* lib/Automake/XFile.pm: remove unnecessary imports of
  500 	Carp, DynaLoader, and File::Basename.
  501 
  502 2020-08-28  Robert Wanamaker  <rlw@nycap.rr.com>
  503 
  504 	docs: automake-history.texi @dircategory Software development.
  505 
  506 	Per thread at:
  507 	https://lists.gnu.org/archive/html/automake-patches/2020-08/msg00006.html
  508 
  509 	* doc/automake-history.texi (@dircategory): Define.
  510 
  511 2020-08-28  Karl Berry  <karl@freefriends.org>
  512 
  513 	automake: if TEST_EXTENSIONS is set to empty, don't look inside it.
  514 
  515 	This change fixes https://bugs.gnu.org/42635.
  516 
  517 	* bin/automake.in (handle_tests): do not use $test_suffixes[0]
  518 	if it does not exist.
  519 	* t/test-extensions-empty.sh: new test.
  520 	* t/list-of-tests.mk (handwritten_TESTS): add it.
  521 
  522 2020-08-13  Felix Yan  <felixonmars@archlinux.org>
  523 
  524 	docs: typo in tap-driver.sh.
  525 
  526 	Per thread at:
  527 	https://lists.gnu.org/archive/html/automake-patches/2020-08/msg00000.html
  528 
  529 	* lib/tap-driver.sh (setup_result_obj): "assing" typo, etc.
  530 
  531 2020-08-01  Paul Eggert  <eggert@cs.ucla.edu>
  532 
  533 	port XFile locking to OpenIndiana
  534 
  535 	I observed this problem on an NFS filesystem on an OpenIndiana
  536 	host (5.11 illumos-dde7ba523f i386).  fcntl (fd, F_SETLK, ...)
  537 	failed with errno == EINVAL, which POSIX allows for files that
  538 	do not support locking.
  539 	* lib/Automake/XFile.pm (lock): Treat EINVAL like ENOLCK.
  540 
  541 2020-07-26  Paul Eggert  <eggert@cs.ucla.edu>
  542 
  543 	* Update scriptversions for install-sh, mkinstalldirs.
  544 
  545 2020-07-26  Paul Eggert  <eggert@cs.ucla.edu>
  546 
  547 	Install directories mode 755 instead of using umask
  548 
  549 	Problem reported by Antoine Amarilli in:
  550 	https://lists.gnu.org/archive/html/automake/2019-01/msg00000.html
  551 	and followed up by Akim Demaille in:
  552 	https://lists.gnu.org/archive/html/bug-bison/2020-07/msg00040.html
  553 	* bin/automake.in: Add a comment about this.
  554 	* lib/install-sh: Ignore umask; just create directories mode 755
  555 	unless overridden via -m (for non-intermediate directories only).
  556 	Also, fix 'umask=$mkdir_umask' typo.
  557 	* lib/mkinstalldirs: Likewise.
  558 
  559 2020-06-29  Paul Eggert  <eggert@cs.ucla.edu>
  560 
  561 	automake: remove stray up_to_date_p
  562 
  563 	* lib/Automake/FileUtils.pm (up_to_date_p):
  564 	Don’t export up_to_date_p, which was removed in
  565 	2020-05-11T00:40:14Z!karl@freefriends.org.
  566 
  567 2020-06-06  Karl Berry  <karl@freefriends.org>
  568 
  569 	tests: support -fno-common in vala-mix2 test.
  570 
  571 	This change fixes https://bugs.gnu.org/41726.
  572 
  573 	* t/vala-mix2.sh: extern in .h, initialization in .c.
  574 	GCC 10 defaults to -fno-common.
  575 
  576 2020-06-06  Karl Berry  <karl@freefriends.org>
  577 
  578 	automake: support AM_TESTSUITE_SUMMARY_HEADER override.
  579 
  580 	This change handles https://bugs.gnu.org/11745.
  581 
  582 	* lib/am/check.am (AM_TESTSUITE_SUMMARY_HEADER): new variable.
  583 	Default value is " for $(PACKAGE_STRING)", including quotes,
  584 	to keep the default output the same.
  585 	($(TEST_SUITE_LOG)): use it, unquoted.
  586 	* doc/automake.texi (Scripts-based Testsuites): document it.
  587 	* NEWS: mention it.
  588 	* t/testsuite-summary-header.sh: new test.
  589 	* t/list-of-tests.mk (handwritten_tests): add it.
  590 	* t/ax/testsuite-summary-checks.sh: fix typo.
  591 
  592 2020-05-28  Akim Demaille  <akim@gnu.org>
  593 
  594 	docs: promote Makefile snippets that work properly with make -n.
  595 
  596 	This change handles https://bugs.gnu.org/10852.
  597 
  598 	* doc/automake.texi (Multiple Outputs): Split commands than
  599 	reinvoke $(MAKE) to avoid file removals during dry runs.
  600 
  601 2020-05-25  Karl Berry  <karl@freefriends.org>
  602 
  603 	docs: forgot TAR in NEWS; fix " -- " in manual.
  604 
  605 	* NEWS: it seems the TAR envvar was never mentioned in NEWS;
  606 	add it, back for 1.11.3 when it was apparently implemented.
  607 	* doc/automake.texi: consistently use "---" instead of " --".
  608 
  609 2020-05-25  Karl Berry  <karl@freefriends.org>
  610 
  611 	docs: TAR envvar overrides "tar" for make dist.
  612 
  613 	This change finishes https://bugs.gnu.org/9822.
  614 
  615 	* doc/automake.texi (Basics of Distribution): mention that
  616 	environment variable TAR overrides "tar".
  617 
  618 2020-05-17  Karl Berry  <karl@freefriends.org>
  619 
  620 	automake: new variable AM_DISTCHECK_DVI_TARGET to override "dvi".
  621 
  622 	This change fixes https://bugs.gnu.org/8289.
  623 
  624 	* lib/am/distdir.am (AM_DISTCHECK_DVI_TARGET): define as dvi.
  625 	(distcheck): use it, isntead of hardcoding dvi.
  626 	* lib/Automake/Variable.pm (%_silent_variable_override): add
  627 	AM_DISTCHECK_DVI_TARGET.
  628 	* t/distcheck-override-dvi.sh: new test.
  629 	* t/list-of-tests.mk (handwritten_TESTS): add it.
  630 	* doc/automake.texi (Checking the Distribution): document this.
  631 	(Third-Party Makefiles): explicitly mention that
  632 	EMPTY_AUTOMAKE_TARGETS is not a built-in or special name.
  633 	Various other index entries and wording tweaks.
  634 	* NEWS (Distribution): mention this.
  635 
  636 2020-05-10  Karl Berry  <karl@freefriends.org>
  637 
  638 	automake: remove unused Automake::FileUtils::up_to_date_p function.
  639 
  640 	Per thread at:
  641 	https://lists.gnu.org/archive/html/automake-patches/2020-04/msg00000.html>
  642 	especially:
  643 	https://lists.gnu.org/archive/html/automake-patches/2020-05/msg00003.html
  644 
  645 	* lib/Automake/FileUtils.pm (up_to_date_p): remove.
  646 	Nothing in Automake itself uses this.  It is used in Autoconf's
  647 	autom4te utility, but Autoconf has its own copy, and the duplication
  648 	was impeding auto4mte development, as discussed in the thread above.
  649 	(While here, insert missing =over/=back to placate pod2text.)
  650 	* NEWS (Miscellanous changes): note this.
  651 
  652 2020-05-07  Karl Berry  <karl@freefriends.org>
  653 
  654 	docs: make dist implies make dvi.
  655 
  656 	This change handles https://bugs.gnu.org/7994.
  657 
  658 	* doc/automake.texi (Preparing Distributions): make distcheck
  659 	runs make dvi.
  660 	(Auxiliary Programs) <texinfo.tex>: mention
  661 	that make dist runs make dvi, and therefore a TeX system is
  662 	required when Texinfo sources are present. Add @cmindex entries
  663 	for all auxiliary programs while we're here.
  664 
  665 2020-05-07  Karl Berry  <karl@freefriends.org>
  666 
  667 	tests: TeX system required for two more tests.
  668 
  669 	* t/instdir-no-empty.sh (required): makeinfo tex texi2dvi dvips.
  670 	* t/txinfo-bsd-make-recurs.sh (required): likewise.
  671 
  672 2020-04-23  Vincent Lefevre  <vincent@vinc17.net>
  673 
  674 	bug#40699: "dist Hook" documentation in manual is incorrect or unclear about write permissions
  675 
  676 	On 2020-04-20 14:59:00 -0600, Karl Berry wrote:
  677 	>     i.e. it does not change the permissions in order to make the removal
  678 	>     work recursively
  679 	>
  680 	> Right, I see it now. Had been testing the wrong thing.
  681 	>
  682 	> So, can you propose a specific change for the manual? -k
  683 
  684 	I think that it is sufficient to fix the example (the explanation
  685 	is just below). BTW, the second example is also incorrect.
  686 
  687 	commit a639e5b51cadbaff88ca4059b4db4571c811070c
  688 	Author: Vincent Lefevre <vincent@vinc17.net>
  689 	Date:   2020-04-23 17:33:54 +0200
  690 
  691 	    doc: fix dist-hook examples
  692 
  693 2020-04-18  Karl Berry  <karl@freefriends.org>
  694 
  695 	cosmetics: spurious word in README, copyright year.
  696 
  697 	* README: delete spurious "that"; update copyright year end to 2020.
  698 	Original suggestion from Vincent Lefevre,
  699 	https://lists.gnu.org/archive/html/automake-patches/2020-04/msg00007.html
  700 
  701 2020-04-08  Samuel Tardieu  <sam@rfc1149.net>
  702 
  703 	docs: test-driver options do not accept =, update --help.
  704 
  705 	This change fixes https://bugs.gnu.org/22445.
  706 
  707 	* lib/test-driver (print_usage): space after --test-name,
  708 	--log-file, --trs-file, not =. Also mention Automake as source.
  709 
  710 2020-04-06  Samy Mahmoudi  <samy.mahmoudi@gmail.com>
  711 
  712 	cosmetics: typo in comment.
  713 
  714 	This change fixes https://bugs.gnu.org/32100.
  715 
  716 	* bin/aclocal.in (install_file): remove duplicate "the" in
  717 	"Using the real the destination file ...".
  718 
  719 2020-04-05  Colomban Wendling  <lists.ban@herbesfolles.org>
  720 
  721 	vala: more precise argument matching.
  722 
  723 	This change fixes https://bugs.gnu.org/18734.
  724 
  725 	* bin/automake.in (lang_vala_finish_target): anchor option regexp
  726 	so that, e.g., an argument "vapi" does not match the option --vapi.
  727 	* NEWS: mention this (and preceding checklinkx change, sorry).
  728 
  729 2020-04-03  Karl Berry  <karl@freefriends.org>
  730 
  731 	doc: update urls in manual and include checklinkx script.
  732 
  733 	* doc/automake.texi: update many urls; http -> https,
  734 	search.cpan.org -> metacpan.org/pod/distribution, node names, etc.
  735 	Remove sourceware.org/cgi-bin/gnatsweb.pl and
  736 	miller.emu.id.au/pmiller/books/rmch/ as these are 404
  737 	and no good replacement is evident.
  738 	s/perl/Perl/ a couple times in text for good measure.
  739 	* contrib/checklinkx: new script, a small modification of
  740 	W3C checklink <https://validator.w3.org/checklink/docs/checklink.html>
  741 	(W3C license, which is free software), starting from version 4.81
  742 	installed from CPAN:
  743 	https://metacpan.org/pod/distribution/W3C-LinkChecker/bin/checklink.pod
  744 	* doc/local.mk (checklinkx): new target to invoke it, with variables.
  745 	* Makefile.am (EXTRA_DIST): distribute it.
  746 	* NEWS: mention all this.
  747 
  748 2020-03-24  Karl Berry  <karl@freefriends.org>
  749 
  750 	tests: require etags for tags-lisp-space test.
  751 
  752 	* t/tags-lisp-space.sh (required): set to etags.
  753 
  754 2020-03-24  Karl Berry  <karl@freefriends.org>
  755 
  756 	Merge branch 'master' of git.savannah.gnu.org:/srv/git/automake
  757 
  758 2020-03-24  Karl Berry  <karl@freefriends.org>
  759 
  760 	doc: forgot Python 3 NEWS entries.
  761 
  762 	* NEWS: item for Python 3 support in 1.16.2.
  763 
  764 2020-03-23  Jim Meyering  <meyering@fb.com>
  765 
  766 	maint: Post-release administrivia
  767 
  768 	* NEWS: Add header line for next release.
  769 	* configure.ac (AC_INIT): Bump version number to 1.16b.
  770 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
  771 
  772 2020-03-16  Jim Meyering  <meyering@fb.com>
  773 
  774 	version 1.16.2
  775 
  776 	* configure.ac (AC_INIT): Bump version number to 1.16.2.
  777 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
  778 	* NEWS: Record release version.
  779 
  780 2020-03-16  Karl Berry  <karl@freefriends.org>
  781 
  782 	maint: sync fdl.texi.
  783 
  784 	* doc/fdl.texi: sync with gnulib/doc/fdl.texi, for https://fsf.org
  785 	and some Texinfo fixes.
  786 
  787 2020-03-12  Karl Berry  <karl@freefriends.org>
  788 
  789 	maint: do not descend into unreadable test directories.
  790 
  791 	* maintainer/syntax-checks.mk (sc_tests_makefile_variable_order):
  792 	find t ! -perm o+r -prune, since some test directories
  793 	are intentionally unreadable.
  794 
  795 2020-03-11  Jim Meyering  <meyering@fb.com>
  796 
  797 	doc: gender tweak
  798 
  799 	* HACKING: s/his/their/
  800 
  801 2020-03-11  Jim Meyering  <meyering@fb.com>
  802 
  803 	maint: run "make fetch" to update files from elsewhere
  804 
  805 	* lib/config.guess: Auto-update.
  806 	* lib/config.sub: Auto-update.
  807 	* lib/gendocs.sh: Auto-update.
  808 	* lib/gendocs_template: Auto-update.
  809 	* lib/gitlog-to-changelog: Auto-update.
  810 	* lib/gnupload: Auto-update.
  811 	* lib/texinfo.tex: Auto-update.
  812 	* lib/update-copyright: Auto-update.
  813 
  814 2020-03-11  Karl Berry  <karl@freefriends.org>
  815 
  816 	maint: typo in comment.
  817 
  818 	* t/CheckListOfTests.am (maintainer-check-list-of-tests): an -> and.
  819 
  820 2020-02-25  Karl Berry  <karl@freefriends.org>
  821 
  822 	python: support both Python 2 and 3 in py-compile
  823 
  824 	* lib/py-compile: check python major version and use imp
  825 	or importlib accordingly, plus related changes. Original
  826 	patch for Python 3 only from Gabriel Ganne at:
  827 	https://lists.gnu.org/archive/html/automake-patches/2019-07/msg00002.html
  828 
  829 2020-02-15  Paul Osmialowski  <pawel.osmialowski@arm.com>
  830 
  831 	configure: add flang compilers to _AM_COMPILER_CAN_FAIL
  832 
  833 	* configure.ac (_AM_COMPILER_CAN_FAIL): add armflang and flang
  834 	(in two places). Original patch at:
  835 	https://lists.gnu.org/archive/html/automake-patches/2019-02/msg00001.html
  836 
  837 2020-02-14  Libor Bukata  <libor.bukata@oracle.com>
  838 
  839 	cosmetics: improve error message when dependency tracking fails
  840 
  841 	This change fixes https://bugs.gnu.org/35848.
  842 
  843 	* m4/depout.m4: Add suggestion to try GNU make to the error message.
  844 	Original patch at
  845 	https://lists.gnu.org/archive/html/automake-patches/2019-05/msg00000.html
  846 
  847 2020-02-13  Vincent Lefevre  <vincent@vinc17.net>
  848 
  849 	maint: update lib/COPYING from canonical source.
  850 
  851 	* lib/COPYING: update from https://www.gnu.org/licenses/gpl-3.0.txt
  852 	(only change is to use https urls).
  853 
  854 2020-02-10  Mathieu Lirzin  <mthl@gnu.org>
  855 
  856 	build: fix 'installcheck' target
  857 
  858 	* t/local.mk (installcheck-testsuite): Do not use 'pre-inst-env' script.
  859 	(AM_TESTS_ENVIRONMENT): Ensure that installed perl modules are found.
  860 
  861 2020-01-29  Lukas Fleischer  <lfleischer@lfos.de>
  862 
  863 	python: properly uninstall __pycache__ in subdirectories
  864 
  865 	This change finishes fixing automake bug#32088:
  866 	https://debbugs.gnu.org/32088
  867 
  868 	* lib/am/python.am (uninstall-%DIR%PYTHON): Uninstall
  869 	byte-compiled files in a '__pycache__' subdirectory
  870 	from sub/__pycache__/, not __pycache__/sub/.
  871 	(am__pep3147_tweak): prefix __pycache__ here, for both .pyc and .pyo.
  872 	https://lists.gnu.org/archive/html/automake-patches/2018-05/msg00000.html
  873 	* t/instmany-python.sh (limit): increase to 4500, following
  874 	Andreas Huettel, https://debbugs.gnu.org/32088#11.
  875 
  876 2020-01-29  Lukas Fleischer  <lfleischer@lfos.de>
  877 
  878 	tests: in python tests, do not require .pyo files (for python3)
  879 
  880 	This change partially fixes automake bug#32088:
  881 	https://debbugs.gnu.org/32088
  882 
  883 	* t/py-compile-basedir.sh: Remove all .pyo checks.
  884 	Also, in this test only, allow for either 4 or 6 files resulting
  885 	from compilation, as new-enough Python3 results in 6
  886 	(per Andreas Huettel, https://debbugs.gnu.org/cgi/bugreport.cgi?bug=32088#17)
  887 	* t/py-compile-basic.sh: Remove all .pyo checks.
  888 	* t/py-compile-destdir.sh: Likewise.
  889 	* t/py-compile-option-terminate.sh: Likewise.
  890 	* t/python-virtualenv.sh: Likewise.
  891 	* t/python10.sh: Likewise.
  892 	* t/python12.sh: Likewise.
  893 	* t/python3.sh: Likewise.
  894 
  895 2020-01-26  Karl Berry  <karl@freefriends.org>
  896 
  897 	doc: clarify build tree location
  898 
  899 	* doc/automake.texi (VPATH Builds): more words about the build
  900 	tree location. Suggestion from Jefferson Carpenter, 31 May 2018:
  901 	<https://lists.gnu.org/archive/html/automake-patches/2018-05/msg00003.html>
  902 
  903 2020-01-19  Karl Berry  <karl@freefriends.org>
  904 
  905 	lint: make syntax-check same as maintainer-check.
  906 
  907 	* maintainer/syntax-checks.mk (syntax-check): new target,
  908 	same as maintainer-check. For consistency with many other
  909 	GNU packages.
  910 	(.PHONY): add here too.
  911 
  912 2020-01-19  Karl Berry  <karl@freefriends.org>
  913 
  914 	tests: use find+rm, not perl, to remove temporary directories.
  915 
  916 	This change fixes automake bug#39078:
  917 	https://debbugs.gnu.org/39078
  918 
  919 	* t/ax/test-lib.sh (rm_rf_): run chmod -R u+rwx and rm -rf,
  920 	instead of calling our t/ax/deltree.pl script.  This reverses the
  921 	change of 2013-05-16.  It made sense to write and use deltree
  922 	at that point, but unfortunately as of perl-5.28.0
  923 	(File::Path.pm version 2.15), rmtree no longer removes a tree with
  924 	unreadable subdirectories, such as we (intentionally) have.  So we
  925 	might as well go back to rm -rf.  The unconditional recursive chmod
  926 	instead of the previously-used more complex find command will
  927 	hopefully prove portable.  See the bug report for more details.
  928 
  929 	Our deltree.pl is not used for anything else, but nevertheless
  930 	leaving it in our source tree for now, for ease of reversion
  931 	and comparison.
  932 
  933 2020-01-08  Karl Berry  <karl@freefriends.org>
  934 
  935 	tests: use skip_ consistently.
  936 
  937 	* t/lisp-loadpath.sh: use skip_ instead of exit 77;
  938 	also, notice and reset if the EMACS envvar is simply "t".
  939 	* t/uninstall-fail.sh: use skip_ instead of the undefined skip.
  940 
  941 2020-01-05  Karl Berry  <karl@freefriends.org>
  942 
  943 	tests: Correctly simulate no emacs in t/nobase-nodist.sh
  944 
  945 	* t/nobase-nodist.sh: export EMACS, not EMCAS, as "no".
  946 
  947 2020-01-04  Karl Berry  <karl@freefriends.org>
  948 
  949 	automake: Support byte compilation in older Emacsen
  950 
  951 	* lib/am/lisp.am (am__emacs_byte_compile_setup) [FIRST]: define
  952 	new make variable, to use byte-compile-dest-file-function if
  953 	available, else byte-compile-dest-file.
  954 	(.el.elc): use it.
  955 	* t/lisp-loadpath.sh: skip test if emacs version is <= 23,
  956 	since their -L ordering is backwards.
  957 	* NEWS: update.
  958 
  959 2020-01-01  Jim Meyering  <meyering@fb.com>
  960 
  961 	maint: make update-copyright
  962 
  963 2019-12-23  Jim Meyering  <meyering@fb.com>
  964 
  965 	maint: make maintainer-check tests pass
  966 
  967 	* maintainer/syntax-checks.mk (sc_sanity_gnu_grep): Remove
  968 	NUL byte from grep output, to avoid shell diagnostic about
  969 	"NUL byte suppressed from expansion."
  970 	(automake_diff_no, aclocal_diff_no): Adjust number of expected
  971 	diff lines.
  972 
  973 2019-12-21  Karl Berry  <karl@freefriends.org>
  974 
  975 	tests: Ensure UTC for Texinfo @UPDATED@ test
  976 
  977 	* t/txinfo-vtexi4.sh: TZ=UTC0; export TZ (same as mdate-sh).
  978 	* NEWS: update.
  979 
  980 2019-12-17  Karl Berry  <karl@freefriends.org>
  981 
  982 	automake: Ensure space after $(LISP) in tags dependencies.
  983 
  984 	This change fixes automake bug#38139.
  985 
  986 	* bin/automake.in (handle_tags): Concatenate with space in " @config".
  987 	* t/tags-lisp-space.sh: New test.
  988 	* t/list-of-tests.mk (handwritten_TESTS): Add it.
  989 
  990 2019-11-17  Giuseppe Scrivano  <gscrivan@redhat.com>
  991 
  992 	dist: add dist-zstd option
  993 
  994 	Add support for using the zstd compression algorithm.
  995 	Use a default compression setting of -19, and ".zst" as the suffix.
  996 
  997 	* bin/automake.in (handle_dist): Add zstd to the list of known dist-
  998 	suffixes.
  999 	(preprocess_file): Map ZSTD to dist-zstd.
 1000 	* doc/automake.texi: Document the new option.
 1001 	* lib/Automake/Options.pm (_is_valid_easy_option): Add dist-zstd.
 1002 	* lib/am/distdir.am (dist-zstd): New rule.
 1003 	(?ZSTD?DIST_TARGETS): Add definition.
 1004 	(distcheck): Add a case for *.tar.zst*.
 1005 	* t/dist-formats.tap: Add tests.
 1006 	* NEWS: Mention the change.
 1007 
 1008 2019-11-11  Paul Eggert  <eggert@cs.ucla.edu>
 1009 
 1010 	config: add msys support
 1011 
 1012 	Requested by Arnold Robbins in:
 1013 	https://lists.gnu.org/r/bug-gnulib/2019-11/msg00008.html
 1014 	* lib/ar-lib, lib/compile (func_file_conv):
 1015 	Treat msys like cygwin.
 1016 
 1017 2019-10-14  Paul Eggert  <eggert@cs.ucla.edu>
 1018 
 1019 	maint: make update-copyright
 1020 
 1021 2019-10-14  Paul Eggert  <eggert@cs.ucla.edu>
 1022 
 1023 	maint: make fetch
 1024 
 1025 2019-09-02  Gavin Smith  <gavinsmith0123@gmail.com>
 1026 
 1027 	automake: do not require @setfilename in Texinfo files
 1028 
 1029 	Texinfo no longer requires a @setfilename directive in each
 1030 	.texi file, so automake now also relaxes its restriction.
 1031 	* bin/automake.in (scan_texinfo_file): Derive name of info file from
 1032 	name of input file if no @setfilename line occurs in the file.
 1033 	* t/txinfo-no-setfilename.sh: New test.
 1034 	* t/list-of-tests.mk: Add it.
 1035 	* NEWS: Mention it.
 1036 
 1037 	Fixes automake bugs #36921 and #34201.
 1038 
 1039 2019-08-26  Paul Eggert  <eggert@cs.ucla.edu>
 1040 
 1041 	install-sh: support -s with read-only source
 1042 
 1043 	Problem reported by Karl Berry in:
 1044 	https://lists.gnu.org/r/bug-gnulib/2019-08/msg00067.html
 1045 	* lib/install-sh: If -s is given, create the temporary file
 1046 	with $cp_umask so that ‘strip’ can write to it.
 1047 
 1048 2019-07-03  Bruno Haible  <bruno@clisp.org>
 1049 
 1050 	ar-lib: Fix for MSVC 14.
 1051 
 1052 	* lib/ar-lib (extract): Convert CRLFs in 'lib -NOLOGO -LIST' output to LFs.
 1053 
 1054 2018-10-23  Paul Hardy  <unifoundry@gmail.com>
 1055 
 1056 	doc: Fix various typos and phrasing
 1057 
 1058 	This change fixes automake bug#32150.
 1059 
 1060 	* doc/automake.texi: Various typos and phrasing changes.
 1061 
 1062 2018-07-08  Mathieu Lirzin  <mthl@gnu.org>
 1063 
 1064 	python: Don't use '\n' in sed substitution
 1065 
 1066 	This change fixes automake bug#31222.
 1067 
 1068 	On macOS, 'sed' interprets '\n' in the substitution text as the letter 'n'
 1069 	instead of as a newline.
 1070 
 1071 	* lib/am/python.am [?FIRST?] (am__pep3147_tweak): Use a space instead of '\n'.
 1072 	* NEWS: Update.
 1073 
 1074 2018-03-24  Mathieu Lirzin  <mthl@gnu.org>
 1075 
 1076 	test-driver.scm: Add "--coverage" option
 1077 
 1078 	* contrib/test-driver.scm: When 'coverage' option is enabled, run tests
 1079 	in the debug vm and trace coverage data.
 1080 	(%options): Add 'coverage'.
 1081 	(show-help): Display option.
 1082 
 1083 2018-03-24  Mathieu Lirzin  <mthl@gnu.org>
 1084 
 1085 	test-driver.scm: Don't guess script name from "--test-name"
 1086 
 1087 	'primitive-load' is used instead of 'load-from-path' since the script is
 1088 	given as a relative file name.  For unknown reason, using 'load' fails
 1089 	with GNU Mcron test suite when running 'make distcheck'.
 1090 
 1091 	* contrib/test-driver.scm: Get the actual script name directly from the
 1092 	command line.  Handle the case where that argument is missing.
 1093 
 1094 2018-03-24  Mathieu Lirzin  <mthl@gnu.org>
 1095 
 1096 	test-driver.scm: Inline 'main' procedure
 1097 
 1098 	Having a main procedure involves passing the '-e main' command-line
 1099 	argument to 'guile' which makes the test driver a bit less easy to use.
 1100 
 1101 	* contrib/test-driver.scm (main): Delete.  Move body to the top-level.
 1102 
 1103 2018-03-11  Mathieu Lirzin  <mthl@gnu.org>
 1104 
 1105 	maint: Use 'before-save-hook' in "contrib/test-driver.scm"
 1106 
 1107 	This a follow-up to commit 6bab5b26a1241b0e0edd058d2a921989f9a1747c.
 1108 
 1109 	Use 'before-save-hook' instead of 'write-file-functions' to match what
 1110 	is done in other scripts.
 1111 
 1112 	* contrib/test-driver.scm: Update hook usage and use 'UTC0' timezone
 1113 	instead of 'UTC'.
 1114 
 1115 2018-03-11  Mathieu Lirzin  <mthl@gnu.org>
 1116 
 1117 	bin: Rely only on the shebang line
 1118 
 1119 	Previously ‘automake’ and ‘aclocal’ were handling the case of being
 1120 	interpreted as a Shell script by using a hack leveraging the fact that
 1121 	Shell and Perl has a compatible syntax intersection allowing those
 1122 	scripts to launch ‘perl’ from the shell.
 1123 
 1124 	* bin/aclocal.in: Remove cryptic launching hack.
 1125 	* bin/automake.in: Likewise.
 1126 
 1127 2018-03-11  Mathieu Lirzin  <mthl@gnu.org>
 1128 
 1129 	maint: Post-release administrivia
 1130 
 1131 	* NEWS: Add header line for next release.
 1132 	* configure.ac (AC_INIT): Bump version number to 1.16a.
 1133 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 1134 
 1135 2018-03-11  Mathieu Lirzin  <mthl@gnu.org>
 1136 
 1137 	version 1.16.1
 1138 
 1139 	* configure.ac (AC_INIT): Bump version number to 1.16.1.
 1140 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 1141 	* NEWS: Record release version.
 1142 
 1143 2018-03-11  Mathieu Lirzin  <mthl@gnu.org>
 1144 
 1145 	maint: Update files from upstream with 'make fetch'
 1146 
 1147 	* lib/config.guess: Update.
 1148 	* lib/config.sub: Likewise.
 1149 	* lib/gendocs.sh: Likewise.
 1150 	* lib/gitlog-to-changelog: Likewise.
 1151 	* lib/gnupload: Likewise.
 1152 	* lib/texinfo.tex: Likewise.
 1153 	* lib/update-copyright: Likewise.
 1154 
 1155 2018-03-11  Pavel Raiskup  <praiskup@redhat.com>
 1156 
 1157 	install-sh: avoid (low risk) race in "/tmp"
 1158 
 1159 	Ensure that nobody can cross privilege boundaries by pre-creating
 1160 	symlink on '$tmpdir' destination directory.
 1161 
 1162 	Just testing 'mkdir -p' by creating "/tmp/ins$RANDOM-$$/d" is not safe
 1163 	because "/tmp" directory is usually world-writeable and
 1164 	"/tmp/ins$RANDOM-$$" content could be pretty easily guessed by
 1165 	attacker (at least for shells where $RANDOM is not supported).  So, as
 1166 	the first step, create the "/tmp/ins$RANDOM-$$" without -p.  This step
 1167 	would fail early if somebody wanted catch us.
 1168 
 1169 	Systems that implement (and have enabled) fs.protected_symlinks kernel
 1170 	feature are not affected even without this commit.
 1171 
 1172 	References:
 1173 	https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=760455
 1174 	https://bugzilla.redhat.com/show_bug.cgi?id=1140725
 1175 
 1176 	* lib/install-sh: Implement safer 'mkdir -p' test by running
 1177 	'$mkdirprog $mkdir_mode "$tmpdir"' first.
 1178 	* NEWS: Update.
 1179 
 1180 2018-03-08  Mathieu Lirzin  <mthl@gnu.org>
 1181 
 1182 	automake: Don't rely on List::Util to provide 'none'
 1183 
 1184 	This change fixes automake bug#30631.
 1185 
 1186 	This removes the use of List::Util which is not supported by Perl 5.6,
 1187 	by reimplementing the 'none' subroutine.
 1188 
 1189 	* lib/Automake/General.pm (none): New subroutine.
 1190 	* bin/automake.in (handle_single_transform): Use it.
 1191 	* t/pm/General.pl: New test.
 1192 	* t/list-of-tests.mk (perl_TESTS): Add it.
 1193 	* NEWS: Update.
 1194 
 1195 2018-03-08  Mathieu Lirzin  <mthl@gnu.org>
 1196 
 1197 	python: Support future python version up to 3.9
 1198 
 1199 	This change fixes automake bug#28160.
 1200 
 1201 	Since AM_PYTHON_PATH macro takes no maximum version argument, there is
 1202 	no need to generate _AM_PYTHON_INTERPRETER_LIST dynamically, like what
 1203 	was previously done by the reverted commit
 1204 	1d60fb72168e62d33fe433380af621de64e22f23.  We could rely on M4 to
 1205 	generate this list statically however this is likely to be a complex
 1206 	solution that would not improve maintainability.
 1207 
 1208 	* m4/python.m4 (_AM_PYTHON_INTERPRETER_LIST): Add 'python3.7',
 1209 	'python3.8', and 'python3.9'.
 1210 	* NEWS: Update.
 1211 
 1212 2018-03-06  Paul Eggert  <eggert@cs.ucla.edu>
 1213 
 1214 	maint: write-file-hooks -> before-save-hook
 1215 
 1216 	write-file-hooks is obsolete since Emacs 22.1 (released June 2007) and
 1217 	it's time to use the recommended replacement.
 1218 	Problem reported by Glenn Morris in:
 1219 	https://lists.gnu.org/r/bug-gnulib/2018-03/msg00008.html
 1220 	* contrib/tap-driver.pl, lib/compile, lib/depcomp, lib/install-sh:
 1221 	* lib/mdate-sh, lib/missing, lib/mkinstalldirs, lib/py-compile:
 1222 	* lib/tap-driver.sh, lib/test-driver, lib/ylwrap:
 1223 	Update hook usage for files where Automake is the canonical source.
 1224 
 1225 2018-03-03  Mathieu Lirzin  <mthl@gnu.org>
 1226 
 1227 	Revert "python: Generate python interpreter list"
 1228 
 1229 	This reverts commit 1d60fb72168e62d33fe433380af621de64e22f23.
 1230 
 1231 2018-02-25  Mathieu Lirzin  <mthl@gnu.org>
 1232 
 1233 	maint: Post-release administrivia
 1234 
 1235 	* NEWS: Add header line for next release.
 1236 	* configure.ac (AC_INIT): Bump version number to 1.16a.
 1237 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 1238 
 1239 2018-02-25  Mathieu Lirzin  <mthl@gnu.org>
 1240 
 1241 	Merge branch 'release'
 1242 
 1243 2018-02-25  Mathieu Lirzin  <mthl@gnu.org>
 1244 
 1245 	version 1.16
 1246 
 1247 	* configure.ac (AC_INIT, APIVERSION): Bump version number to 1.16.
 1248 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 1249 	* NEWS: Record release version.
 1250 
 1251 2018-02-25  Mathieu Lirzin  <mthl@gnu.org>
 1252 
 1253 	maint: Update INSTALL
 1254 
 1255 	* INSTALL: Use single quotes.
 1256 
 1257 2018-02-25  Mathieu Lirzin  <mthl@gnu.org>
 1258 
 1259 	tests: Remove tests intended only for the 'next' branch
 1260 
 1261 	The commit 199e7a445040270fa5ef67623c56cde40d765199 "Prefer https: URLS"
 1262 	which is a cherry-pick of b09d945b795ab9deed2bc457289cd5f41c506b50 from
 1263 	'next' to 'master' has mistakenly copied some tests only present on the
 1264 	'next' branch.
 1265 
 1266 	* t/am-prog-mkdir-p.sh: Delete.
 1267 	* t/txinfo-no-split.sh: Likewise.
 1268 
 1269 2018-02-25  Mathieu Lirzin  <mthl@gnu.org>
 1270 
 1271 	maint: Update files from upstream with 'make fetch'
 1272 
 1273 	* lib/config.guess: Update
 1274 	* lib/config.sub: Likewise.
 1275 	* lib/gendocs.sh: Likewise.
 1276 	* lib/gendocs_template: Likewise.
 1277 	* lib/gitlog-to-changelog: Likewise.
 1278 	* lib/gnupload: Likewise.
 1279 	* lib/texinfo.tex: Likewise.
 1280 	* lib/update-copyright: Likewise.
 1281 
 1282 2018-02-18  Mathieu Lirzin  <mthl@gnu.org>
 1283 
 1284 	python: Avoid exceeding command-line length limit
 1285 
 1286 	With Python implementations following PEP-3174, a large number of files
 1287 	are installed in the ‘__pycache__’ directory.  As a consequence
 1288 	“t/instmany-python.sh” test was failing due to the
 1289 	‘uninstall-pythonPYTHON’ target deleting installed files in a single
 1290 	‘rm’ command.  Doing that in multiple steps avoids exceeding the
 1291 	command-line length limit.  This fixes bug#30335.
 1292 
 1293 	* lib/am/python.am (uninstall-%DIR%PYTHON): For byte-compiled files
 1294 	installed in '__pycache__' directory, uninstall them by batch of 40.
 1295 	[?FIRST?] (am__pep3147_tweak): Adapt.
 1296 
 1297 2018-02-18  Mathieu Lirzin  <mthl@gnu.org>
 1298 
 1299 	tests: Don't check 'Getopt::Long' corner cases
 1300 
 1301 	Depending on the installed 'Getopt::Long' perl module, command-line
 1302 	handling may vary a bit.  As a consequence we prefer not to check
 1303 	command-line corners cases.  This change fixes automake bug#29638.
 1304 
 1305 	* t/aclocal.sh (am_create_testdir): Don't expect "--versi" to be
 1306 	interpreted as "--version".
 1307 	* t/automake-cmdline.tap: Don't expect "--vers" to be interpreted as
 1308 	"--version" and things after "--" to be interpreted as file arguments.
 1309 	(do_check): Display the actual command output.
 1310 	* t/maken3.sh (check_targets): "--force" is not a documented option, so
 1311 	don't use it.
 1312 
 1313 2018-02-03  Mathieu Lirzin  <mthl@gnu.org>
 1314 
 1315 	python: Generate python interpreter list
 1316 
 1317 	_AM_PYTHON_INTERPRETER_LIST is used by AM_PYTHON_PATH to autodetect
 1318 	Python programs whose names correspond to a specific Python
 1319 	version (e.g. python3.6).  Previously this list was updated manually.
 1320 	The automatic support of newer versions (up to 4.0 excluded) fixes
 1321 	bug#28160.
 1322 
 1323 	* m4/python.m4 (am_py_min_ver, am_py_max_ver): New macros.
 1324 	(_AM_PYTHON_INTERPRETER_LIST): Generate this list instead of hard-coding
 1325 	it.  Implementation is taken from GNU Pyconfigure.
 1326 
 1327 2018-01-19  Mathieu Lirzin  <mthl@gnu.org>
 1328 
 1329 	tests: Improve comment in 'txinfo-many-output-formats.sh'
 1330 
 1331 	* t/txinfo-many-output-formats.sh: Explain why AM_MAKEINFOFLAGS is set
 1332 	with an invalid option.
 1333 
 1334 2018-01-18  Mathieu Lirzin  <mthl@gnu.org>
 1335 
 1336 	tests: Distribute DEJATOOL files manually
 1337 
 1338 	"t/check12.sh" was failing because files declared in DEJATOOL are not
 1339 	automatically distributed.  The test running 'make distcheck' couldn't succeed
 1340 	since some scripts were not distributed.  This fixes automake bug#26738.
 1341 
 1342 	* t/check12.sh (Makefile.am): Distribute files from DEJATOOL.
 1343 
 1344 2018-01-18  Mathieu Lirzin  <mthl@gnu.org>
 1345 
 1346 	tests: Let 'ltorder.sh' run successfully with Guix dynamic loader
 1347 
 1348 	* t/ltorder.sh: Set GUIX_LD_WRAPPER_ALLOW_IMPURITIES environment variable to
 1349 	unlock the dynamic loader provided by GNU Guix.
 1350 
 1351 2018-01-18  Mathieu Lirzin  <mthl@gnu.org>
 1352 
 1353 	tests: Fix various 'flex' compilation issues
 1354 
 1355 	* t/lex-clean-cxx.sh (parsefoo.lxx): Declare 'yylex':
 1356 	(mainfoo.cc): Make declaration compatible with C++.
 1357 	* t/lex-depend-cxx.sh (joe.ll): Declare 'yylex'.
 1358 	* t/silent-many-languages.sh (Makefile.am, sub/Makefile.am): Link -lfl only
 1359 	with 'fo2' and 'sub/ba2' which are the only program needing it.
 1360 	(foo5.l): Define 'isatty'.
 1361 	(foo6.y): Declare 'yylex'.
 1362 
 1363 2018-01-18  Mathieu Lirzin  <mthl@gnu.org>
 1364 
 1365 	tests: Check GCS conformance of 'aclocal' command-line interface
 1366 
 1367 	* t/aclocal.sh: Check that 'aclocal' support the --version and --help
 1368 	command-line options.
 1369 
 1370 2018-01-17  Mathieu Lirzin  <mthl@gnu.org>
 1371 
 1372 	contrib: Add Guile custom test driver using SRFI-64 test harness
 1373 
 1374 	This sets a home for a script already used by GNU Guix and GNU Mcron.
 1375 
 1376 	* contrib/test-driver.scm: New test driver script.
 1377 	* NEWS: Update.
 1378 
 1379 2018-01-17  Mathieu Lirzin  <mthl@gnu.org>
 1380 
 1381 	maint: Document how to use Guix for Automake development
 1382 
 1383 	* HACKING <Setting the development environment>: New part.
 1384 
 1385 2018-01-04  Khem Raj  <raj.khem@gmail.com>  (tiny change)
 1386 
 1387 	automake: Add default libtool_tag to cppasm
 1388 
 1389 	* bin/automake.in (register_language): Define default libtool tag to be
 1390 	CC since CPPASCOMPILE is using CC to call assembler.
 1391 
 1392 2018-01-04  Bruno Haible  <bruno@clisp.org>
 1393 
 1394 	doc: Document the portability of various tar formats better
 1395 
 1396 	* doc/automake.texi (List of Automake options): Document the portability of
 1397 	the tar-ustar and tar-pax options better.
 1398 
 1399 2018-01-04  Simon Sobisch  <simonsobisch@web.de>  (tiny change)
 1400 
 1401 	missing: Update displayed URLs
 1402 
 1403 	* lib/missing (perl_URL): Use HTTPS.
 1404 	(flex_URL): Use new Github homepage.
 1405 
 1406 2018-01-04  Mathieu Lirzin  <mthl@gnu.org>
 1407 
 1408 	maint: Update copyright years to 2018
 1409 
 1410 	This update has been made with 'make update-copyright'.
 1411 
 1412 2018-01-04  Mathieu Lirzin  <mthl@gnu.org>
 1413 
 1414 	maint: Exclude ".dir-locals.el" from copyright updates
 1415 
 1416 	* maintainer/maint.mk (files_without_copyright): Add ".dir-locals.el".
 1417 
 1418 2017-12-16  Jim Meyering  <meyering@fb.com>
 1419 
 1420 	port elisp-compilation support to emacs-23.1 and newer
 1421 
 1422 	In May of 2017, emacs.master support for using the long-deprecated
 1423 	byte-compile-dest-file function was removed, and that removal broke
 1424 	automake's elisp-compiling rule for any .el file not in the current
 1425 	directory.  In emacs-23.1 (July 2009) byte-compile-dest-file-function
 1426 	became the recommended way to adjust the byte-compiler's destination.
 1427 	The removed functionality has been restored for Emacs-26, albeit with
 1428 	dissuasive diagnostics warning about the imminent removal of this
 1429 	functionality.  It will be removed in Emacs-27.
 1430 	* lib/am/lisp.am (.el.elc): Use byte-compile-dest-file-function, rather
 1431 	than byte-compile-dest-file.  Also, use "-f batch-byte-compile '$<'"
 1432 	rather than open-coding it, as suggested by Glenn Morris.
 1433 	* t/lisp-readonly-srcdir.sh: New file, to test for the above.
 1434 	* t/list-of-tests.mk (handwritten_TESTS): Add it.
 1435 	* NEWS (Bugs fixed): Mention this problem.
 1436 
 1437 2017-11-28  Jim Meyering  <meyering@fb.com>
 1438 
 1439 	"make dist" did not depend on $(BUILT_SOURCES)
 1440 
 1441 	* lib/am/distdir.am (distdir-am): New intermediate target.
 1442 	Interpose this target between $(distdir) and its dependency
 1443 	on $(DISTFILES), so that we can ensure $(BUILT_SOURCES) are
 1444 	all created before we begin creating $(DISTFILES).
 1445 	* t/dist-vs-built-sources.sh: Test for this.
 1446 	* t/list-of-tests.mk (handwritten_TESTS): Add it.
 1447 	* NEWS (Bugs fixed): Mention it.
 1448 	Assaf Gordon reported that "make dist" (after ./configure
 1449 	from a pristine clone of GNU hello) would fail due to the
 1450 	absence of configmake.h while compiling lib/localcharset.c.
 1451 	https://lists.gnu.org/r/bug-hello/2014-03/msg00016.html
 1452 
 1453 2017-11-24  Mathieu Lirzin  <mthl@gnu.org>
 1454 
 1455 	maint: Update HACKING
 1456 
 1457 	* HACKING (Working with git): Remove reference to the 'micro' branch and
 1458 	adapt branch descriptions to the current branching scheme.
 1459 
 1460 2017-09-23  Mathieu Lirzin  <mthl@gnu.org>
 1461 
 1462 	maint: Make Emacs use 'makefile-automake-mode'
 1463 
 1464 	* bin/local.mk: Specify mode name in the first line.
 1465 	* contrib/t/local.mk: Likewise.
 1466 	* doc/local.mk: Likewise.
 1467 	* lib/Automake/local.mk: Likewise.
 1468 	* lib/am/local.mk: Likewise.
 1469 	* lib/local.mk: Likewise.
 1470 	* m4/local.mk: Likewise.
 1471 	* t/local.mk: Likewise.
 1472 
 1473 2017-09-23  Paul Eggert  <eggert@cs.ucla.edu>
 1474 
 1475 	maint: update .gitignore
 1476 
 1477 	* .gitignore: Add pre-inst-env, and sort.
 1478 
 1479 2017-09-23  Paul Eggert  <eggert@cs.ucla.edu>
 1480 
 1481 	install-sh: do not assume / = //
 1482 
 1483 	* lib/install-sh: Do not append / to destination
 1484 	directory if it already ends in /.  This supports
 1485 	a destination directory of // on hosts where / and //
 1486 	are distinct directories, as POSIX allows.
 1487 
 1488 2017-09-23  Paul Eggert  <eggert@cs.ucla.edu>
 1489 
 1490 	maint: fix two more http: URLs
 1491 
 1492 	* m4/init.m4: Change http: to https: in comments.
 1493 
 1494 2017-09-22  Mathieu Lirzin  <mthl@gnu.org>
 1495 
 1496 	maint: Configure Emacs automatically with ".dir-locals.el"
 1497 
 1498 	* .dir-locals.el: New Emacs directory configuration file.  All perl
 1499 	files adapted.
 1500 
 1501 2017-09-22  Mathieu Lirzin  <mthl@gnu.org>
 1502 
 1503 	maint: 'master' should be merged into 'next'
 1504 
 1505 	* HACKING: Fix instructions merge instructions.
 1506 
 1507 2017-09-21  Paul Eggert  <eggert@cs.ucla.edu>
 1508 
 1509 	* lib/mdate.sh (TZ): Use portable setting.
 1510 
 1511 2017-09-19  Mathieu Lirzin  <mthl@gnu.org>
 1512 
 1513 	Merge branch 'minor'
 1514 
 1515 2017-09-19  Mathieu Lirzin  <mthl@gnu.org>
 1516 
 1517 	maint: Document new Branch names convention
 1518 
 1519 	* HACKING: Update to new Branch names.
 1520 
 1521 2017-09-19  Mathieu Lirzin  <mthl@gnu.org>
 1522 
 1523 	maint: Reset master
 1524 
 1525 2017-09-19  Mathieu Lirzin  <mthl@gnu.org>
 1526 
 1527 	Merge branch 'minor'
 1528 
 1529 2017-09-19  Paul Eggert  <eggert@cs.ucla.edu>
 1530 
 1531 	make fetch
 1532 
 1533 2017-09-19  Paul Eggert  <eggert@cs.ucla.edu>
 1534 
 1535 	Fix a couple more http: URLs
 1536 
 1537 2017-09-19  Paul Eggert  <eggert@cs.ucla.edu>
 1538 
 1539 	Prefer https: URLs
 1540 
 1541 	In Gnulib, Emacs, etc. we are changing ftp: and http: URLs to use
 1542 	https:, to discourage man-in-the-middle attacks when downloading
 1543 	software. The attached patch propagates these changes upstream to
 1544 	Automake.  This patch does not affect files that Automake is
 1545 	downstream of, which I'll patch separately.
 1546 
 1547 	Althouth the resources are not secret, plain HTTP is vulnerable to
 1548 	malicious routers that tamper with responses from GNU servers,
 1549 	and this sort of thing is all too common when people in some other
 1550 	countries browse US-based websites. See, for example:
 1551 
 1552 	Aceto G, Botta A, Pescapé A, Awan MF, Ahmad T, Qaisar
 1553 	S. Analyzing internet censorship in Pakistan. RTSI
 1554 	2016. https://dx.doi.org/10.1109/RTSI.2016.7740626
 1555 
 1556 	HTTPS is not a complete solution here, but it can be a significant
 1557 	help. The GNU project regularly serves up code to users, so we should
 1558 	take some care here.
 1559 
 1560 2017-09-16  Paul Eggert  <eggert@cs.ucla.edu>
 1561 
 1562 	make fetch
 1563 
 1564 2017-09-16  Paul Eggert  <eggert@cs.ucla.edu>
 1565 
 1566 	Fix a couple more http: URLs
 1567 
 1568 2017-09-16  Paul Eggert  <eggert@cs.ucla.edu>
 1569 
 1570 	Prefer https: URLs
 1571 
 1572 	In Gnulib, Emacs, etc. we are changing ftp: and http: URLs to use
 1573 	https:, to discourage man-in-the-middle attacks when downloading
 1574 	software. The attached patch propagates these changes upstream to
 1575 	Automake.  This patch does not affect files that Automake is
 1576 	downstream of, which I'll patch separately.
 1577 
 1578 	Althouth the resources are not secret, plain HTTP is vulnerable to
 1579 	malicious routers that tamper with responses from GNU servers,
 1580 	and this sort of thing is all too common when people in some other
 1581 	countries browse US-based websites. See, for example:
 1582 
 1583 	Aceto G, Botta A, Pescapé A, Awan MF, Ahmad T, Qaisar
 1584 	S. Analyzing internet censorship in Pakistan. RTSI
 1585 	2016. https://dx.doi.org/10.1109/RTSI.2016.7740626
 1586 
 1587 	HTTPS is not a complete solution here, but it can be a significant
 1588 	help. The GNU project regularly serves up code to users, so we should
 1589 	take some care here.
 1590 
 1591 2017-09-15  Mathieu Lirzin  <mthl@gnu.org>
 1592 
 1593 	maint: Document how to handle assignment exemption
 1594 
 1595 	* HACKING: Update.
 1596 
 1597 2017-09-15  Reiner Herrmann  <reiner@reiner-h.de>  (tiny change)
 1598 
 1599 	mdate-sh: Ensure reproducible time output
 1600 
 1601 	This change fixes automake bug#20314.
 1602 
 1603 	'mdate-sh' pretty-prints the modification time of a file.  But it's
 1604 	output can vary depending on the timezone of the caller. Someone in
 1605 	timezone GMT-12 will get a different result (day) than someone in
 1606 	timezone GMT+12.  As this output is also used to create/update stamp
 1607 	files, which influence the further build process, the build result can
 1608 	vary.
 1609 
 1610 	* lib/mdate-sh: Set 'TZ' to UTC which ensures reproducible output.
 1611 	* NEWS: Announce bug fix.
 1612 
 1613 2017-09-15  Michael Haubenwallner  <michael.haubenwallner@ssi-schaefer.com>
 1614 
 1615 	automake: Depend on LIBOBJDIR for LIBOBJS and ALLOCA
 1616 
 1617 	This change fixes automake bug#27781.
 1618 
 1619 	* bin/automake.in: Add Makefile dependency on LIBOBJDIR/dirstamp for
 1620 	each LIBOBJS/ALLOCA source file found.
 1621 	* t/list-of-tests.mk (XFAIL_TESTS): Drop
 1622 	t/libobj-no-dependency-tracking.sh.
 1623 	* NEWS: Announce bug fix.
 1624 
 1625 2017-09-14  Mathieu Lirzin  <mthl@gnu.org>
 1626 
 1627 	Revert "automake: Handle LTLIBOBJS more specifically"
 1628 
 1629 	This reverts commit 5521219348c55af354878583b99c5f9d66d6d38a.
 1630 
 1631 2017-08-31  Mathieu Lirzin  <mthl@gnu.org>
 1632 
 1633 	Merge branch 'minor'
 1634 
 1635 2017-08-31  Mathieu Lirzin  <mthl@gnu.org>
 1636 
 1637 	build: Rename "Makefile.inc" Makefile fragments to "local.mk"
 1638 
 1639 	This is done to follow a convention used by a lot of GNU packages.
 1640 
 1641 	* bin/Makefile.inc: Rename to ...
 1642 	* bin/local.mk: ... this.
 1643 	* doc/Makefile.inc: Rename to ...
 1644 	* doc/local.mk: ... this.
 1645 	* lib/Automake/Makefile.inc: Rename to ...
 1646 	* lib/Automake/local.mk: ... this.
 1647 	* lib/am/Makefile.inc: Rename to ...
 1648 	* lib/am/local.mk: ... this.
 1649 	* lib/Makefile.inc: Rename to ...
 1650 	* lib/local.mk: ... this.
 1651 	* m4/Makefile.inc: Rename to ...
 1652 	* m4/local.mk: ... this.
 1653 	* contrib/t/Makefile.inc: Rename to ...
 1654 	* contrib/t/local.mk: ... this.
 1655 	* t/Makefile.inc: Rename to ...
 1656 	* t/local.mk: ... this.  Adapt.
 1657 	* Makefile.am: Adapt.
 1658 
 1659 2017-08-31  Mathieu Lirzin  <mthl@gnu.org>
 1660 
 1661 	Merge branch 'micro' into minor
 1662 
 1663 2017-08-31  Mathieu Lirzin  <mthl@gnu.org>
 1664 
 1665 	maint: Define API version manually
 1666 
 1667 	* configure.ac (APIVERSION): Define it to 1.15a.
 1668 	* bootstrap (APIVERSION): Parse 'configure.ac' to get it.
 1669 
 1670 2017-08-31  Mathieu Lirzin  <mthl@gnu.org>
 1671 
 1672 	build: Inline perl prototypes in sources
 1673 
 1674 	Prototypes allows us to avoid using the '&foo' invocation form when
 1675 	invoking a subroutine before its definition.  Previously those
 1676 	prototypes were generated to prevent them from falling out-of-sync
 1677 	with actual definitions.  Now we provide a 'check-perl-protos' lint
 1678 	script to ensure that this is not the case.  This has the same
 1679 	benefits as generating prototypes while simplifying the
 1680 	bootstrap/build process.
 1681 
 1682 	* bin/gen-perl-protos: Remove.
 1683 	* bin/Makefile.inc: Adapt.
 1684 	* bootstrap: Likewise.
 1685 	* bin/aclocal.in: Inline prototypes.
 1686 	* bin/automake.in: Likewise.
 1687 	* maintainer/check-perl-protos: New lint script.
 1688 	* maintainer/syntax-checks.mk (sc_perl_protos): New target.
 1689 	(syntax_check_rules): Add it.
 1690 
 1691 2017-08-30  Mathieu Lirzin  <mthl@gnu.org>
 1692 
 1693 	build: Generate versioned scripts at make time
 1694 
 1695 	* configure.ac: Don't generate 'bin/aclocal-${APIVERSION}' and
 1696 	'bin/automake-${APIVERSION}'.
 1697 	* bin/wrap-aclocal.in: Delete.
 1698 	* bin/wrap-automake.in: Likewise.
 1699 	* bin/Makefile.inc (%D%/automake-$(APIVERSION))
 1700 	(%D%/aclocal-$(APIVERSION)): New targets.
 1701 	(CLEANFILES): Add them.
 1702 
 1703 2017-08-30  Mathieu Lirzin  <mthl@gnu.org>
 1704 
 1705 	aclocal: Support ACLOCAL_AUTOMAKE_DIR environment variable
 1706 
 1707 	* bin/aclocal.in: Reset '@automake_includes' and '@system_includes' in
 1708 	for build environment.  Allow setting '@automake_includes' with
 1709 	ACLOCAL_AUTOMAKE_DIR environment variable.
 1710 	(parse_arguments): Ignore 'dirlist' when '@system_includes' is empty.
 1711 	* doc/automake.texi (aclocal Options): Document ACLOCAL_AUTOMAKE_DIR.
 1712 	* bin/wrap-aclocal.in: Remove extra command line options.
 1713 	* pre-inst-env.in: Set ACLOCAL_AUTOMAKE_DIR and ACLOCAL_PATH
 1714 	environment variables.
 1715 	* t/ax/test-defs.in: Adapt.
 1716 	* t/ansi2knr-no-more.sh (warn_rx): Likewise.
 1717 
 1718 2017-08-30  Mathieu Lirzin  <mthl@gnu.org>
 1719 
 1720 	config: Support AUTOMAKE_LIBDIR environment variable
 1721 
 1722 	* lib/Automake/Config.in: Let AUTOMAKE_LIBDIR environment variable
 1723 	override the default location for '$libdir'.
 1724 	* doc/automake.texi (automake Invocation): Document AUTOMAKE_LIBDIR.
 1725 	* pre-inst-env.in (AUTOMAKE_LIBDIR): Set AUTOMAKE_LIBDIR.
 1726 	* bin/wrap-automake.in: Don't use "--libdir" option.
 1727 
 1728 2017-08-29  Mathieu Lirzin  <mthl@gnu.org>
 1729 
 1730 	build: Use PERL5LIB from 'pre-inst-env' script
 1731 
 1732 	* bin/aclocal.in: Detect build environment with AUTOMAKE_UNINSTALLED.
 1733 	Let 'PERL5LIB' from 'pre-inst-env' define the perl module directories.
 1734 	* bin/automake.in: Likewise.
 1735 	* bin/wrap-aclocal.in: Don't set 'Automake::perl_libdirs' anymore.
 1736 	* bin/wrap-automake.in: Likewise.
 1737 	* t/Makefile.inc (LOG_COMPILER, PL_LOG_COMPILER): Use 'pre-inst-env'.
 1738 	(AM_PL_LOG_FLAGS): Don't set perl module path.
 1739 
 1740 2017-08-29  Mathieu Lirzin  <mthl@gnu.org>
 1741 
 1742 	build: Extend PATH from 'pre-inst-env' script
 1743 
 1744 	* configure.ac (ACLOCAL, AUTOMAKE): Use 'pre-inst-env'.
 1745 	* Makefile.am (extend_PATH): Delete.
 1746 	* doc/Makefile.inc (update_mans, $(srcdir)/%D%/amhello-1.0.tar.gz)
 1747 	(setup_autotools_paths): Use 'pre-inst-env' instead of $(extend_PATH).
 1748 
 1749 2017-08-29  Mathieu Lirzin  <mthl@gnu.org>
 1750 
 1751 	build: Add 'pre-inst-env' wrapper script
 1752 
 1753 	This script allows someone building Automake from source to easily run
 1754 	'automake' and 'aclocal' without having to install it or knowing the
 1755 	existence of "wrapped" scripts in "bin" directory.
 1756 
 1757 	* pre-inst-env.in: New script.
 1758 	* configure.ac: Generate 'pre-inst-env' at configure time.
 1759 	* Makefile.am (nodist_noinst_SCRIPTS, CLEANFILES): Add it.
 1760 
 1761 2017-08-29  Mathieu Lirzin  <mthl@gnu.org>
 1762 
 1763 	build: Move wrapped scripts in "bin" directory
 1764 
 1765 	Those scripts are used both in the build process and for the tests, so
 1766 	it seems clearer to not hide them in "t/wrap" directory.
 1767 
 1768 	* t/wrap/aclocal.in: Rename to ...
 1769 	* bin/wrap-aclocal.in: ... this.
 1770 	* t/wrap/automake.in: Rename to ...
 1771 	* bin/wrap-automake.in: ... this.
 1772 	* configure.ac: Generate "bin/aclocal-${APIVERSION}" and
 1773 	"bin/automake-${APIVERSION}" at configure time.
 1774 	* t/Makefile.inc (nodist_noinst_SCRIPTS): Move wrapped scripts ...
 1775 	* bin/Makefile.inc (nodist_noinst_SCRIPTS): ... here.
 1776 	* Makefile.am (extend_PATH): Use "bin" directory.
 1777 	* t/ax/test-defs.in (am_bindir): Likewise.
 1778 	* t/get-sysconf.sh: Likewise.
 1779 	* .gitignore: Update.
 1780 
 1781 2017-08-27  Mathieu Lirzin  <mthl@gnu.org>
 1782 
 1783 	automake: Handle LTLIBOBJS more specifically
 1784 
 1785 	* bin/automake.in (handle_LIBOBJS_or_ALLOCA): Remove complex regexp
 1786 	substitution when handling LTLIBOBJS.
 1787 
 1788 2017-08-21  Michael Haubenwallner  <michael.haubenwallner@ssi-schaefer.com>
 1789 
 1790 	automake: Ensure that LIBOBJ_DIR builddir is created
 1791 
 1792 	* t/libobj-no-dependency-tracking.sh: New test reproducing the issue
 1793 	described in automake bug#27781.
 1794 	* t/list-of-tests.mk (handwritten_TESTS, XFAIL_TESTS): Add it.
 1795 
 1796 2017-08-08  Mathieu Lirzin  <mthl@gnu.org>
 1797 
 1798 	doc: Update to latest help2man
 1799 
 1800 	This fixes bug#27773.
 1801 
 1802 	* doc/help2man: Sync with version 1.47.3 to support reproducible
 1803 	builds by using $SOURCE_DATE_EPOCH.
 1804 
 1805 2017-07-15  Mathieu Lirzin  <mthl@gnu.org>
 1806 
 1807 	doc: Fix typo "nothing" => "noting"
 1808 
 1809 	This fixes bug#24476.
 1810 
 1811 	* doc/automake.texi (Checking the Distribution): Fix typo.
 1812 	* THANKS: Update.
 1813 
 1814 2017-07-15  Mathieu Lirzin  <mthl@gnu.org>
 1815 
 1816 	doc: Fix typos
 1817 
 1818 	This fixes bug#23099.
 1819 
 1820 	* doc/automake.texi (Scripts-based Testsuites)
 1821 	(Parallel Test Harness): Fix typos.
 1822 	* THANKS: Update.
 1823 
 1824 2017-07-15  Mathieu Lirzin  <mthl@gnu.org>
 1825 
 1826 	check: Make 'check-TESTS' target depend on test dependencies
 1827 
 1828 	The 'check' and 'recheck' targets are already depending on the progams
 1829 	and scripts required to run the test suite.  The 'check-TESTS' target
 1830 	while not in the public API might be used by uninformed users, so it
 1831 	makes sense to add the same dependency to it.  This fixes bug#27186.
 1832 
 1833 	* lib/am/check.am [!%?SERIAL_TESTS%] (check-TESTS): Add dependency to
 1834 	'%CHECK_DEPS%'.
 1835 	* THANKS: Update.
 1836 
 1837 2017-07-15  Mathieu Lirzin  <mthl@gnu.org>
 1838 
 1839 	dejagnu: Allow AM_RUNTESTFLAGS to override default RUNTEST arguments
 1840 
 1841 	This fixes bug#25335.
 1842 
 1843 	* lib/am/dejagnu.am (check-DEJAGNU): Swap $(RUNTESTDEFAULTFLAGS) and
 1844 	$(AM_RUNTESTFLAGS).
 1845 	* THANKS: Update.
 1846 
 1847 2017-07-15  Thomas Jahns  <jahns@dkrz.de>
 1848 
 1849 	aclocal: Avoid problematic string interpolation
 1850 
 1851 	This fixes bug#20903.  Thanks to Eric Bavier for reporting this issue.
 1852 
 1853 	* bin/aclocal.in (automake_includes): Avoid problematic string
 1854 	interpolation.
 1855 	* NEWS: Update.
 1856 	* THANKS: Thank Eric Bavier.
 1857 
 1858 2017-07-15  Mathieu Lirzin  <mthl@gnu.org>
 1859 
 1860 	python: Add python3.6
 1861 
 1862 	This fixes bug#27458.
 1863 
 1864 	* m4/python.m4 (AM_PATH_PYTHON): Add python3.6.
 1865 
 1866 2017-07-06  Thomas Martitz  <kugel@rockbox.org>
 1867 
 1868 	automake: Shorter object file names under subdir-objects
 1869 
 1870 	Combining the 'subdir-objects' option with target-specific flags had
 1871 	the consequence of producing long object file names.  This was done to
 1872 	preventively ensure the uniqueness of object file names.  We are now
 1873 	using shorter names by default, and handle long names when an actual
 1874 	conflict is detected.  This will hopefully reduce the necessity of
 1875 	using the 'prog_SHORTNAME' facility.
 1876 
 1877 	Example:
 1878 	previously:
 1879 	  AUTOMAKE_OPTIONS = subdir-objects
 1880 	  bin_PROGRAMS += path/to/foo
 1881 	  path_to_foo_CFLAGS = $(AM_CFLAGS) -g
 1882 
 1883 	resulted in objects:
 1884 	  sub/path_to_foo-foo.o
 1885 
 1886 	now object file name is:
 1887 	  sub/foo-foo.o
 1888 
 1889 	* bin/automake.in (proglist, liblist, ltliblist)
 1890 	(dup_shortnames): New globals.
 1891 	(initialize_per_input): Initialize them.
 1892 	(handle_targets): New subroutine.
 1893 	(handle_single_transform): Truncate object file names when possible.
 1894 	* t/subobj-objname-clash.sh: New test.
 1895 	* t/list-of-tests.mk (handwritten_TESTS): Add it.
 1896 	* NEWS: Update.
 1897 
 1898 2017-06-18  Mathieu Lirzin  <mthl@gnu.org>
 1899 
 1900 	maint: Post-release administrivia
 1901 
 1902 	* NEWS: Add header line for next release.
 1903 	* configure.ac (AC_INIT): Bump version number to 1.15.1a.
 1904 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 1905 
 1906 2017-06-18  Mathieu Lirzin  <mthl@gnu.org>
 1907 
 1908 	version 1.15.1
 1909 
 1910 	* configure.ac (AC_INIT): Bump version number to 1.15.1.
 1911 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 1912 
 1913 2017-06-18  Mathieu Lirzin  <mthl@gnu.org>
 1914 
 1915 	maint: Update 'git-tag-release' rule
 1916 
 1917 	* maintainer/maint.mk (git-tag-release): Use a Git tag message similar
 1918 	to what the 'do-release-commit-and-tag' script from Gnulib do.
 1919 
 1920 2017-06-17  Mathieu Lirzin  <mthl@gnu.org>
 1921 
 1922 	maint: Update files from upstream with 'make fetch'
 1923 
 1924 	* lib/config.guess: Update.
 1925 	* lib/config.sub: Likewise.
 1926 	* lib/texinfo.tex: Likewise.
 1927 
 1928 2017-06-17  Mathieu Lirzin  <mthl@gnu.org>
 1929 
 1930 	maint: Update NEWS
 1931 
 1932 	* NEWS: Announce important bug fixes.
 1933 
 1934 2017-06-16  Mathieu Lirzin  <mthl@gnu.org>
 1935 
 1936 	Merge branch 'minor'
 1937 
 1938 2017-06-16  Mathieu Lirzin  <mthl@gnu.org>
 1939 
 1940 	Merge branch 'micro' into minor
 1941 
 1942 2017-05-20  Mathieu Lirzin  <mthl@gnu.org>
 1943 
 1944 	bootstrap: Add a rationale for the bootstrap process.
 1945 
 1946 	* bootstrap: Explain why we can't simply run 'autoreconf -i'.
 1947 
 1948 2017-05-20  Mathieu Lirzin  <mthl@gnu.org>
 1949 
 1950 	bootstrap: Rename 'bootstrap.sh' to 'bootstrap'.
 1951 
 1952 	Follow Gnulib's convention of using either "bootstrap" or "autogen.sh"
 1953 	file names for development bootstrap scripts.
 1954 
 1955 	* bootstrap.sh: Rename to ...
 1956 	* bootstrap: ... this.
 1957 	* GNUmakefile (bootstrap): Adapt.
 1958 	* HACKING: Likewise.
 1959 	* Makefile.am (EXTRA_DIST): Likewise.
 1960 	* doc/automake.texi (Future of aclocal)
 1961 	(Error required file ltmain.sh not found): Likewise.
 1962 	* maintainer/maint.mk (autodiffs, update-copyright): Likewise.
 1963 
 1964 2017-05-20  Mathieu Lirzin  <mthl@gnu.org>
 1965 
 1966 	automake: Update 'read_am_file' docstring.
 1967 
 1968 	* bin/automake.in (read_am_file): Update docstring which was referring
 1969 	to a non existent '%contents' variable.
 1970 
 1971 2017-04-16  Bruno Haible  <bruno@clisp.org>
 1972 
 1973 	tests: Never invoke gettextize.
 1974 
 1975 	This change fixes automake bug#26514.
 1976 
 1977 	* t/gettext-macros.sh: Never invoke gettextize.
 1978 
 1979 2017-04-13  Mathieu Lirzin  <mthl@gnu.org>
 1980 
 1981 	tests: Update "t/txinfo-no-clutter.sh".
 1982 
 1983 	This is workaround to a regression introduced by
 1984 	48107579abadbe857c8299c38d2ca0a3f8f354c8.  With current
 1985 	"lib/texinfo.tex", 'texi2dvi' is not able to produce a valid output
 1986 	for a Texinfo document containing a @synindex command.  As a
 1987 	consequence the test "t/txinfo-no-clutter.sh" was failing.
 1988 
 1989 	* t/txinfo-no-clutter.sh: Remove use of @synindex command.
 1990 
 1991 2017-03-05  Mathieu Lirzin  <mthl@gnu.org>
 1992 
 1993 	maint: Update "lib/texinfo.tex".
 1994 
 1995 	This fixes a regression in 48107579abadbe857c8299c38d2ca0a3f8f354c8
 1996 	where 'texi2dvi' was not able to produce a valid output for a Texinfo
 1997 	document without text.  As a consequence the test "t/silent-texi.sh"
 1998 	was failing.
 1999 
 2000 	* lib/texinfo.tex: Update to 2017-03-05.09.
 2001 
 2002 2017-03-02  Mathieu Lirzin  <mthl@gnu.org>
 2003 
 2004 	maint: Update copyright years to 2017.
 2005 
 2006 	This update has been made with 'make update-copyright'.
 2007 
 2008 2017-03-02  Mathieu Lirzin  <mthl@gnu.org>
 2009 
 2010 	maint: Update files from upstream with 'make fetch'.
 2011 
 2012 	* lib/INSTALL: Update.
 2013 	* lib/config.guess: Likewise.
 2014 	* lib/config.sub: Likewise.
 2015 	* lib/gendocs.sh: Likewise.
 2016 	* lib/gendocs_template: Likewise.
 2017 	* lib/gitlog-to-changelog: Likewise.
 2018 	* lib/gnupload: Likewise.
 2019 	* lib/texinfo.tex: Likewise.
 2020 	* lib/update-copyright: Likewise.
 2021 
 2022 2017-03-02  Mathieu Lirzin  <mthl@gnu.org>
 2023 
 2024 	maint: Update fetch URLs.
 2025 
 2026 	* maintainer/maint.mk (SV_CVS, SV_GIT_AC): Remove.
 2027 	(SV_GIT_CF, SV_GIT_GL): Use https for Savannah Git repositories.
 2028 	(FETCHFILES): Use Gnulib repository as the source of "texinfo.tex",
 2029 	"gendocs.sh", and "gendocs_template" files.
 2030 
 2031 2017-02-23  Mathieu Lirzin  <mthl@gnu.org>
 2032 
 2033 	doc: fix typo "requited" => "required"
 2034 
 2035 	This change fixes automake bug#25413.
 2036 
 2037 	* doc/automake.texi (Serial Test Harness): Fix typo.
 2038 
 2039 2017-01-01  Jim Meyering  <meyering@fb.com>
 2040 
 2041 	maint: update copyright dates for 2017
 2042 
 2043 	* all files: Run this command, using update-copyright from gnulib:
 2044 	  UPDATE_COPYRIGHT_FORCE=1 \
 2045 	  UPDATE_COPYRIGHT_USE_INTERVALS=2 \
 2046 	  UPDATE_COPYRIGHT_MAX_LINE_LENGTH=79 \
 2047 	   update-copyright $(git ls-files)
 2048 
 2049 2016-09-08  Jim Meyering  <meyering@fb.com>
 2050 
 2051 	automake: do not emit rule with two dependents
 2052 
 2053 	On a system using our replacement alloca, make would emit
 2054 	this warning:
 2055 	  target '.deps/alloca.Po' given more than once in the same rule
 2056 	That arose because automake would emit a rule depending on both
 2057 	./.deps/alloca.Po and .deps/alloca.Po. Normally, duplicate
 2058 	dependents are avoided by virtue of their names being keys in
 2059 	the %dep_files hash, but in this case, that particular file
 2060 	was specified in two different ways.
 2061 	* bin/automake.in (handle_ALLOCA): When the $dir prefix is empty,
 2062 	make it './', to ensure that when we add ./.deps/alloca.Po
 2063 	it is deduped.
 2064 	See: http://bugs/gnu/org/22702
 2065 
 2066 2016-06-01  Jim Meyering  <meyering@fb.com>
 2067 
 2068 	compile: remove .Tpo file upon failure
 2069 
 2070 	When generating a .deps/base.Po file, our emitted rule first writes
 2071 	to a temporary .Tpo file, so the final creation can be atomic, via
 2072 	mv's rename.  However, when generation of the .Tpo fails, it stop the
 2073 	process and does not remove that temporary file.  And nothing else
 2074 	ever deletes it, either.  Hence, in the unusual case in which one
 2075 	expects a compilation to fail during a successful build (gnulib's
 2076 	test-verify.sh does precisely this), a .deps/test-verify.Tpo file
 2077 	would be left behind, and that would in turn cause a "make distcheck"
 2078 	failure because that file would exist in one build+distclean hierarchy
 2079 	but not in the distribution tarball.
 2080 	* lib/am/depend2.am: Ensure that the temporary file is removed
 2081 	upon failure.
 2082 	* t/distcheck-Tpo.sh: New file.
 2083 	* t/list-of-tests.mk (handwritten_TESTS): Add it.
 2084 	This addresses bug#23661.
 2085 
 2086 2016-04-20  Paul Eggert  <eggert@cs.ucla.edu>
 2087 
 2088 	python: add python3.5, python3.4
 2089 
 2090 	* m4/python.m4 (AM_PATH_PYTHON): Add python3.5, python3.4.
 2091 
 2092 2016-03-31  Paul Eggert  <eggert@cs.ucla.edu>
 2093 
 2094 	Merge branch 'minor'
 2095 
 2096 2016-03-31  Paul Eggert  <eggert@cs.ucla.edu>
 2097 
 2098 	Merge branch 'micro' into minor
 2099 
 2100 2016-03-31  Paul Eggert  <eggert@cs.ucla.edu>
 2101 
 2102 	automake: port to Perl 5.22 and later
 2103 
 2104 	Without this change, Perl 5.22 complains "Unescaped left brace in
 2105 	regex is deprecated" and this is planned to become a hard error in
 2106 	Perl 5.26.  See:
 2107 	http://search.cpan.org/dist/perl-5.22.0/pod/perldelta.pod#A_literal_%22{%22_should_now_be_escaped_in_a_pattern
 2108 	* bin/automake.in (substitute_ac_subst_variables): Escape left brace.
 2109 
 2110 2016-03-28  Paul Eggert  <eggert at>
 2111 
 2112 	automake: port better to future gzip
 2113 
 2114 	* lib/am/distdir.am (dist-gzip, dist-shar, distcheck):
 2115 	Port better to future versions of gzip, which are planned to
 2116 	deprecate the GZIP environment variable (Bug#20132).
 2117 
 2118 2016-01-12  Paul Eggert  <eggert@cs.ucla.edu>
 2119 
 2120 	Merge branch 'minor'
 2121 
 2122 2016-01-12  Paul Eggert  <eggert@cs.ucla.edu>
 2123 
 2124 	Merge branch 'micro' into minor
 2125 
 2126 2016-01-12  Paul Eggert  <eggert@cs.ucla.edu>
 2127 
 2128 	maint: port time-stamp-time-zone to strict POSIX
 2129 
 2130 	Set time-stamp-time-zone to "UTC0", not to "UTC", as POSIX defines
 2131 	TZ="UTC0" not TZ="UTC".
 2132 
 2133 2015-11-24  Peter Rosin  <peda@lysator.liu.se>
 2134 
 2135 	Merge branch 'minor'
 2136 
 2137 	* minor:
 2138 	  compile: add icl to compile wrapper script
 2139 	  tests: fix a typo-induced bug
 2140 
 2141 2015-11-24  Peter Rosin  <peda@lysator.liu.se>
 2142 
 2143 	Merge branch 'micro' into minor
 2144 
 2145 	* micro:
 2146 	  compile: add icl to compile wrapper script
 2147 
 2148 2015-11-24  Jonathan L Peyton  <jonathan.l.peyton@intel.com>  (tiny change)
 2149 
 2150 	compile: add icl to compile wrapper script
 2151 
 2152 	* lib/compile: Have icl be treated similarly to cl
 2153 	(scriptversion): Update.
 2154 	* t/ax/am-test-lib.sh (require_tool): Handle icl.
 2155 	* t/compile7.sh: Add new test file for icl...
 2156 	* t/list-of-tests.mk (handwritten_TESTS): ...and use it.
 2157 	* NEWS: Update.
 2158 	* THANKS: Update.
 2159 
 2160 2015-01-06  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2161 
 2162 	tests: fix a typo-induced bug
 2163 
 2164 	* t/subobj-vpath-pr13928.sh: Here; we were using $FGREP instead of
 2165 	$EGREP, oops.
 2166 
 2167 2015-01-06  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2168 
 2169 	Merge branch 'minor'
 2170 
 2171 	* minor:
 2172 	  deps: fix corner-case "make distclean" bug
 2173 	  compile: don't place built object files in $(srcdir), ever ...
 2174 	  tests: fix some bugs in an XFAILing test
 2175 	  deps: 'subdir-object' option now works when foo_SOURCES contains $(var)
 2176 	  NEWS: fix a typo
 2177 
 2178 2015-01-06  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2179 
 2180 	Merge branch 'deps-pr13928' into minor
 2181 
 2182 	* deps-pr13928:
 2183 	  deps: fix corner-case "make distclean" bug
 2184 	  compile: don't place built object files in $(srcdir), ever ...
 2185 	  tests: fix some bugs in an XFAILing test
 2186 	  deps: 'subdir-object' option now works when foo_SOURCES contains $(var)
 2187 
 2188 2015-01-06  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2189 
 2190 	Merge branch 'micro' into minor
 2191 
 2192 	* micro:
 2193 	  NEWS: fix a typo
 2194 
 2195 2015-01-06  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2196 
 2197 	deps: fix corner-case "make distclean" bug
 2198 
 2199 	Assume we have package satisfying the following conditions:
 2200 	  (1) automatic dependency tracking is enabled;
 2201 	  (2) the 'subdir-objects' Automake option is enabled;
 2202 	  (3) the package uses a recursive make setup.
 2203 
 2204 	Also assume that:
 2205 	  (a) a subdir Makefile declares a foo_SOURCES variable containing
 2206 	      a source file in the parent directory;
 2207 	  (b) that parent Makefile declare a compiled program itself.
 2208 
 2209 	Then BSD and Solaris make used to fail when running "make distclean",
 2210 	because the 'distclean' target of the subdir Makefile removed the
 2211 	whole '.deps' directory before the parent Makefile was done with the
 2212 	included '.Po' makefile fragments in that directory. This issue was
 2213 	revealed by failures in the 'subobj-vpath-pr13928.sh' test when those
 2214 	make implementations were used.
 2215 
 2216 	We fix the issue by ensuring the 'distclean' target of any Makefile
 2217 	only removed the '.Po' makefile fragments included by it, rather than
 2218 	the whole '.deps' directory where such files resides.
 2219 
 2220 	This change should be the last step in fixing automake bug#13928
 2221 	for good.
 2222 
 2223 	* bin/automake.in (handle_languages), lib/am/depend.am: Adjust
 2224 	to implement the new 'distclean' logic.
 2225 	* t/pr224.sh: Adjust to avoid a spurious failure.
 2226 	* PLANS/subdir-objects.txt: Update.
 2227 
 2228 2015-01-06  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2229 
 2230 	compile: don't place built object files in $(srcdir), ever ...
 2231 
 2232 	... even when a source file is specified as '$(srdir)/foo.c' or
 2233 	'$(top_srcdir)/bar.c'. And ditto for dependency-tracking makefile
 2234 	fragments (those under '.deps' directories).
 2235 
 2236 	Such issues used to occur when the 'subdir-objects' option was given.
 2237 
 2238 	This change should fix the second and last part of automake bug#13928.
 2239 	See also bug#16375 and bug#15293.
 2240 
 2241 	* NEWS: Update.
 2242 	* bin/automake.in (handle_single_transform): Make sure object files
 2243 	and dependency-tracking makefile fragments coming from source like
 2244 	'$(srcdir)/foo.c' and '$(top_srcdir)/bar.c' are placed respectively
 2245 	under $(builddir) and $(top_builddir).
 2246 	* t/subobj-vpath-pr13928.sh: Enhance to expose even more aspects
 2247 	of the bug we've just fixed.
 2248 	* t/subobj-pr13928-more-langs.sh: New test, similar to the one above,
 2249 	but with non-C languages as well.
 2250 	* t/list-of-tests.mk (XFAIL_TESTS): Remove 'subobj-vpath-pr13928.sh',
 2251 	it's now supposed to pass.
 2252 	(handwritten_TESTS): Add 'subobj-pr13928-more-langs.sh'.
 2253 
 2254 2015-01-06  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2255 
 2256 	tests: fix some bugs in an XFAILing test
 2257 
 2258 	* t/subobj-vpath-pr13928.sh: This one. The test would have failed (or
 2259 	hung!) even if the bug it was testing against were fixed.
 2260 
 2261 2015-01-06  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2262 
 2263 	deps: 'subdir-object' option now works when foo_SOURCES contains $(var)
 2264 
 2265 	Following a suggestions of Johan Kristensen, we have config.status use
 2266 	'make' invocations rather than Makefile-parsing 'sed' hacks to bootstrap
 2267 	the dependency-tracking '.Po' and '.Plo' makefile fragments. To handle
 2268 	the inclusion of such files that are still missing when make is first
 2269 	we basically generate a temporary Makefile without these includes, and
 2270 	call 'make' on that Makefile.
 2271 
 2272 	This fixes the serious bug bug#13928, which was an hard blocker to make the
 2273 	behavior mandated by the 'subdir-object' active by default (which we want
 2274 	to do in Automake 2.0).
 2275 
 2276 	The issue has also been reported in bug#15919.
 2277 
 2278 	* NEWS, THANKS: Update.
 2279 	* bin/automake.in (handle_languages): Add a trailing "marking" comment
 2280 	("# am--include-marker") to the generated Makefile lines issuing 'include'
 2281 	directives for the dependency-tracking '.Po' and '.Plo' makefile fragments.
 2282 	Also rename the generated Makefile variable 'am__depfiles_maybe' to the
 2283 	clearer 'am__maybe_remake_depfiles'.
 2284 	Minor unrelated refactoring.
 2285 	* lib/am/configure.am: Adjust to account for the 'am__depfiles_maybe' ->
 2286 	'am__maybe_remake_depfiles' renaming.
 2287 	* lib/am/depend.am: Add rules to generate a dummy version of all the
 2288 	dependency-tracking '.Po' and '.Plo' makefile fragments.
 2289 	* m4/depout.m4: Use make invocations rather than Makefile-parsing sed hacks
 2290 	to bootstrap the dependency-tracking '.Po' and '.Plo' makefile fragments.
 2291 	We still use some sed trickery in order to remove the inclusion of the
 2292 	still non existing .Po and .Plo files from the Makefile we invoke make
 2293 	upon; this is done stripping lines that contain the magic string
 2294 	"# am--include-marker".
 2295 	* m4/make.m4 (AM_MAKE_INCLUDE): Given that now automake generates Makefiles
 2296 	containing include statements with trailing comment, adjust the checks done
 2297 	here to make sure $MAKE support that; e.g., "include foo.mk # comment"
 2298 	rather than just "include foo.mk".
 2299 	Also refactor and adjust to leave better debugging info in config.log.
 2300 	* t/postproc.sh: Rename ...
 2301 	* t/depend-postproc.sh: ... to this, and adjust and enhance.
 2302 	* t/list-of-tests.mk (handwritten_TESTS): Adjust.
 2303 	(XFAIL_TESTS): Remove 't/subobj-indir-pr13928.sh', which is now succeeding.
 2304 	* t/subobj-indir-pr13928.sh: Simplify slightly, now that we expect it to
 2305 	pass.
 2306 	* t/depcomp8a.sh: Adjust grepping check to account for the changes in
 2307 	the generated Makefile, and tp be somewhat more robust in light of possible
 2308 	future modifications.
 2309 	* t/depcomp8b.sh: Likewise.
 2310 	* t/subobj11b.sh: Likewise.
 2311 	* t/subobj11c.sh: Likewise.
 2312 	* t/extra-sources.sh: Likewise.
 2313 	* t/lex-depend-grep.sh: Likewise.
 2314 	* t/lex-depend-cxx.sh: Add a command to help debugging in case of test
 2315 	failure.
 2316 
 2317 	Helped-by: Johan Kristensen <johankristensen@gmail.com>
 2318 
 2319 2015-01-06  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2320 
 2321 	NEWS: fix a typo
 2322 
 2323 2015-01-05  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2324 
 2325 	maint: update copyright years to 2015 (branch 'master')
 2326 
 2327 	* t/am-prog-mkdir-p.sh: Here.
 2328 	* t/txinfo-no-split.sh: And here.
 2329 
 2330 2015-01-05  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2331 
 2332 	Merge branch 'minor'
 2333 
 2334 	* minor:
 2335 	  maint: update copyright years to 2015 (branch 'micro')
 2336 
 2337 2015-01-05  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2338 
 2339 	Merge branch 'micro' into minor
 2340 
 2341 	* micro:
 2342 	  maint: update copyright years to 2015 (branch 'micro')
 2343 
 2344 2015-01-05  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2345 
 2346 	maint: update copyright years to 2015 (branch 'micro')
 2347 
 2348 2015-01-05  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2349 
 2350 	Merge branch 'minor'
 2351 
 2352 	* minor:
 2353 	  sync: update third-part files from upstream
 2354 	  maint: fix typo in error message of a maintainer-only rule
 2355 	  maint: bump version number 1.15 -> 1.15a ('minor' Git branch)
 2356 	  release: stable minor release 1.15
 2357 	  NEWS: minor improvements and fixed some typos and grammaros
 2358 
 2359 2015-01-05  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2360 
 2361 	Merge branch 'micro' into minor
 2362 
 2363 	* micro:
 2364 	  sync: update third-part files from upstream
 2365 	  maint: fix typo in error message of a maintainer-only rule
 2366 
 2367 2015-01-05  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2368 
 2369 	sync: update third-part files from upstream
 2370 
 2371 	* lib/config.sub: This.
 2372 	* lib/INSTALL, lib/config.guess, lib/gitlog-to-changelog, lib/gnupload,
 2373 	lib/update-copyright: And this (but only for copyright year update, no
 2374 	real semantic change)
 2375 
 2376 2015-01-05  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2377 
 2378 	maint: fix typo in error message of a maintainer-only rule
 2379 
 2380 	* maintainer/maint.mk (web-manual-update): Here.
 2381 
 2382 2015-01-05  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2383 
 2384 	maint: bump version number 1.15 -> 1.15a ('minor' Git branch)
 2385 
 2386 	* configure.ac (AC_INIT): Bump version number to 1.15a.
 2387 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 2388 
 2389 2015-01-05  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2390 
 2391 	Merge branch 'minor' into micro after 1.15 release
 2392 
 2393 	And bump version number: 1.15 -> 1.15.0a
 2394 
 2395 	* minor:
 2396 	  release: stable minor release 1.15
 2397 	  NEWS: minor improvements and fixed some typos and grammaros
 2398 	  docs: "make distcheck" implementation details are not to be abused
 2399 	  NEWS: improve and adjust in light of the oncoming 1.15 release
 2400 	  Fix dumb logic error preventing $install_sh from being be overridden
 2401 	  Expose automake bug#19311
 2402 	  build: fix race in parallel builds
 2403 	  init: ensure $ac_aux_dir is defined before being used
 2404 	  plans: enabling subdir-object by default is blocked on bug#13928
 2405 	  maint: update copyright years
 2406 	  maint: sync files from upstream ("make fetch")
 2407 	  Typofixes in warning messages and manual
 2408 	  NEWS: a typofix, and better word wrapping
 2409 	  parallel-tests: avoid possible implicit "make all" in test-suite.log rule
 2410 	  Allow user to extend .PRECIOUS target
 2411 	  cosmetics: remove a couple of extra trailing white spaces
 2412 	  tests: fix a spurious failure on Mac OS X
 2413 	  docs: make clear the JAVA primary is frozen
 2414 	  install-sh: a slightly better diagnostic, and tests enhancements
 2415 	  install-sh: be stricter in catching invalid usages
 2416 	  tests: more significant names for some tests
 2417 	  tests: some cosmetic fixes
 2418 	  tests: more significant names for a test
 2419 	  docs: drop a few obsolescent FIXME/TODO comments, and associated text
 2420 	  testsuite harness: report test exit status in log file
 2421 	  TAP driver: no need to invoke AC_PROG_AWK directly
 2422 	  TAP driver: remove perl implementation (move it into contrib/)
 2423 	  NEWS: stop reporting "new" Automake versioning scheme
 2424 	  cosmetics: untabify the install-sh script
 2425 	  install-sh: assume that "set -f" and "set +f" work...
 2426 	  install-sh: assume ${var:-value} works as expected
 2427 	  install-sh: assume 'dirname' is available and working correctly
 2428 	  post-release: micro version bump (1.14a)
 2429 
 2430 2014-12-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2431 
 2432 	release: stable minor release 1.15
 2433 
 2434 	* configure.ac (AC_INIT): Bump version number to 1.15.
 2435 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 2436 
 2437 2014-12-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2438 
 2439 	NEWS: minor improvements and fixed some typos and grammaros
 2440 
 2441 2014-12-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2442 
 2443 	Merge branch 'minor'
 2444 
 2445 	* minor:
 2446 	  docs: "make distcheck" implementation details are not to be abused
 2447 	  NEWS: improve and adjust in light of the oncoming 1.15 release
 2448 	  shell-no-trail-bslash: improve diagnostic in case of failure
 2449 
 2450 2014-12-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2451 
 2452 	docs: "make distcheck" implementation details are not to be abused
 2453 
 2454 	* doc/automake.texi: State explicitly and in detail that the exact location
 2455 	and the exact structure of the subdirectory used by "make distcheck" is to
 2456 	be considered an implementation detail, which can change at any time.
 2457 
 2458 2014-12-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2459 
 2460 	NEWS: improve and adjust in light of the oncoming 1.15 release
 2461 
 2462 2014-12-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2463 
 2464 	Merge branch 'micro' into minor
 2465 
 2466 	* micro:
 2467 	  shell-no-trail-bslash: improve diagnostic in case of failure
 2468 
 2469 2014-12-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2470 
 2471 	shell-no-trail-bslash: improve diagnostic in case of failure
 2472 
 2473 	* t/ax/shell-no-trail-bslash.in: Here, by fixing a typo in a
 2474 	variable name and a logic error.
 2475 	* t/self-check-shell-no-trail-bslash.sh: Enhance to catch the
 2476 	issue.
 2477 
 2478 2014-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2479 
 2480 	Merge branches 'texi-in-builddir-remove-distclean-hack' and 'minor'
 2481 
 2482 	* texi-in-builddir-remove-distclean-hack:
 2483 	  texinfo: remove hack about info files in CLEANFILES variables
 2484 
 2485 	* minor:
 2486 	  tests: make script 'shell-no-trail-bslash' simpler and more robust
 2487 
 2488 2014-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2489 
 2490 	Merge branch 'micro' into minor
 2491 
 2492 	* micro:
 2493 	  tests: make script 'shell-no-trail-bslash' simpler and more robust
 2494 
 2495 2014-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2496 
 2497 	tests: make script 'shell-no-trail-bslash' simpler and more robust
 2498 
 2499 	This solves spurious failure in the 'check-no-trailing-backslash-in-recipes'
 2500 	target for Automake-NG.
 2501 
 2502 	This is basically a backport of Automake-NG commit v1.14.1-1010-g85aae58;
 2503 	the point is to minimize the amount of spurious diffs between the mainline
 2504 	Automake and the Automake-NG source trees.
 2505 
 2506 	* t/ax/shell-no-trail-bslash.in: Simplify and fortify.
 2507 	* t/self-check-shell-no-trail-bslash.sh: Enhance.
 2508 
 2509 2014-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2510 
 2511 	texinfo: remove hack about info files in CLEANFILES variables
 2512 
 2513 	Automake used to implement an undocumented hack causing '.info' files
 2514 	that appeared to be cleaned (by e.g. being listed in the CLEANFILES
 2515 	variable) to also be built in the builddir rather than in the srcdir;
 2516 	this was for backward compatibility with packages such as Texinfo,
 2517 	which did things like:
 2518 
 2519 	    info_TEXINFOS = texinfo.txi info-stnd.texi info.texi
 2520 	    DISTCLEANFILES = texinfo texinfo-* info*.info*
 2521 	    # Do not create info files for distribution.
 2522 	    dist-info:
 2523 
 2524 	in order not to distribute .info files.
 2525 
 2526 	Now that we have the 'info-in-builddir' option that explicitly causes
 2527 	generated '.info' files to be placed in the builddir, this hack is no
 2528 	longer necessary, and we can remove it (after having deprecated it in
 2529 	the Automake 1.14 release already).
 2530 
 2531 	* bin/automake.in (handle_texinfo_helper): Remove the hack.  Adjust
 2532 	comments accordingly.
 2533 	* NEWS: Update.
 2534 	* t/txinfo23.sh: Delete as obsolete.
 2535 	* t/txinfo25.sh: Likewise.
 2536 	* t/txinfo24.sh: Likewise.
 2537 	* t/txinfo28.sh: Delete as mostly obsolete, its only still relevant
 2538 	parts moved ...
 2539 	* t/mdate5.sh: ... into this test.
 2540 	* t/txinfo-clean.sh: Remove references to deleted tests.
 2541 	* t/list-of-tests.mk: Adjust.
 2542 
 2543 2014-12-27  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2544 
 2545 	tests: avoid a possible failure due to autom4te cache staleness
 2546 
 2547 	* t/includes-deprecation.sh: Here.
 2548 
 2549 2014-12-27  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2550 
 2551 	Merge branch 'minor'
 2552 
 2553 	* minor:
 2554 	  tests: fix spurious failure in test on TEXINFO_TEX overriding
 2555 	  tests: avoid some spurious failures on AIX 7.1
 2556 
 2557 2014-12-27  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2558 
 2559 	Merge branch 'micro' into minor
 2560 
 2561 	* micro:
 2562 	  tests: fix spurious failure in test on TEXINFO_TEX overriding
 2563 	  tests: avoid some spurious failures on AIX 7.1
 2564 
 2565 2014-12-27  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2566 
 2567 	tests: fix spurious failure in test on TEXINFO_TEX overriding
 2568 
 2569 	* t/txinfo-override-texinfo-tex.sh: Here.  The issue was pre-existing, but
 2570 	has been only recently exposed by the fix for automake bug#18286 "distcheck
 2571 	fails to detect missing files" (see commit v1.14.1-4-g01a7a4a) and by the
 2572 	BSD make semantics.  To convince yourself this change actually makes sense
 2573 	semantically, see https://sourceware.org/ml/binutils/2012-06/msg00004.html
 2574 
 2575 2014-12-27  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2576 
 2577 	Merge branch 'aix-testsuite-failures' into micro
 2578 
 2579 	* aix-testsuite-failures:
 2580 	  tests: avoid some spurious failures on AIX 7.1
 2581 
 2582 2014-12-27  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2583 
 2584 	tests: avoid some spurious failures on AIX 7.1
 2585 
 2586 	* t/lex-noyywrap.sh: Here.
 2587 	* t/instmany-mans.sh: And here.
 2588 	* t/instmany-python.sh: And here.
 2589 	* t/instmany.sh: And here.
 2590 	* t/parallel-tests-concurrency.sh: And here.
 2591 
 2592 2014-12-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2593 
 2594 	Merge branch 'minor'
 2595 
 2596 	* minor:
 2597 	  dist: fix bug#18286 "distcheck fails to detect missing files"
 2598 	  tests: expose automake bug#18286 "distcheck fails to detect missing files"
 2599 	  include: fix bug in handling of user-defined makefile fragments generation
 2600 	  tests: expose bug in handling of user-defined makefile fragments generation
 2601 
 2602 2014-12-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2603 
 2604 	Merge branch 'micro' into minor
 2605 
 2606 	* micro:
 2607 	  dist: fix bug#18286 "distcheck fails to detect missing files"
 2608 	  tests: expose automake bug#18286 "distcheck fails to detect missing files"
 2609 	  include: fix bug in handling of user-defined makefile fragments generation
 2610 	  tests: expose bug in handling of user-defined makefile fragments generation
 2611 
 2612 2014-12-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2613 
 2614 	Merge branch 'distcheck-pr18286' into micro
 2615 
 2616 	* distcheck-pr18286:
 2617 	  dist: fix bug#18286 "distcheck fails to detect missing files"
 2618 	  tests: expose automake bug#18286 "distcheck fails to detect missing files"
 2619 
 2620 2014-12-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2621 
 2622 	dist: fix bug#18286 "distcheck fails to detect missing files"
 2623 
 2624 	BTW, this issue had been already reported in the past:
 2625 	http://lists.gnu.org/archive/html/automake/2006-09/msg00008.html
 2626 	http://lists.gnu.org/archive/html/automake/2013-01/msg00049.html
 2627 
 2628 	"make distcheck" could sometimes fail to detect missing files in the
 2629 	distribution tarball, especially in those cases where both the generated
 2630 	files and their dependencies are explicitly in $(srcdir).  An important
 2631 	example of this are *generated* makefile fragments included at Automake
 2632 	time in Makefile.am.  A basic example:
 2633 
 2634 	    # -*- Makefile.am -*-
 2635 
 2636 	    $(srcdir)/fragment.am: $(srcdir)/data.txt $(srcdir)/preproc.sh
 2637 	        cd $(srcdir) && $(SHELL) preproc.sh <data.txt >fragment.am
 2638 
 2639 	    include $(srcdir)/fragment.am
 2640 
 2641 	    ...
 2642 
 2643 	If the use forgot to add data.txt and/or preproc.sh in the distribution
 2644 	tarball, "make distcheck" would have erroneously succeeded!
 2645 
 2646 	The reason is that, while $(srcdir)/data.txt does not exist, make also
 2647 	looks in $(srcdir)/$(srcdir)/data.txt, and in the distcheck-issued
 2648 	VPATH build where $(srcdir) is '..', that file exists, as it is
 2649 	part of the original development directory.
 2650 
 2651 	* t/distdir.am (distcheck): Adjust to have the build directory be
 2652 	'$(distdir)/_build/sub' rather than just '$(distdir)/_build'.  Thanks
 2653 	Nicola Fontana for the suggestion.
 2654 	* t/distcheck-pr18286.sh: Enhance and tighten a little.
 2655 	* t/list-of-tests.mk (XFAIL_TESTS): Remove 't/distcheck-pr18286.sh',
 2656 	as it's now passing.
 2657 	* t/subdir-am-cond.sh: Adjust to avoid a fully spurious failure due
 2658 	to the new distcheck semantics.
 2659 	* t/subdir-ac-subst.sh: Likewise.
 2660 	* t/dejagnu-relative-srcdir.sh: Likewise.
 2661 	* t/txinfo-builddir.sh: Likewise.
 2662 	* NEWS: Update.
 2663 
 2664 	Helped-by: Nicola Fontana <ntd@entidi.it>
 2665 	Helped-by: Peter Johansson <trojkan@gmail.com>
 2666 
 2667 2014-12-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2668 
 2669 	tests: expose automake bug#18286 "distcheck fails to detect missing files"
 2670 
 2671 	* t/distcheck-pr18286.sh: New test, still XFAILing.
 2672 	* t/list-of-tests.mk: Add it.
 2673 
 2674 2014-12-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2675 
 2676 	Merge branch 'am-deps' into micro
 2677 
 2678 	* am-deps:
 2679 	  include: fix bug in handling of user-defined makefile fragments generation
 2680 	  tests: expose bug in handling of user-defined makefile fragments generation
 2681 
 2682 2014-12-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2683 
 2684 	include: fix bug in handling of user-defined makefile fragments generation
 2685 
 2686 	If a user defined one single Makefile fragment to be included (via Automake
 2687 	includes) in his main Makefile.am, and gave a rule to generate that file
 2688 	from other data, Automake used to spuriously complain about with something
 2689 	like "overrides Automake target '$(srcdir)/foo.am".  This change remove that
 2690 	spurious error (via a simple hack rather than a systematic change, but oh
 2691 	well).
 2692 
 2693 	* lib/am/configure.am (%MAKEFILE-IN-DEPS%) [?HAVE-MAKEFILE-IN-DEPS?]: Add
 2694 	a trailing "$(am__empty)" to the list of targets, which is enough to trick
 2695 	Automake into not complaining about "duplicated targets" in case the
 2696 	'%MAKEFILE-IN-DEPS%' list expands to a single target that is also declared
 2697 	in some user-defined rule.
 2698 	* t/list-of-tests.mk (XFAIL_TESTS): Remove now-passing test
 2699 	't/am-include-only-one-generated-fragment.sh'.
 2700 	* NEWS: Update.
 2701 
 2702 2014-12-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2703 
 2704 	tests: expose bug in handling of user-defined makefile fragments generation
 2705 
 2706 	If a user defines one single Makefile fragment to be included (via Automake
 2707 	includes) in his main Makefile.am, and givse a rule to generate that file
 2708 	from other data, Automake will spuriously complain about with something
 2709 	like "overrides Automake target '$(srcdir)/foo.am".
 2710 
 2711 	* t/am-include-only-one-generated-fragment.sh: Expose the bug (this test
 2712 	is still XFAILing).
 2713 	* t/list-of-tests.mk: Add the new test.
 2714 
 2715 2014-12-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2716 
 2717 	Merge branch 'minor'
 2718 
 2719 	* minor:
 2720 	  cleanup: refactor code to initialize DIST_COMMON
 2721 	  dist: ordering of files in DIST_COMMON is deterministic now
 2722 	  tests: refactor some tests on DIST_COMMON
 2723 	  maint: make output of 'gen-testsuite-part' deterministic
 2724 	  When computing lispdir, don't load emacs site wide init file.
 2725 	  PATH: quote $(PATH_SEPARATOR) as well
 2726 	  Improve detection of GNU make, avoiding "Arg list too long" errors.
 2727 
 2728 2014-12-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2729 
 2730 	Merge branch 'micro' into minor
 2731 
 2732 	* micro:
 2733 	  cleanup: refactor code to initialize DIST_COMMON
 2734 	  dist: ordering of files in DIST_COMMON is deterministic now
 2735 	  tests: refactor some tests on DIST_COMMON
 2736 	  maint: make output of 'gen-testsuite-part' deterministic
 2737 	  When computing lispdir, don't load emacs site wide init file.
 2738 	  PATH: quote $(PATH_SEPARATOR) as well
 2739 	  Improve detection of GNU make, avoiding "Arg list too long" errors.
 2740 
 2741 2014-12-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2742 
 2743 	cleanup: refactor code to initialize DIST_COMMON
 2744 
 2745 	There is not need to make that an Automake variable early,
 2746 	only to later get and munge its contents, and use the new
 2747 	content to redefine the variable.
 2748 
 2749 	* bin/automake.in (@dist_common): New global variable.
 2750 	(push_dist_common, handle_dist): Use it.
 2751 	(handle_dist): Define am__DIST_COMMON instead of DIST_COMMON
 2752 	directly.
 2753 	(initialize_per_input): Reset it to empty.
 2754 	($configure_dist_common): Turn this scalar variable ...
 2755 	(@configure_dist_common): ... into this array variable.
 2756 	(handle_dist): Adjust.
 2757 	(required_file_check_or_copy): Update and wrap some comments.
 2758 	* lib/am/distdir.am (DIST_COMMON): Append $(am__DIST_COMMON).
 2759 	* t/distcom2.sh: Tighten a little.
 2760 
 2761 2014-12-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2762 
 2763 	dist: ordering of files in DIST_COMMON is deterministic now
 2764 
 2765 	It had likely stopped being deterministic due to the new perl behavior
 2766 	of having non-deterministic order of numerating hash keys:
 2767 	<http://search.cpan.org/dist/perl-5.18.0/pod/perldelta.pod#Hash_randomization>
 2768 	<http://onionstand.blogspot.ie/2012/12/are-you-relying-on-hash-keys-being.html>
 2769 	See also similar commit v1.14-19-g52e6404, albeit in this case the issue
 2770 	is likely coming from autom4te/autoconf, not from automake itself.
 2771 
 2772 	Fixes automake bug http://debbugs.gnu.org/17908
 2773 
 2774 	* bin/automake.in (handle_dist): Sort @dist_common.
 2775 	(print_autodist_files): Swap invocations of 'sort' and 'uniq', for
 2776 	consistency with the new code in 'handle_dist' and to get rid of a
 2777 	minor hack.
 2778 	* NEWS: Update.
 2779 
 2780 2014-12-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2781 
 2782 	tests: refactor some tests on DIST_COMMON
 2783 
 2784 	So that they prefer checking the semantics of the generated Makefiles,
 2785 	rather than grepping their content.  This will be useful in an upcoming
 2786 	refactoring.
 2787 
 2788 	* t/distcom-subdir.sh: Adjust this test.
 2789 	* t/distcom2.sh: And this.
 2790 	* t/distcom3.sh: And this.
 2791 	* t/distcom4.sh: And this.
 2792 	* t/distcom5.sh: And this.
 2793 
 2794 2014-12-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2795 
 2796 	maint: make output of 'gen-testsuite-part' deterministic
 2797 
 2798 	So that diffs displayed by the 'compare-autodiffs' target are
 2799 	less spurious and more useful.
 2800 
 2801 	* gen-testsuite-part: Sort keys of %deps_extractor, %wrapper_setups
 2802 	and %depmodes before iterating on them.
 2803 
 2804 2014-12-19  Andrew Burgess  <andrew.burgess@embecosm.com>  (tiny change)
 2805 
 2806 	When computing lispdir, don't load emacs site wide init file.
 2807 
 2808 	When computing the lispdir emacs was previously invoked with the '-q'
 2809 	option to avoid loading the users initialisation files, however, the
 2810 	site wide initialisation file was still loaded, in some cases this can
 2811 	cause emacs to hang, with the result that a configure can also hang.
 2812 
 2813 	The lisp code that aclocal causes to be executed reduces the load-path
 2814 	list (in emacs) to empty.  The load-path is used by emacs to find
 2815 	packages which it wants to load.  Currently, if emacs tries to auto
 2816 	load a package during shut down, and the package is not found, then
 2817 	emacs will hang.  This does seem like an emacs bug, but protecting
 2818 	against this in aclocal is simply a case of not loading the site wide
 2819 	initialisation file.
 2820 
 2821 	In this patch then the '-q' option to emacs is replaced with '-Q',
 2822 	this has the same, the '-Q' option is similar to '-q --no-site-file
 2823 	--no-splash'.
 2824 
 2825 	* doc/automake.texi (Hard-Coded Install Paths): Update explanation of
 2826 	emacs code used to get lispdir.
 2827 	* m4/lispdir.m4 (AM_PATH_LISPDIR): Update emacs flags.
 2828 
 2829 2014-12-19  KO Myung-Hun  <komh78@gmail.com>  (tiny change)
 2830 
 2831 	PATH: quote $(PATH_SEPARATOR) as well
 2832 
 2833 	On OS/2, $(PATH_SEPARATOR) is ';'. Without quote, it is recognized as
 2834 	a mark of end of sentence.
 2835 
 2836 	* Makefile.am: quote $(PATH_SEPARATOR) as well.
 2837 	* t/Makefile.inc: Likewise.
 2838 
 2839 2014-12-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2840 
 2841 	Improve detection of GNU make, avoiding "Arg list too long" errors.
 2842 
 2843 	Such errors could take place when the main makefile included too many
 2844 	sub-makefiles, making $(MAKEFILE_LIST) too long and causing the
 2845 	recipes $(am__is_gnu_make) to exceed the shell's command-line length
 2846 	limits.  This is not a theoretical issue: it could happen for projects
 2847 	having lots of C/C++ sources and using automatic dependency tracking,
 2848 	which created an included .Po sub-makefile for each of such sources.
 2849 
 2850 	Fixes http://debbugs.gnu.org/18744
 2851 
 2852 	* lib/am/header-vars.am (am__is_gnu_make): Fix the logic to avoid
 2853 	the use of $(MAKEFILE_LIST).
 2854 	* NEWS: Update.
 2855 
 2856 2014-12-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2857 
 2858 	Merge branch 'minor'
 2859 
 2860 	* minor:
 2861 	  Fix stupid typo in test, causing spurious failure
 2862 	  sync: update third-part files from upstream
 2863 	  Make sure AM_INIT_AUTOMAKE has a trailing newline
 2864 	  dist: adjust warning messages about shar and tarZ deprecation
 2865 	  docs: improve description of ${PACKAGE}, ${VERSION}, and similar variables
 2866 
 2867 2014-12-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2868 
 2869 	Merge branch 'micro' into minor
 2870 
 2871 	* micro:
 2872 	  Fix stupid typo in test, causing spurious failure
 2873 
 2874 2014-12-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2875 
 2876 	Fix stupid typo in test, causing spurious failure
 2877 
 2878 	* t/aminit-trailing-dnl-comment-pr16841.sh: s/greop/grep/
 2879 
 2880 2014-12-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2881 
 2882 	Merge branch 'micro' into minor
 2883 
 2884 	* micro:
 2885 	  sync: update third-part files from upstream
 2886 	  Make sure AM_INIT_AUTOMAKE has a trailing newline
 2887 	  dist: adjust warning messages about shar and tarZ deprecation
 2888 	  docs: improve description of ${PACKAGE}, ${VERSION}, and similar variables
 2889 
 2890 2014-12-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2891 
 2892 	Merge branch 'pr16841' into micro
 2893 
 2894 	* pr16841:
 2895 	  Make sure AM_INIT_AUTOMAKE has a trailing newline
 2896 
 2897 2014-12-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2898 
 2899 	sync: update third-part files from upstream
 2900 
 2901 	* lib/config.guess: This.
 2902 	* lib/config.sub: And this.
 2903 	* lib/gitlog-to-changelog: And this.
 2904 	* lib/gnupload: And this.
 2905 	* lib/update-copyright: And this.
 2906 
 2907 2014-12-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2908 
 2909 	Make sure AM_INIT_AUTOMAKE has a trailing newline
 2910 
 2911 	This used to be the case until Automke 1.13, but we broke it in
 2912 	Automake 1.14 (see commit v1.13.1-71-gf78b0f0).  This caused
 2913 	issues like http://debbugs.gnu.org/16841
 2914 
 2915 	* m4/init.m4 (AM_INIT_AUTOMAKE): Adjust.
 2916 	* t/aminit-trailing-dnl-comment-pr16841.sh: New test.
 2917 	* t/list-of-tests.mk: Add it.
 2918 	* NEWS, THANKS: Update.
 2919 
 2920 2014-12-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2921 
 2922 	Merge branches 'dist-warn-pr19108' and 'docs-pr16623' into micro
 2923 
 2924 	* dist-warn-pr19108:
 2925 	  dist: adjust warning messages about shar and tarZ deprecation
 2926 
 2927 	* docs-pr16623:
 2928 	  docs: improve description of ${PACKAGE}, ${VERSION}, and similar variables
 2929 
 2930 2014-12-19  Aharon Robbins  <arnold@skeeve.com>
 2931 
 2932 	dist: adjust warning messages about shar and tarZ deprecation
 2933 
 2934 	They were swapped.  Reported in http://debbugs.gnu.org/19108.
 2935 
 2936 2014-12-17  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2937 
 2938 	docs: improve description of ${PACKAGE}, ${VERSION}, and similar variables
 2939 
 2940 	In particular, suggesting that $PACKAGE and $VERSION should be aliases of
 2941 	the $PACKAGE_TARNAME and $PACKAGE_VERSION definitions coming from AC_INIT,
 2942 	and not be defined via an obsolete 2-argument invocation of AM_INIT_AUTOMAKE;
 2943 	and why that is the best default, given all our historical baggage.
 2944 
 2945 	See discussion in http://debbugs.gnu.org/16623 for more information and
 2946 	background.
 2947 
 2948 	* doc/automake.texi: Adjust.
 2949 	* THANKS: Update.
 2950 
 2951 2014-12-17  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2952 
 2953 	Merge branch 'minor'
 2954 
 2955 	* minor:
 2956 	  Fix dumb logic error preventing $install_sh from being be overridden
 2957 	  Automake docs: fix typos and use of British English
 2958 	  Expose automake bug#19311
 2959 	  build: fix race in parallel builds
 2960 	  build: fix race in parallel builds
 2961 
 2962 2014-12-17  Thomas Jahns  <jahns@dkrz.de>
 2963 
 2964 	Fix dumb logic error preventing $install_sh from being be overridden
 2965 
 2966 	* m4/install-sh.m4: Here.
 2967 	* THANKS: Update.
 2968 
 2969 2014-12-17  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2970 
 2971 	Merge branches 'pr19311' and 'micro' into minor
 2972 
 2973 	* pr19311:
 2974 	  Expose automake bug#19311
 2975 	  build: fix race in parallel builds
 2976 
 2977 	* micro:
 2978 	  Automake docs: fix typos and use of British English
 2979 	  build: fix race in parallel builds
 2980 
 2981 2014-12-17  Karl Berry  <karl@freefriends.org>
 2982 
 2983 	Automake docs: fix typos and use of British English
 2984 
 2985 	* doc/automake.texi: Here.
 2986 
 2987 2014-12-17  Stefano Lattarini  <stefano.lattarini@gmail.com>
 2988 
 2989 	Expose automake bug#19311
 2990 
 2991 	AC_PROG_CC called before AC_CONFIG_AUX_DIR can silently force wrong
 2992 	$ac_aux_dir definition.
 2993 
 2994 	* t/auxdir-pr19311.sh: New.
 2995 	* t/list-of-tests.mk: Add it as an XFAIL test.
 2996 
 2997 2014-12-17  Paul Eggert  <eggert@cs.ucla.edu>
 2998 
 2999 	build: fix race in parallel builds
 3000 
 3001 	Reported by Friedrich Beckmann in: http://bugs.gnu.org/18301
 3002 	* lib/am/texi-vers.am (?DIRSTAMP?): Put the process-ID into the
 3003 	temporary file name.  Use a similar temporary in the source dir.
 3004 
 3005 2014-08-23  Paul Eggert  <eggert@cs.ucla.edu>
 3006 
 3007 	build: fix race in parallel builds
 3008 
 3009 	Reported by Friedrich Beckmann in: http://bugs.gnu.org/18301
 3010 	* lib/am/texi-vers.am (?DIRSTAMP?): Put the process-ID into the
 3011 	temporary file name.  Use a similar temporary in the source dir.
 3012 
 3013 2014-04-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3014 
 3015 	Merge branch 'minor'
 3016 
 3017 	* minor:
 3018 	  init: ensure $ac_aux_dir is defined before being used
 3019 
 3020 2014-04-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3021 
 3022 	init: ensure $ac_aux_dir is defined before being used
 3023 
 3024 	Since we use '$ac_aux_dir' to define '$am_aux_dir', we need
 3025 	to ensure the former has been initialized before we try to
 3026 	define the latter, otherwise the definition:
 3027 
 3028 	   am_aux_dir=`cd $ac_aux_dir && pwd`
 3029 
 3030 	will set '$am_aux_dir' to '$HOME', likely causing weird and
 3031 	unexpected behaviours.
 3032 
 3033 	This change fixes automake bug#15981.
 3034 
 3035 	* m4/auxdir.m4 (AM_AUX_DIR_EXPAND): AC_REQUIRE expansion
 3036 	of 'AC_CONFIG_AUX_DIR_DEFAULT'. Fix redundant comment and
 3037 	AC_PREREQ, add extra quoting around '$ac_aux_dir'.
 3038 	* t/auxdir-pr15981.sh: New test.
 3039 	* t/auxdir-cc-pr15981.sh: Likewise.
 3040 	* t/list-of-tests.mk (handwritten_TESTS): Add them.
 3041 	* THANKS, NEWS: Update.
 3042 	* bin/automake.in: Fix a harmless typo in comments, that
 3043 	I happened to notice while writing this patch.
 3044 
 3045 2014-04-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3046 
 3047 	Merge branch 'minor'
 3048 
 3049 	* minor:
 3050 	  plans: enabling subdir-object by default is blocked on bug#13928
 3051 
 3052 2014-04-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3053 
 3054 	plans: enabling subdir-object by default is blocked on bug#13928
 3055 
 3056 2014-04-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3057 
 3058 	maint: update copyright (for files in 'master' only)
 3059 
 3060 2014-04-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3061 
 3062 	Merge branch 'minor'
 3063 
 3064 	* minor:
 3065 	  maint: update copyright years
 3066 	  maint: sync files from upstream ("make fetch")
 3067 
 3068 2014-04-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3069 
 3070 	maint: update copyright years
 3071 
 3072 	We've been in 2014 already for few months now...
 3073 
 3074 2014-04-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3075 
 3076 	maint: sync files from upstream ("make fetch")
 3077 
 3078 2014-04-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3079 
 3080 	Merge branch 'minor'
 3081 
 3082 	* minor:
 3083 	  Typofixes in warning messages and manual
 3084 
 3085 2014-04-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3086 
 3087 	Typofixes in warning messages and manual
 3088 
 3089 	Fixes automake bug#16827 and bug#16997.
 3090 
 3091 2014-04-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3092 
 3093 	Merge branch 'minor'
 3094 
 3095 	* minor:
 3096 	  doc: fix encoding error with UTF-8 characters
 3097 	  NEWS: a typofix, and better word wrapping
 3098 	  parallel-tests: avoid possible implicit "make all" in test-suite.log rule
 3099 
 3100 2014-04-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3101 
 3102 	Merge branch 'micro' into minor
 3103 
 3104 	* micro:
 3105 	  doc: fix encoding error with UTF-8 characters
 3106 
 3107 2014-01-04  Paul Eggert  <eggert@cs.ucla.edu>
 3108 
 3109 	doc: fix encoding error with UTF-8 characters
 3110 
 3111 	* doc/automake.texi: Specify @documentencoding and
 3112 	@documentlanguage, to prevent encoding errors for parts of this
 3113 	input file that are UTF-8.  This also causes the .info output to
 3114 	use curly quotes, which is easier to read though it does assume
 3115 	UTF-8 support.
 3116 
 3117 2013-12-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3118 
 3119 	NEWS: a typofix, and better word wrapping
 3120 
 3121 2013-12-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3122 
 3123 	parallel-tests: avoid possible implicit "make all" in test-suite.log rule
 3124 
 3125 	This change fixes automake bug#16302.
 3126 
 3127 	* lib/am/check.am ($(TEST_SUITE_LOG)): Avoid running "make $redo_logs"
 3128 	when $redo_logs expands to empty, since in that case we are actually
 3129 	ending up invoking a full "make all".  That shouldn't be required, and
 3130 	can cause slowdowns for people implementing their extra "laziness
 3131 	wrappers" around check-TESTS (automake bug#16302).
 3132 	* NEWS: Update.
 3133 
 3134 2013-12-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3135 
 3136 	Merge branch 'minor'
 3137 
 3138 	* minor:
 3139 	  Allow user to extend .PRECIOUS target
 3140 	  cosmetics: remove a couple of extra trailing white spaces
 3141 	  tests: fix a spurious failure on Mac OS X
 3142 	  docs: make clear the JAVA primary is frozen
 3143 	  install-sh: a slightly better diagnostic, and tests enhancements
 3144 	  install-sh: be stricter in catching invalid usages
 3145 	  tests: more significant names for some tests
 3146 	  tests: some cosmetic fixes
 3147 	  tests: more significant names for a test
 3148 	  docs: drop a few obsolescent FIXME/TODO comments, and associated text
 3149 	  testsuite harness: report test exit status in log file
 3150 	  TAP driver: no need to invoke AC_PROG_AWK directly
 3151 	  TAP driver: remove perl implementation (move it into contrib/)
 3152 
 3153 2013-12-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3154 
 3155 	Allow user to extend .PRECIOUS target
 3156 
 3157 	References:
 3158 	<http://lists.freedesktop.org/archives/systemd-devel/2013-July/012155.html>
 3159 	<http://lists.gnu.org/archive/html/automake/2013-07/msg00011.html>
 3160 
 3161 	* bin/automake.in: Adjust to ensure we handle '.PRECIOUS' the same way
 3162 	we do for '.PHONY' and '.MAKE'.
 3163 	* lib/Automake/Rule.pm: Likewise.
 3164 	* t/precious.sh: New test.
 3165 	* t/list-of-tests.mk: Add it.
 3166 	* t/phony.sh: Enhance a little while at it.
 3167 	* NEWS: Update.
 3168 	* THANKS: Likewise.
 3169 
 3170 	Reported-by: Holger Hans Peter Freyther <holger@freyther.de>
 3171 
 3172 2013-12-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3173 
 3174 	cosmetics: remove a couple of extra trailing white spaces
 3175 
 3176 2013-12-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3177 
 3178 	tests: fix a spurious failure on Mac OS X
 3179 
 3180 	This change fixes bug#14706.
 3181 
 3182 	* lib/depcomp2.sh: Also cater to spurious diagnostic from GNU rm,
 3183 	not only from Apple's rm.
 3184 
 3185 2013-12-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3186 
 3187 	docs: make clear the JAVA primary is frozen
 3188 
 3189 	* doc/automake.texi: Here.  The JAVA primary is broken in several ways,
 3190 	and will no longer be developed, not even for bug fixes.
 3191 
 3192 	See also automake bugs #9088, #8662 and #8540.
 3193 
 3194 2013-12-25  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3195 
 3196 	Merge branch 'install-sh-improvements' into minor
 3197 
 3198 	* install-sh-improvements:
 3199 	  install-sh: a slightly better diagnostic, and tests enhancements
 3200 	  install-sh: be stricter in catching invalid usages
 3201 
 3202 2013-12-25  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3203 
 3204 	install-sh: a slightly better diagnostic, and tests enhancements
 3205 
 3206 	* lib/install-sh: When called with no non-option arguments and the '-t'
 3207 	option with an argument that is not an existing directory, have the
 3208 	diagnostic output complain about the lack of required arguments rather
 3209 	than about the bad argument passed to '-t'.
 3210 	* t/install-sh-unittests.sh: Enhance to also check diagnostic printed
 3211 	in cases of expected failure.
 3212 
 3213 2013-12-25  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3214 
 3215 	install-sh: be stricter in catching invalid usages
 3216 
 3217 	Such usages (which are rejected by GNU install as well) are:
 3218 
 3219 	  - options -d and -t used together;
 3220 
 3221 	  - argument passed to option -t must be a directory;
 3222 
 3223 	  - if there are two or more SOURCEFILE arguments, the
 3224 	    DESTINATION argument must be a directory.
 3225 
 3226 	Note that we still allow the use of options -d and -T together, by
 3227 	making -d take the precedence; this is for compatibility with GNU
 3228 	install.
 3229 
 3230 	This change fixes, among other things, automake bug#15376.
 3231 
 3232 	* lib/install-sh: Adjust.
 3233 	* t/install-sh-unittests.sh: Enhance.
 3234 	* NEWS: Update.
 3235 	* THANKS: Add reporter of bug#15376.
 3236 
 3237 	Helped-by: Tobias Hansen <thansen@debian.org>
 3238 
 3239 2013-12-25  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3240 
 3241 	tests: more significant names for some tests
 3242 
 3243 	* t/insh2.sh: Rename...
 3244 	* t/dist-install-sh.sh: ... like this.
 3245 	* t/instsh.sh: Rename...
 3246 	* t/add-missing-install-sh.sh: ... like this.
 3247 	* t/instsh2.sh: Rename...
 3248 	* t/install-sh-unittests.sh: ... like this.
 3249 	* t/instsh3.sh: Rename...
 3250 	* t/install-sh-option-C.sh: ... like this.
 3251 	* t/list-of-tests.mk: Adjust.
 3252 
 3253 2013-12-25  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3254 
 3255 	tests: some cosmetic fixes
 3256 
 3257 	* t/instdat.sh: Here.
 3258 	* t/instdat2.sh: And here.
 3259 	* t/instsh.sh: And here.
 3260 	* t/instsh3.sh: And here.
 3261 
 3262 2013-12-25  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3263 
 3264 	tests: more significant names for a test
 3265 
 3266 	* t/install2.sh: Rename...
 3267 	* t/dist-with-unreadable-makefile-fails.sh: ... like this.
 3268 	* t/list-of-tests.mk: Adjust.
 3269 
 3270 2013-12-25  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3271 
 3272 	Merge branches 'fix-pr11814' and 'drop-perl-tap-driver' into minor
 3273 
 3274 	* fix-pr11814:
 3275 	  docs: drop a few obsolescent FIXME/TODO comments, and associated text
 3276 	  testsuite harness: report test exit status in log file
 3277 
 3278 	* drop-perl-tap-driver:
 3279 	  TAP driver: remove perl implementation (move it into contrib/)
 3280 
 3281 2013-12-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3282 
 3283 	docs: drop a few obsolescent FIXME/TODO comments, and associated text
 3284 
 3285 2013-12-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3286 
 3287 	testsuite harness: report test exit status in log file
 3288 
 3289 	The exit status of a test should be reported in the test logs, so
 3290 	that one can see at a glance whether the test has succeeded or failed,
 3291 	without having to look also into the corresponding .trs file.
 3292 
 3293 	This fixes automake bug#11814.
 3294 
 3295 	* lib/test-driver: Also report the test script exit status in the
 3296 	test log (as the last line).
 3297 	* t/check-exit-status-reported.sh: Test this new behaviour.
 3298 	* t/list-of-tests.mk: Add the new test.
 3299 	* t/ax/test-lib.sh( am_exit_trap): No longer log the test exit status;
 3300 	this has been made redundant by the change to 'test-driver'.  While at
 3301 	it, fix an imperfect quoting.
 3302 
 3303 2013-12-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3304 
 3305 	TAP driver: no need to invoke AC_PROG_AWK directly
 3306 
 3307 	It is already required by AM_INIT_AUTOMAKE anyway.
 3308 
 3309 	* doc/automake.texi: Adjust examples.
 3310 	* t/tap-doc2.sh: Adjust documentation-tracking test.
 3311 	* m4/init.m4 (AM_INIT_AUTOMAKE): Explicitly tell that the AC_PROG_AWK
 3312 	requirement is also needed whenever the TAP driver is used.
 3313 
 3314 2013-12-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3315 
 3316 	TAP driver: remove perl implementation (move it into contrib/)
 3317 
 3318 	That implementation was only meant as a standard against which the
 3319 	portable awk+shell implementation was to be measured.  Now, since
 3320 	Automake 1.12, the latter implementation is fully functional and
 3321 	already used in the wild, and in fact feature-par with the perl
 3322 	implementation.  So the perl implementation is now just slowing
 3323 	down and complicating our testsuite.  Let's move it to 'contrib/'
 3324 	(we don't want to remove it, in case someone is actually using it
 3325 	in the wild).
 3326 
 3327 	* lib/tap-driver.pl: Move it ...
 3328 	* contrib/tap-driver.pl: ... here.  While at it, convert quoting
 3329 	`like this' to quoting 'like this', and remove an obsolescent FIXME
 3330 	comment.
 3331 	* lib/Makefile.inc (dist_script_DATA): Drop '%D%/tap-driver.pl'.
 3332 	* Makefile.am (EXTRA_DIST): Add 'contrib/tap-driver.pl'.
 3333 	* doc/automake.texi: Remove one stray reference to 'tap-driver.pl',
 3334 	and reference 'tap-driver.sh' instead, as intended.
 3335 	* t/ax/am-test-lib.sh ($am_tap_implementation): Delete definition and
 3336 	uses.
 3337 	(fetch_tap_driver): Simplify to unconditionally assume the shell+awk
 3338 	implementation of the TAP driver is used.
 3339 	(get_shell_script): Make more flexible so that it can cater to the
 3340 	needs of 'fetch_tap_driver()'.
 3341 	* t/tap-bad-prog.tap: Likewise.
 3342 	* t/tap-bailout-leading-space.sh: Likewise.
 3343 	* t/tap-signal.tap: Likewise.
 3344 	* t/tap-test-number-0.sh: Likewise.
 3345 	* t/test-driver-cond.sh: Use 'tap-driver.sh' instead of 'tap-driver.pl'.
 3346 	* gen-testsuite-part (%test_generators): Do not generate sister tests
 3347 	that use the perl TAP driver rather than the shell+awk one.
 3348 	* NEWS: Update.
 3349 
 3350 2013-12-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3351 
 3352 	Merge branch 'minor'
 3353 
 3354 	* minor:
 3355 	  NEWS: stop reporting "new" Automake versioning scheme
 3356 	  post-release: micro version bump to 1.14.1a devel version
 3357 	  release: stable micro release 1.14.1
 3358 	  HACKING: minor clarification
 3359 	  tests: make install-info-dir.sh print more debugging info
 3360 	  tests: remove too-brittle test tap-realtime.sh
 3361 	  maintainer: am-ft: add option to cater to clock skews
 3362 	  sync: update INSTALL, config.guess and config.sub from upstream
 3363 	  TAP driver: cosmetic fixes
 3364 
 3365 2013-12-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3366 
 3367 	NEWS: stop reporting "new" Automake versioning scheme
 3368 
 3369 	It's actually old news by now.
 3370 
 3371 2013-12-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3372 
 3373 	Merge branch 'micro' into minor
 3374 
 3375 	* micro:
 3376 	  post-release: micro version bump to 1.14.1a devel version
 3377 	  release: stable micro release 1.14.1
 3378 	  HACKING: minor clarification
 3379 	  tests: make install-info-dir.sh print more debugging info
 3380 	  tests: remove too-brittle test tap-realtime.sh
 3381 	  maintainer: am-ft: add option to cater to clock skews
 3382 	  sync: update INSTALL, config.guess and config.sub from upstream
 3383 	  TAP driver: cosmetic fixes
 3384 
 3385 2013-12-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3386 
 3387 	post-release: micro version bump to 1.14.1a devel version
 3388 
 3389 	* configure.ac (AC_INIT): Bump version number to 1.14.1a.
 3390 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 3391 
 3392 2013-12-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3393 
 3394 	release: stable micro release 1.14.1
 3395 
 3396 	* configure.ac (AC_INIT): Bump version number to 1.14.1.
 3397 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 3398 
 3399 2013-12-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3400 
 3401 	HACKING: minor clarification
 3402 
 3403 2013-12-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3404 
 3405 	tests: make install-info-dir.sh print more debugging info
 3406 
 3407 	With the hope that this will shed more light on bug#14601
 3408 
 3409 2013-12-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3410 
 3411 	tests: remove too-brittle test tap-realtime.sh
 3412 
 3413 	* t/tap-realtime.sh: Delete.  It has always been brittle, but now
 3414 	it's also causing spurious failures when mawk is used as the awk
 3415 	implementation in tap-driver.sh (see bug#14601).
 3416 	* t/list-of-tests.mk: Adjust.
 3417 
 3418 2013-12-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3419 
 3420 	maintainer: am-ft: add option to cater to clock skews
 3421 
 3422 	* maintainer/am-ft: Add option '-S', giving a number of seconds to sleep
 3423 	after copying the tarball to the remote system and before unpacking,
 3424 	building and testing it.  This is to cater to situations where the clock
 3425 	of the remote system is skewed (in the past) w.r.t. the clock the local
 3426 	system the tarball has been built on.
 3427 
 3428 2013-12-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3429 
 3430 	sync: update INSTALL, config.guess and config.sub from upstream
 3431 
 3432 2013-12-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3433 
 3434 	TAP driver: cosmetic fixes
 3435 
 3436 	* lib/tap-driver.sh: Quote 'like this', not `like this'.  Remove an
 3437 	obsolete FIXME.  Correctly mark another comment as a TODO rather
 3438 	than as a FIXME.
 3439 
 3440 2013-11-02  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3441 
 3442 	Merge branch 'minor'
 3443 
 3444 	* minor:
 3445 	  cosmetics: fix typo in a user-facing message in tests
 3446 	  automake: account for perl hash order randomization
 3447 	  tests: avoid use of intervals to capitalize letters
 3448 	  cosmetics: untabify the install-sh script
 3449 	  install-sh: assume that "set -f" and "set +f" work...
 3450 	  install-sh: assume ${var:-value} works as expected
 3451 	  install-sh: assume 'dirname' is available and working correctly
 3452 	  distcheck: don't allow overriding of --prefix and --srcdir by the user
 3453 	  tests: expose bug#14991 (relates to 'distcheck')
 3454 
 3455 2013-11-01  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3456 
 3457 	Merge branch 'modernize-install-sh' into minor
 3458 
 3459 	* modernize-install-sh:
 3460 	  cosmetics: untabify the install-sh script
 3461 	  install-sh: assume that "set -f" and "set +f" work...
 3462 	  install-sh: assume ${var:-value} works as expected
 3463 	  install-sh: assume 'dirname' is available and working correctly
 3464 
 3465 2013-11-01  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3466 
 3467 	Merge branch 'micro' into minor
 3468 
 3469 	* micro:
 3470 	  cosmetics: fix typo in a user-facing message in tests
 3471 	  automake: account for perl hash order randomization
 3472 	  tests: avoid use of intervals to capitalize letters
 3473 	  distcheck: don't allow overriding of --prefix and --srcdir by the user
 3474 	  tests: expose bug#14991 (relates to 'distcheck')
 3475 
 3476 2013-11-01  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3477 
 3478 	cosmetics: fix typo in a user-facing message in tests
 3479 
 3480 	* t/lex-header.sh: A "skip" message in this test, precisely.
 3481 
 3482 2013-11-01  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3483 
 3484 	Merge branches 'fix-pr14991' and 'fix-pr14891' into micro
 3485 
 3486 	* fix-pr14991:
 3487 	  distcheck: don't allow overriding of --prefix and --srcdir by the user
 3488 	  tests: expose bug#14991 (relates to 'distcheck')
 3489 
 3490 	* fix-pr14891:
 3491 	  automake: account for perl hash order randomization
 3492 	  tests: avoid use of intervals to capitalize letters
 3493 
 3494 2013-10-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3495 
 3496 	automake: account for perl hash order randomization
 3497 
 3498 	Try to explicitly order the keys of some perl hashes when looping
 3499 	on them to do sanity/correctness checks and possibly display warning
 3500 	messages; this should ensure a more reproducible output.  Not really
 3501 	a big deal, but I prefer to keep the order of such output reproducible
 3502 	if possible.
 3503 
 3504 	Issue revealed by spurious testsuite failures with perl 5.18, as
 3505 	reported in automake bug#14891.  See also:
 3506 	<http://search.cpan.org/dist/perl-5.18.0/pod/perldelta.pod#Hash_randomization>
 3507 	<http://onionstand.blogspot.ie/2012/12/are-you-relying-on-hash-keys-being.html>
 3508 
 3509 	* lib/Automake/Variable.pm (variables): Explicitly order the values of
 3510 	the returned Automake::Variable instances.
 3511 	(variables_dump): Simplify, using the knowledge that 'variables()' now
 3512 	sorts its output.
 3513 	* t/preproc-errmsg.sh: Adjust.
 3514 
 3515 2013-10-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3516 
 3517 	tests: avoid use of intervals to capitalize letters
 3518 
 3519 	It was causing spurious failures with with Solaris 8 'tr'.
 3520 	See automake bug#14891.
 3521 
 3522 	* t/test-extensions.sh: Adjust.
 3523 
 3524 2013-10-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3525 
 3526 	cosmetics: untabify the install-sh script
 3527 
 3528 	* lib/install-sh: Here, plus a couple of related formatting tweaks.
 3529 
 3530 2013-10-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3531 
 3532 	install-sh: assume that "set -f" and "set +f" work...
 3533 
 3534 	... and disable/enable shell globbing, respectively.  This is
 3535 	mandated by POSIX, and supported even by Solaris 9 /bin/sh (one
 3536 	of the most braindead shells we still support).
 3537 
 3538 	* lib/install.sh: Adjust.
 3539 	* NEWS: Update.
 3540 
 3541 2013-10-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3542 
 3543 	install-sh: assume ${var:-value} works as expected
 3544 
 3545 	The Autoconf manual says it is OK these days.
 3546 
 3547 	* lib/install.sh: Adjust.
 3548 	* NEWS: Update.
 3549 
 3550 2013-10-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3551 
 3552 	install-sh: assume 'dirname' is available and working correctly
 3553 
 3554 	Really, we no longer care about hosts so outdated/broken to miss
 3555 	fundamental utilities like basename or dirname.
 3556 
 3557 	* lib/install.sh: Adjust.
 3558 	* NEWS, THANKS: Update.
 3559 
 3560 	Suggested-by: Philipp A. Hartmann <philipp.hartmann@offis.de>
 3561 
 3562 2013-10-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3563 
 3564 	distcheck: don't allow overriding of --prefix and --srcdir by the user
 3565 
 3566 	Not through AM_DISTCHECK_FLAGS, nor through DISTCHECK_FLAGS.  Apparently,
 3567 	some packages got in the habit of relaying all the options passed to the
 3568 	original ./configure invocation through to the configure invocations
 3569 	in "make distcheck".  This was causing problems, because it also passed
 3570 	through the original --srcdir and --prefix options.
 3571 
 3572 	Fixes: expose bug#14991 (relates to 'distcheck')
 3573 
 3574 	* lib/am/distdir.am (distcheck): Pass the hard-coded --srcdir and
 3575 	--prefix options *after* both the developer-defined options in
 3576 	$(AM_DISTCHECK_FLAGS) and the user-defined options in $(DISTCHECK_FLAGS).
 3577 	* t/list-of-tests.mk (XFAIL_TESTS): Remove the now-passing test
 3578 	'distcheck-no-destdist-or-srcdir-override.sh'.
 3579 	* doc/automake.texi (Checking the Distribution): Update.
 3580 	* NEWS: Likewise.
 3581 
 3582 2013-10-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3583 
 3584 	tests: expose bug#14991 (relates to 'distcheck')
 3585 
 3586 	* t/distcheck-no-prefix-or-srcdir-override.sh: New, expose the bug.
 3587 	* t/list-of-tests.mk (handwritten_TESTS, XFAIL_TESTS): Add it.
 3588 
 3589 2013-10-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3590 
 3591 	Merge branch 'minor'
 3592 
 3593 	* minor:
 3594 	  tests: fix spurious failure when zip is present but unzip is not
 3595 	  tests: fix spurious failure due to localization issues
 3596 	  NEWS: update with the changes since v1.14
 3597 	  docs: correct typos in the fix-timestamp.sh script
 3598 	  python: byte-compile nobase_*_PYTHON files only once
 3599 	  cosmetics: typofix in the 'missing' script
 3600 	  test: avoid false positives in 'cc-no-c-o' script
 3601 	  test harness: improve catching of usage errors in script 'test-driver'
 3602 	  tests: fix a spurious failure on NetBSD-current
 3603 	  am-ft: make the environment available earlier
 3604 	  NEWS: post-release tweaks (for 1.14.x series)
 3605 	  tests: avoid a spurious failure on MacOS X 10.6.8
 3606 	  tests: don't risk hanging on the 'cl' requirement
 3607 
 3608 2013-10-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3609 
 3610 	Merge branch 'micro' into minor
 3611 
 3612 	* micro:
 3613 	  tests: fix spurious failure when zip is present but unzip is not
 3614 	  tests: fix spurious failure due to localization issues
 3615 	  NEWS: update with the changes since v1.14
 3616 	  docs: correct typos in the fix-timestamp.sh script
 3617 	  python: byte-compile nobase_*_PYTHON files only once
 3618 	  cosmetics: typofix in the 'missing' script
 3619 	  test: avoid false positives in 'cc-no-c-o' script
 3620 	  test harness: improve catching of usage errors in script 'test-driver'
 3621 	  tests: fix a spurious failure on NetBSD-current
 3622 	  am-ft: make the environment available earlier
 3623 	  NEWS: post-release tweaks (for 1.14.x series)
 3624 	  tests: avoid a spurious failure on MacOS X 10.6.8
 3625 	  tests: don't risk hanging on the 'cl' requirement
 3626 
 3627 2013-10-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3628 
 3629 	tests: fix spurious failure when zip is present but unzip is not
 3630 
 3631 	Fixes automake bug#15181.
 3632 
 3633 	* t/dist-formats.tap (have_compressor): When checking that zip(1), also
 3634 	check for unzip(1), otherwise "make distcheck" will be unable to extract
 3635 	the zip tarball it creates, which will cause spurious failures.  While
 3636 	at it, reorganize the existing code a bit.
 3637 	* THANKS, NEWS: Update.
 3638 
 3639 2013-10-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3640 
 3641 	tests: fix spurious failure due to localization issues
 3642 
 3643 	Fixes automake bug#15237.
 3644 
 3645 	* t/autohdr-subdir-pr12495.sh: Ensure make is run in the C locale, so that
 3646 	we can expect error messages in English when grepping its output.
 3647 	* THANKS, NEWS: Update.
 3648 
 3649 2013-10-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3650 
 3651 	NEWS: update with the changes since v1.14
 3652 
 3653 2013-10-28  Peter Breitenlohner  <peb@mppmu.mpg.de>  (tiny change)
 3654 
 3655 	docs: correct typos in the fix-timestamp.sh script
 3656 
 3657 	* doc/automake.texi: Here.  The original version of this example script
 3658 	makes no sense at all, using 'configure' instead of the intended 'touch'
 3659 	in few key places.
 3660 
 3661 2013-10-28  Benoit Sigoure  <tsunanet@gmail.com>  (tiny change)
 3662 
 3663 	python: byte-compile nobase_*_PYTHON files only once
 3664 
 3665 	* lib/am/python.am: Here. Byte-compiling was occurring inside of
 3666 	a loop, causing an O(n^2) number of byte-compilations instead of
 3667 	O(n).
 3668 
 3669 2013-10-28  Václav Zeman  <vhaisman@gmail.com>  (tiny change)
 3670 
 3671 	cosmetics: typofix in the 'missing' script
 3672 
 3673 	* lib/missing: Here, in a message printed to the user.
 3674 	* THANKS: Update.
 3675 
 3676 2013-07-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3677 
 3678 	Merge branch 'fix-pr14911' into micro
 3679 
 3680 	* fix-pr14911:
 3681 	  test: avoid false positives in 'cc-no-c-o' script
 3682 
 3683 2013-07-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3684 
 3685 	Merge branch 'fix-half-pr14760' into micro
 3686 
 3687 	* fix-half-pr14760:
 3688 	  tests: fix a spurious failure on NetBSD-current
 3689 
 3690 2013-07-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3691 
 3692 	test: avoid false positives in 'cc-no-c-o' script
 3693 
 3694 	Fixes automake bug#14911.
 3695 
 3696 	* t/ax/cc-no-c-o.in: Be more careful in determining whether both the
 3697 	'-c' and '-o' options have been passed on the command line to the
 3698 	compiler.  In particular, do not spuriously complain in the face of
 3699 	options like '-compatibility_version' or '-current_version' (seen on
 3700 	Mac OS X 10.7).
 3701 	* THANKS: Update.
 3702 
 3703 2013-07-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3704 
 3705 	test harness: improve catching of usage errors in script 'test-driver'
 3706 
 3707 	Fixes automake bug#14840.
 3708 
 3709 	* lib/test-driver: Catch and report usage errors where the caller has
 3710 	forgotten to specify one of the mandatory options (--test-name,
 3711 	--log-file, --trs-file) or has not passed any non-option argument.
 3712 	Also, be sure to work correctly even when no '--' special argument
 3713 	is passed to separate option from non-options arguments.
 3714 	* THANKS: Update.
 3715 
 3716 2013-07-07  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3717 
 3718 	tests: fix a spurious failure on NetBSD-current
 3719 
 3720 	Reported in automake bug#14760.
 3721 
 3722 	* tests/silent-custom.sh: Be prepared to handle creative
 3723 	quoting in the output of the shell run for the make recipes
 3724 	when the shell traces are active ("set -x").
 3725 
 3726 2013-06-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3727 
 3728 	am-ft: make the environment available earlier
 3729 
 3730 	So that, with a PATH that is updated earlier, we can find the 'xz'
 3731 	program even on systems where it isn't in the default PATH.
 3732 
 3733 	* maintainer/am-ft: Adjust accordingly.
 3734 
 3735 2013-06-27  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3736 
 3737 	NEWS: post-release tweaks (for 1.14.x series)
 3738 
 3739 2013-06-27  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3740 
 3741 	tests: avoid a spurious failure on MacOS X 10.6.8
 3742 
 3743 	Fixes automake bug#14706.
 3744 
 3745 	* t/depcomp2.sh: Strip, from the redirected ./configure stderr, the
 3746 	possible error message "rm: conftest.dSYM: is a directory", generated
 3747 	by cleanup code that doesn't cater to the existence of *.dSYM
 3748 	directories sometimes created by the compiler on MacOS X.  This
 3749 	"massaging" of ./configure stderr is legitimate, since the spurious
 3750 	error message is due not to automake-related code, but to a know
 3751 	buglet/limitation of either Autoconf or Mac OS X bundles gcc:
 3752 	<http://lists.gnu.org/archive/html/bug-autoconf/2007-11/msg00017.html>
 3753 	Actually, from that link it appears that the original Autoconf issue
 3754 	had been fixed, but it must have been re-introduced in the meantime :-(
 3755 
 3756 2013-06-27  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3757 
 3758 	tests: don't risk hanging on the 'cl' requirement
 3759 
 3760 	On the GNU/Linux boxes of some users that run our testsuite there
 3761 	is a '/usr/local/bin/cl' executable, from the IRAF package:
 3762 
 3763 	    <http://iraf.noao.edu/>
 3764 
 3765 	The test 'compile4.sh' (and other tests) try to invoke the 'cl'
 3766 	command to check whether it's a Microsoft compiler; the IRAF cl
 3767 	is an interactive program, so it hangs on such invocation.  In
 3768 	conclusion, the testsuite hangs for those users which have the
 3769 	IRAF cl early in PATH.
 3770 
 3771 	Fix the issue by redirecting the input of cl from /dev/null when
 3772 	invoking it, which is enough to prevent the cl program from IRAF
 3773 	from hanging, and should have no effect on the behaviour of the
 3774 	Microsoft compiler.
 3775 
 3776 	This change fixes automake bug#14707.
 3777 
 3778 	* t/ax/am-test-lib.sh (require_tool): Adjust the handling of
 3779 	the 'cl' requirement.
 3780 
 3781 2013-06-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3782 
 3783 	Merge branch 'minor'
 3784 
 3785 	* minor:
 3786 	  post-release: micro version bump (1.14a)
 3787 	  post-release: micro version bump (1.14.0a)
 3788 	  release: stable minor release 1.14
 3789 
 3790 2013-06-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3791 
 3792 	Merge branch 'micro' into minor
 3793 
 3794 	* micro:
 3795 	  post-release: micro version bump (1.14.0a)
 3796 
 3797 2013-06-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3798 
 3799 	post-release: micro version bump (1.14a)
 3800 
 3801 	* configure.ac: Bump version: 1.14 -> 1.14a
 3802 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 3803 
 3804 2013-06-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3805 
 3806 	post-release: micro version bump (1.14.0a)
 3807 
 3808 	* configure.ac: Bump version: 1.14 -> 1.14.0a
 3809 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 3810 
 3811 2013-06-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3812 
 3813 	Merge branch 'maint' into micro
 3814 
 3815 	* maint:
 3816 	  release: stable minor release 1.14
 3817 	  NEWS: one more minor fixlet
 3818 	  maint: port check-minimal-autoconf to VPATH builds
 3819 	  tests: avoid spurious failures in Linux -> MinGW cross-compilation mode
 3820 	  tests: simplify checks for some expected variables values in Makefiles
 3821 	  NEWS: improve and update wording
 3822 	  tests: fix/improve few heading comments
 3823 	  coverage: new test on Texinfo @include support
 3824 	  tests: tighten a grepping check
 3825 	  tests: more significant names for some tests
 3826 	  rename-tests: rework some code for clarity and safety
 3827 	  tests: cosmetic changes in t/extra-sources.sh
 3828 	  rename-tests: inform the user about the pre-filled commit msg
 3829 	  typofix: in comments in 'maintainer/rename-tests'
 3830 	  rename-tests: also "git add" list-of-tests.mk
 3831 	  tests: rename t/exsource.sh -> t/extra-sources.sh
 3832 	  tests: some improvements to Gettext tests
 3833 	  maint: version bump after beta release 1.13b
 3834 	  release: beta release 1.13b (will become 1.14)
 3835 	  automake: assume we can always pass '-o' to the C compiler
 3836 	  NEWS: fix typo
 3837 	  NEWS: on assuming "rm -f" without arguments work
 3838 	  docs: AM_PROG_CC_C_O: correct imprecise statements about it
 3839 	  NEWS: document deprecation of 'shar' and 'compress' dist formats
 3840 	  m4: rename minuso.m4 -> prog-cc-c-o.m4
 3841 	  tests: some tests make no sense if "$CC -c -o" doesn't work
 3842 	  AM_PROG_CC_C_O: don't rely on AC_PROG_CC_C_O, re-implement similar logic
 3843 	  compile: rewrite AC_PROG_CC with AM_PROG_CC_C_O contents
 3844 	  tests: avoid few lingering $MAKE redirections
 3845 	  tests: fix a spurious failure on NetBSD 5.1
 3846 	  texi: build version.texi and stamp-vti in srcdir
 3847 	  tests: fix a botched heading comment
 3848 	  tests: fix another spurious with FreeBSD make
 3849 	  tests: fix a spurious failure with FreeBSD make
 3850 	  tests: remove remaining exec bits ('maint' branch)
 3851 	  PLANS: subdir-objects: various updates
 3852 	  THANKS: update Akim's e-mail address
 3853 	  tests: less uses of "make -e"; avoid spurious failures in 'check-cc-no-c-o'
 3854 	  build: be more respectful of user-specified verbosity
 3855 	  check-cc-no-c-o: unify initializations in a single place
 3856 	  check-cc-no-c-o: avoid a spurious failure
 3857 	  build: fixup for building in a VPATH setup
 3858 	  Use AC_DEFUN_ONCE to define AM_PROG_CC_C_O
 3859 	  compile: avoid AC_PROG_CC messy rewrite
 3860 	  options: tiny simplification in dealing with incompatible versions
 3861 	  options: try to report as much errors as possible
 3862 	  refactor: fix few "inverted boolean" usages
 3863 	  options: better name for an internal function
 3864 	  options: more consistency in use of return statuses to report errors
 3865 	  options: tiny simplification in dealing with erroneous opts
 3866 	  options: consistently use return statuses to report errors
 3867 	  options: re-enable some sanity checks
 3868 	  THANKS: update Eric Blake's e-mail address
 3869 	  NEWS: typofix
 3870 	  news: document new 'subdir-objects' warning
 3871 	  PLANS: one minor fixlet (mostly cosmetic)
 3872 	  PLANS: we have already dropped support for split info files in master
 3873 	  NEWS: fix a reference to Automake 1.14 where Automake 2.0 was intended
 3874 	  PLANS: fix reference to non-existent 'next' branch
 3875 	  PLANS: fix botched version reference
 3876 	  maintcheck: fix two references to old location of aclocal and automake
 3877 	  dist: deprecated shar and tar+compress formats
 3878 	  am: prefer a shorter idiom where possible
 3879 	  maint: re-run "make update-copyright" ...
 3880 	  tests: avoid spurious failure with older flex (2.5.4)
 3881 	  build: move automake and aclocal in 'bin' subdir
 3882 	  build: break up monolithic Makefile.am in subdir-specific fragments
 3883 	  maint branch: we are going to become Automake 1.14
 3884 	  tests: typofixes in comments in t/preproc-c-compile.sh
 3885 	  tests: remove bashism from a test
 3886 	  tests: rename some with more descriptive names
 3887 	  typofix: in comments in t/extra2.sh
 3888 	  tests: fix botched cross-reference in a heading comment
 3889 	  automake: refactoring: factor out common cpp-like flags
 3890 	  NEWS (mint): reflect new Automake versioning scheme
 3891 	  maintcheck: avoid spurious failure
 3892 	  perl: perl subroutine prototypes are problematic, don't use them
 3893 	  maint: more adjustments to the new versioning scheme
 3894 	  cosmetics: fix some "docstring-like" comments in automake
 3895 	  style: call perl functions 'like_this()', not '&like_this()'
 3896 	  preproc: enhance and extend tests
 3897 	  preproc: add support for relative names in included fragments
 3898 	  maint: use more perl subroutines prototypes in the automake script
 3899 	  build: auto-generate perl subroutines prototypes for automake and aclocal
 3900 	  refactor: rip module Automake::Language out of automake script
 3901 	  tests: more information about Lex and Yacc programs
 3902 	  lint: fix spurious failure for 'sc_rm_minus_f' syntax check
 3903 	  maint: bump version 1.13.1a -> 1.13.2a
 3904 	  plans: we are not going to remove AM_PROG_MKDIR_P in Automake 1.14
 3905 	  init.m4: add probe to check "rm -f" without args work
 3906 	  subdir-objects: complain if it isn't enabled
 3907 	  plans: update w.r.t. latest changes
 3908 	  ywrap: remove an obsolete FIXME comment
 3909 	  ywrap: style fixes (no semantic change intended)
 3910 	  convenience: "make lint" as an alias for "make maintainer-check"
 3911 	  docs: typofix in manual
 3912 	  coverage: using multiple lexers in a single program
 3913 	  tests: remove most uses of the AM_PROG_CC_C_O obsolete macro
 3914 	  coverage: obsolete macro AM_PROG_CC_C_O should cause no warning nor errors
 3915 	  INSTALL: update copyright years
 3916 	  ithreads: use runtime (not configure time) detection of perl threads
 3917 	  copyright: add few missing copyright notices
 3918 	  maint: files in PLANS are to be exempted from copyright notice
 3919 	  maint: consistently honor the UPDATE_COPYRIGHT_YEAR environment variable
 3920 	  copyright: update some copyright years
 3921 	  compile: use 'compile' script when "-c -o" is used with losing compilers
 3922 	  HACKING: suggest more checks before releasing
 3923 	  tests: can fake a compiler not grasping "-c -o" -- globally in all tests
 3924 	  sync: update files from upstream with "make fetch"
 3925 	  typofix: in comments in GNUmakefile
 3926 	  Rename 'maint/' -> 'maintainer/', for Git's sake
 3927 	  HACKING: minor typofix
 3928 	  HACKING: bug-tracker, the PLANS directory, and how to plan "big" changes
 3929 	  HACKING: rewindable branches should live in the 'experimental/*' namespace
 3930 	  HACKING: fixlets about git branch rewinding policy
 3931 	  HACKING: commit messages are not to follow GCS ChangeLog rules too strongly
 3932 	  HACKING: "detailed explanation" in commit messages is almost mandatory
 3933 	  HACKING: we use "merge --log" even when merging master
 3934 	  HACKING: typofix
 3935 	  depend2.am: fix comments on verbosity of compilation rules
 3936 	  depend2.am: improve comments a little
 3937 	  plans: automake 1.14 is to assume "rm -f" with no args is OK
 3938 	  plans: we want to active subdir-objects unconditionally in automake 1.14
 3939 	  tests: adjust stale references to old test names
 3940 	  tests: rename the last aclocal test with dumb name
 3941 	  tests: fix an old botched change to an aclocal test
 3942 	  tests: fix some botched inter-test references in heading comments
 3943 	  coverage: compile rules used "-c -o" also with losing compilers
 3944 	  texi: remove extra verbosity in creation of dirstamp directory
 3945 	  coverage: user can avoid distributing '.info' pages
 3946 	  plans: add some on-going plans (already registered on the bug tracker)
 3947 	  docs: mention dist-hook help for EXTRA_DIST
 3948 	  texi: remove workaround for older Texinfo (4.1)
 3949 	  NEWS: improve wordings in entry deprecating suffix-less info files
 3950 	  build: don't enable 'color-tests' automake option explicitly
 3951 	  build: enable all warnings as fatal in our own build system
 3952 	  texi: Texinfo sources and CLEANFILES definition should co-exist peacefully
 3953 	  tests: make two new test executable
 3954 	  runtest: better command line API
 3955 	  tests: move runtest.in away from the top-lever directory
 3956 	  maint: move more maintainer files in the 'maint/' subdir
 3957 	  plans: add the "PLANS" directory
 3958 	  tests: more significant names for some tests
 3959 	  maint: add some of my maintainer-specific scripts
 3960 	  texi: deprecate hack about info files in CLEANFILES variables
 3961 	  texi: info files can be generated in the builddir
 3962 
 3963 2013-06-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3964 
 3965 	release: stable minor release 1.14
 3966 
 3967 	* configure.ac (AC_INIT): Bump version number to 1.14.
 3968 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 3969 
 3970 2013-06-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3971 
 3972 	Merge branch 'maint'
 3973 
 3974 	* maint:
 3975 	  NEWS: one more minor fixlet
 3976 	  maint: port check-minimal-autoconf to VPATH builds
 3977 	  tests: avoid spurious failures in Linux -> MinGW cross-compilation mode
 3978 	  tests: simplify checks for some expected variables values in Makefiles
 3979 	  NEWS: improve and update wording
 3980 	  tests: fix/improve few heading comments
 3981 	  coverage: new test on Texinfo @include support
 3982 	  tests: tighten a grepping check
 3983 	  tests: more significant names for some tests
 3984 	  rename-tests: rework some code for clarity and safety
 3985 	  tests: cosmetic changes in t/extra-sources.sh
 3986 	  rename-tests: inform the user about the pre-filled commit msg
 3987 	  typofix: in comments in 'maintainer/rename-tests'
 3988 	  rename-tests: also "git add" list-of-tests.mk
 3989 	  tests: rename t/exsource.sh -> t/extra-sources.sh
 3990 
 3991 2013-06-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3992 
 3993 	NEWS: one more minor fixlet
 3994 
 3995 2013-06-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 3996 
 3997 	maint: port check-minimal-autoconf to VPATH builds
 3998 
 3999 	* maintainer/maint.mk (check-minimal-autoconf): Here.
 4000 
 4001 2013-06-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4002 
 4003 	tests: avoid spurious failures in Linux -> MinGW cross-compilation mode
 4004 
 4005 	* t/ccnoc-deps.sh: Here.
 4006 	* t/preproc-demo.sh: And here.
 4007 
 4008 2013-06-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4009 
 4010 	tests: simplify checks for some expected variables values in Makefiles
 4011 
 4012 	Do so by using our custom 'is' auxiliary script rather than grepping
 4013 	the output from make.  This is more natural, more robust, and often
 4014 	shorter to write.
 4015 
 4016 	Unfortunately, we can't do that in all cases: sometimes we really need
 4017 	to match the content of a variable against a regular expressions, and
 4018 	we can't know nor are interested in its exact value.
 4019 
 4020 	This is basically a follow-up on commit v1.11-1830-g96401cb of
 4021 	2012-02-08 (tests: better way to compare lists in Makefile rules).
 4022 
 4023 	* t/subst-no-trailing-empty-line.sh: Adjust.
 4024 	* t/pluseq10.sh: Likewise.
 4025 	* t/check5.sh: Likewise, and enhance a little while at it.
 4026 	* t/check7.sh: Likewise.
 4027 	* t/exeext.sh: Likewise.
 4028 
 4029 2013-06-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4030 
 4031 	NEWS: improve and update wording
 4032 
 4033 2013-06-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4034 
 4035 	tests: fix/improve few heading comments
 4036 
 4037 	* t/lflags.sh: Here.
 4038 	* t/lflags-cxx.sh: And here.
 4039 	* t/yflags.sh: And here.
 4040 	* t/yflags-cxx.sh: And here.
 4041 
 4042 2013-06-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4043 
 4044 	coverage: new test on Texinfo @include support
 4045 
 4046 	Backported from the Automake-NG testsuite.
 4047 
 4048 	* t/txinfo-include.sh: New test.
 4049 	* t/list-of-tests.mk: Add it.
 4050 
 4051 2013-06-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4052 
 4053 	tests: tighten a grepping check
 4054 
 4055 	* t/yflags-force-override.sh: Here, by being sure to correctly
 4056 	match an expected literal dot.
 4057 
 4058 2013-06-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4059 
 4060 	tests: more significant names for some tests
 4061 
 4062 	* t/extra2.sh: Rename...
 4063 	* t/extra-sources-no-spurious.sh: ... like this.
 4064 	* t/yflags2.sh: Rename...
 4065 	* t/yflags-cxx.sh: ... like this.
 4066 	* t/lflags2.sh: Rename...
 4067 	* t/lflags-cxx.sh: ... like this.
 4068 
 4069 2013-06-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4070 
 4071 	rename-tests: rework some code for clarity and safety
 4072 
 4073 2013-06-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4074 
 4075 	tests: cosmetic changes in t/extra-sources.sh
 4076 
 4077 	* t/extra-sources.sh: Do not create unneeded C sources.  Add
 4078 	trailing ':' command.
 4079 
 4080 2013-06-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4081 
 4082 	rename-tests: inform the user about the pre-filled commit msg
 4083 
 4084 2013-06-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4085 
 4086 	typofix: in comments in 'maintainer/rename-tests'
 4087 
 4088 2013-06-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4089 
 4090 	rename-tests: also "git add" list-of-tests.mk
 4091 
 4092 2013-06-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4093 
 4094 	tests: rename t/exsource.sh -> t/extra-sources.sh
 4095 
 4096 	* t/exsource.sh: Rename ...
 4097 	* t/extra-sources.sh: ... like this.
 4098 	* t/list-of-tests.mk: Adjust.
 4099 
 4100 2013-06-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4101 
 4102 	Merge branch 'maint'
 4103 
 4104 	* maint:
 4105 	  tests: some improvements to Gettext tests
 4106 
 4107 2013-06-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4108 
 4109 	Merge branch 'gettext-tests' into maint
 4110 
 4111 	* gettext-tests:
 4112 	  tests: some improvements to Gettext tests
 4113 
 4114 2013-06-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4115 
 4116 	tests: some improvements to Gettext tests
 4117 
 4118 	Mostly to bring them more in sync with the ones in Automake-NG.
 4119 	See also commit v1.12.2-824-g5468d52 of 2012-08-10([ng] tests:
 4120 	reorganize gettext tests a bit) in Automake-NG.
 4121 
 4122 	* t/gettext.sh: Rename ...
 4123 	* t/gettext-basics.sh: ... like this, enhance a little, and
 4124 	move checks on requirement of 'config.rpath' out into ...
 4125 	* t/gettext-config-rpath.sh: ... into this new test, and move
 4126 	checks about PR/381...
 4127 	* t/gettext-pr381.sh: ... into this new test.
 4128 	* t/gettext2.sh: Rename ...
 4129 	* t/gettext-external-pr338.sh: ... like this, and enhance a
 4130 	little.
 4131 	* t/gettext3.sh: Rename ...
 4132 	* t/gettext-intl-subdir.sh: ... like this, and add trailing
 4133 	':' command.
 4134 
 4135 2013-06-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4136 
 4137 	Merge branch 'maint'
 4138 
 4139 	* maint:
 4140 	  post-release: micro version bump (1.13.4a)
 4141 	  release: stable micro release 1.13.4
 4142 	  sync: update config.guess
 4143 
 4144 2013-06-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4145 
 4146 	Merge branch 'micro' into maint
 4147 
 4148 	* micro:
 4149 	  post-release: micro version bump (1.13.4a)
 4150 	  release: stable micro release 1.13.4
 4151 	  sync: update config.guess
 4152 
 4153 2013-06-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4154 
 4155 	post-release: micro version bump (1.13.4a)
 4156 
 4157 	* configure.ac: Bump version: 1.13.4 -> 1.13.4a
 4158 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 4159 
 4160 2013-06-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4161 
 4162 	release: stable micro release 1.13.4
 4163 
 4164 	* configure.ac (AC_INIT): Bump version number to 1.13.4.
 4165 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 4166 
 4167 2013-06-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4168 
 4169 	sync: update config.guess
 4170 
 4171 	* lib/config.guess: This.  Actually, only the timestamp has been
 4172 	updated (apparently, it was mistakenly not updated in the previous
 4173 	real change to the script).
 4174 
 4175 2013-06-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4176 
 4177 	Merge branch 'maint'
 4178 
 4179 	* maint:
 4180 	  THANKS: update e-mall address for Ralf Corsepius
 4181 	  lang, suffix rules: don't require C stuff needlessly
 4182 	  tests: expose automake bug#14560
 4183 	  maint: add a missing copyright notice
 4184 	  sync: update config.guess from upstream
 4185 	  tests: expose automake bug#13928
 4186 	  comments: fix some out-of-sync refs to test scripts
 4187 	  tests: expose automake bug#13940
 4188 
 4189 2013-06-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4190 
 4191 	Merge branch 'micro' into maint
 4192 
 4193 	* micro:
 4194 	  THANKS: update e-mall address for Ralf Corsepius
 4195 	  lang, suffix rules: don't require C stuff needlessly
 4196 	  tests: expose automake bug#14560
 4197 
 4198 2013-06-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4199 
 4200 	Merge branch 'fix-pr14560' into micro
 4201 
 4202 	* fix-pr14560:
 4203 	  lang, suffix rules: don't require C stuff needlessly
 4204 	  tests: expose automake bug#14560
 4205 
 4206 2013-06-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4207 
 4208 	THANKS: update e-mall address for Ralf Corsepius
 4209 
 4210 2013-06-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4211 
 4212 	lang, suffix rules: don't require C stuff needlessly
 4213 
 4214 	This change fixes automake bug#14560: when two or more user-defined suffix
 4215 	rules were present in a single Makefile.am, automake would needlessly
 4216 	include definition of some make variables related to C compilation in the
 4217 	generated Makefile.in.
 4218 
 4219 	* automake.in (handle_languages): Fix logic to decide whether or not to
 4220 	include definitions of C compilation related variables in the generated
 4221 	Makefile.in: instead of doing so when two or more user-defined suffix
 4222 	rules are seen (which is a completely bogus criterion), do so when two
 4223 	or more compiled languages are used.
 4224 	* lib/Automake/Rule.pm (suffix_rules_count): Remove as no longer used.
 4225 	(@EXPORT): Adjust.
 4226 	* t/list-of-tests.mk (XFAIL_TESTS): No longer list the test script
 4227 	'suffix-extra-c-stuff-pr14560.sh', which now passes.
 4228 	* NEWS: Update.
 4229 
 4230 2013-06-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4231 
 4232 	tests: expose automake bug#14560
 4233 
 4234 	Automake needlessly generates definition of make variables related
 4235 	to C compilation when two or more user-defined suffix rules are
 4236 	present in a single Makefile.am.
 4237 
 4238 	* t/suffix-extra-c-stuff-pr14560.sh: New test, exposing the bug.
 4239 	This test is still xfailing.
 4240 	* t/no-extra-c-stuff.sh: New test, check for a potential related
 4241 	regression.  This regression is not actually present here (so this
 4242 	test passes), but it still took place in our first attempt at
 4243 	fixing bug#14560 -- so this test has proven to be actually useful.
 4244 	* t/no-extra-makefile-code.sh: Improve comments, and tighten the
 4245 	grepping checks a little.
 4246 	* t/list-of-tests.mk (handwritten_TESTS): Add the new tests.
 4247 	(XFAIL_TESTS): Add the new xfailing test.
 4248 
 4249 2013-06-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4250 
 4251 	Merge branch 'micro' into maint
 4252 
 4253 	* micro:
 4254 	  maint: add a missing copyright notice
 4255 	  sync: update config.guess from upstream
 4256 	  tests: expose automake bug#13928
 4257 	  comments: fix some out-of-sync refs to test scripts
 4258 	  tests: expose automake bug#13940
 4259 
 4260 2013-06-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4261 
 4262 	maint: add a missing copyright notice
 4263 
 4264 	* t/ax/deltree.pl: Here.  Issue revealed by "make update-copyright".
 4265 
 4266 2013-06-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4267 
 4268 	sync: update config.guess from upstream
 4269 
 4270 	* lib/config.guess: Here.
 4271 
 4272 2013-06-08  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4273 
 4274 	tests: expose automake bug#13928
 4275 
 4276 	* t/subobj-indir-pr13928.sh: New test, still xfailing.
 4277 	* t/subobj-vpath-pr13928.sh: Likewise.
 4278 	* t/list-of-tests.mk (XFAIL_TESTS, handwritten_TESTS): Update.
 4279 
 4280 2013-06-08  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4281 
 4282 	comments: fix some out-of-sync refs to test scripts
 4283 
 4284 	Those script has been renamed since those comments where written.
 4285 
 4286 	* lib/Automake/Rule.pm: Adjust.
 4287 	* lib/am/distdir.am: Likewise.
 4288 
 4289 2013-06-08  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4290 
 4291 	tests: expose automake bug#13940
 4292 
 4293 	* t/override-conditional-pr13940.sh: New test, still xfailing.
 4294 	* t/list-of-tests.mk (XFAIL_TESTS, handwritten_TESTS): Add it.
 4295 
 4296 2013-06-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4297 
 4298 	fixup: botched merge for 'bin/automake.in'
 4299 
 4300 	* bin/automake.in (print_autodist_files): Here, we had unwittingly
 4301 	botched up a 'format' invocation during the latest maint -> master
 4302 	merge.
 4303 
 4304 2013-06-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4305 
 4306 	Merge branch 'maint'
 4307 
 4308 	* maint:
 4309 	  configure: remove an obsolete TODO comment
 4310 	  post-release: micro version bump (1.13.3a)
 4311 	  release: stable micro release 1.13.3
 4312 	  typofix: fix grammaro in comments in t/tags-pr12372.sh
 4313 	  NEWS: minor fixlets, re-wording, and better text wrapping
 4314 	  tests: slightly stricter checks in t/cxx-demo.sh
 4315 	  tests: fix spurious failure when 'etags' program is Exuberant Ctags
 4316 	  tests: fix spurious failure due to missing sleeps
 4317 	  maint: version bump after beta release 1.13b
 4318 	  release: beta release 1.13b (will become 1.14)
 4319 	  maint: support new Automake versioning scheme in tagging/uploading rules
 4320 	  announcement: can be generated from development snapshots as well
 4321 	  announcement: can be generated from development snapshots as well
 4322 	  announcement: be less strict in the paring of NEWS
 4323 	  maint: version bump after beta release 1.13.2b
 4324 	  release: beta release 1.13.2b (will become 1.13.3)
 4325 	  sync: update config.guess from upstream
 4326 	  automake: assume we can always pass '-o' to the C compiler
 4327 	  NEWS: document recent testsuite fixes (MinGW/MSYS related)
 4328 	  depcomp: avoid trailing backslash in depfile for depmode=msvc7
 4329 	  tests: prune some weed in a non-POSIX test
 4330 	  tests: avoid a spurious failure on MSYS
 4331 	  t/README: document "run_make", discourage "make -e"
 4332 	  cosmetics: tiny whitespace fixlets
 4333 	  tests: avoid a spurious failure with MSVC
 4334 	  tests: avoid a spurious failure on non-POSIX systems
 4335 
 4336 2013-06-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4337 
 4338 	Merge branch 'micro' into maint
 4339 
 4340 	* micro:
 4341 	  configure: remove an obsolete TODO comment
 4342 	  post-release: micro version bump (1.13.3a)
 4343 	  release: stable micro release 1.13.3
 4344 	  typofix: fix grammaro in comments in t/tags-pr12372.sh
 4345 
 4346 2013-06-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4347 
 4348 	configure: remove an obsolete TODO comment
 4349 
 4350 	* configure.ac: Here.
 4351 
 4352 2013-06-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4353 
 4354 	post-release: micro version bump (1.13.3a)
 4355 
 4356 	* configure.ac: Bump version: 1.13.3 -> 1.13.3a
 4357 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 4358 
 4359 2013-06-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4360 
 4361 	release: stable micro release 1.13.3
 4362 
 4363 	* configure.ac (AC_INIT): Bump version number to 1.13.3.
 4364 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 4365 
 4366 2013-06-03  Peter Rosin  <peda@lysator.liu.se>
 4367 
 4368 	typofix: fix grammaro in comments in t/tags-pr12372.sh
 4369 
 4370 2013-06-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4371 
 4372 	Merge branch 'micro' into maint
 4373 
 4374 	* micro:
 4375 	  NEWS: minor fixlets, re-wording, and better text wrapping
 4376 	  tests: slightly stricter checks in t/cxx-demo.sh
 4377 	  tests: fix spurious failure when 'etags' program is Exuberant Ctags
 4378 	  tests: fix spurious failure due to missing sleeps
 4379 
 4380 2013-06-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4381 
 4382 	NEWS: minor fixlets, re-wording, and better text wrapping
 4383 
 4384 2013-06-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4385 
 4386 	tests: slightly stricter checks in t/cxx-demo.sh
 4387 
 4388 	This is a follow-up to the commit fixing automake bug#14493.
 4389 
 4390 	* t/cxx-demo.sh: Also check that the built program returns the
 4391 	correct (i.e., zero) exit status when run.  And improve comments
 4392 	a little while at it.
 4393 
 4394 2013-06-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4395 
 4396 	tests: fix spurious failure when 'etags' program is Exuberant Ctags
 4397 
 4398 	Fixes automake bug#14517.
 4399 
 4400 	* t/tags-pr12372.sh: If the 'etags' program in use supports the
 4401 	'--langmap' option, use it.  That is required to avoid spurious
 4402 	failures with Exuberant Ctags (at least version 5.8), which by
 4403 	default do not generate any tags for file extensions it doesn't
 4404 	recognize.
 4405 
 4406 2013-06-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4407 
 4408 	tests: fix spurious failure due to missing sleeps
 4409 
 4410 	Fixes automake bug#14528.
 4411 
 4412 	* THANKS: Give credit to that bug's original reporter.
 4413 
 4414 	* t/remake-configure-dependencies.sh: Add few missing '$sleep'
 4415 	invocations.  I thought that the sleeps implicit in the configure
 4416 	invocation were enough, but they were not, actually.  Here is what
 4417 	can happen:
 4418 
 4419 	  1. The config.status script is generated by a configure run.
 4420 	  2. ./config.status and make are run.
 4421 	  3. The 'print-version' script is modified.
 4422 	  4. Since that script is listed in $(CONFIGURE_DEPENDENCIES),
 4423 	     autoconf is re-run.
 4424 	  5. On a fast-enough machine, the three steps 2-4 above, even
 4425 	     combined, might have taken less than a second to run;
 4426 	  6. If the filesystem doesn't have a sub-second timestamp
 4427 	     resolution, that means the newly-generated configure has
 4428 	     the same timestamp of the old config.status;
 4429 	  7. So, config.status is not re-run, and the Makefiles are
 4430 	     not updated.
 4431 	  8. Spurious failure!
 4432 
 4433 	So we really need more explicit sleeps.
 4434 
 4435 2013-05-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4436 
 4437 	maint: version bump after beta release 1.13b
 4438 
 4439 	* configure.ac (AC_INIT): Bump version number to 1.13c.
 4440 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 4441 
 4442 2013-05-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4443 
 4444 	release: beta release 1.13b (will become 1.14)
 4445 
 4446 	* configure.ac (AC_INIT): Bump version number to 1.13b.
 4447 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 4448 
 4449 2013-05-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4450 
 4451 	Merge branch 'micro' into maint
 4452 
 4453 	* micro:
 4454 	  maint: support new Automake versioning scheme in tagging/uploading rules
 4455 	  announcement: can be generated from development snapshots as well
 4456 	  announcement: can be generated from development snapshots as well
 4457 	  announcement: be less strict in the paring of NEWS
 4458 	  maint: version bump after beta release 1.13.2b
 4459 	  release: beta release 1.13.2b (will become 1.13.3)
 4460 	  sync: update config.guess from upstream
 4461 
 4462 2013-05-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4463 
 4464 	maint: support new Automake versioning scheme in tagging/uploading rules
 4465 
 4466 	* maint.mk (base_version_rx: Drop.
 4467 	(rx-0, rx-1): New.
 4468 	(stable_major_version_rx, stable_minor_version_rx, beta_version_rx):
 4469 	Update.
 4470 	(stable_micro_version_rx, alpha_version_rx: New.
 4471 	(determine_release_type): Adjust.
 4472 	(web-manual-update): Likewise.
 4473 
 4474 2013-05-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4475 
 4476 	announcement: can be generated from development snapshots as well
 4477 
 4478 	And not only from checkouts corresponding exactly from a beta
 4479 	or stable release.  That was only getting in the way of proper
 4480 	testing for the 'announcement' recipe.
 4481 
 4482 	* maint.mk (determine_release_type): If the make variable
 4483 	DEVEL_SNAPSHOT is set, do not error out if the current version
 4484 	denotes a development snapshot (e.g., "1.13.2c" or "1.99a").
 4485 	(announcement): Relax, by also accepting to run from development
 4486 	snapshots, not only stable or beta releases.  Do so by defining
 4487 	the target-specific variable DEVEL_SNAPSHOT to "yes".
 4488 	(print-release-type): Micro enhancement while at it.
 4489 
 4490 2013-05-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4491 
 4492 	announcement: can be generated from development snapshots as well
 4493 
 4494 	And not only from checkouts corresponding exactly from a beta
 4495 	or stable release.  That was only getting in the way of proper
 4496 	testing for the 'announcement' recipe.
 4497 
 4498 	* maint.mk (determine_release_type): If the make variable
 4499 	DEVEL_SNAPSHOT is set, do not error out if the current version
 4500 	denotes a development snapshot (e.g., "1.13.2c" or "1.99a").
 4501 	(announcement): Relax, by also accepting to run from development
 4502 	snapshots, not only stable or beta releases.  Do so by defining
 4503 	the target-specific variable DEVEL_SNAPSHOT to "yes".
 4504 
 4505 2013-05-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4506 
 4507 	announcement: be less strict in the paring of NEWS
 4508 
 4509 	So that the file generated by "make announcement" is correct also
 4510 	for beta releases.
 4511 
 4512 	* maint.mk (announcement): Relax the awk program processing NEWS a
 4513 	little.
 4514 
 4515 2013-05-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4516 
 4517 	maint: version bump after beta release 1.13.2b
 4518 
 4519 	* configure.ac (AC_INIT): Bump version number to 1.13.2c.
 4520 	* m4/amversion.m4: Likewise (autoupdated by "make bootstrap").
 4521 
 4522 2013-05-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4523 
 4524 	release: beta release 1.13.2b (will become 1.13.3)
 4525 
 4526 	* configure.ac (AC_INIT): Bump version number to 1.13.2b.
 4527 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 4528 
 4529 2013-05-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4530 
 4531 	sync: update config.guess from upstream
 4532 
 4533 2013-05-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4534 
 4535 	Merge branch 'micro' into maint
 4536 
 4537 	* micro:
 4538 	  NEWS: document recent testsuite fixes (MinGW/MSYS related)
 4539 	  depcomp: avoid trailing backslash in depfile for depmode=msvc7
 4540 	  tests: prune some weed in a non-POSIX test
 4541 	  tests: avoid a spurious failure on MSYS
 4542 
 4543 2013-05-30  Peter Rosin  <peda@lysator.liu.se>
 4544 
 4545 	automake: assume we can always pass '-o' to the C compiler
 4546 
 4547 	It is assumed that we can pass -c -o to the C compiler, so remove
 4548 	some special casing and always do that.
 4549 
 4550 	This change is similar in spirit to v1.13.1d-217-g7299c4d "depend: assume
 4551 	we can always pass '-o' to the C compiler"
 4552 
 4553 	This change also happen to fix a testsuite failure
 4554 	(t/silent-many-languages.sh) when mixing MSVC and GNU fortran, which
 4555 	have different default object file extensions (.obj vs. .o). This
 4556 	difference in object file extension is not handled well and caused
 4557 	Automake to look for MSVC objects with .o extension. Always using -o
 4558 	makes MSVC create .o object files and linking succeeds. Not that
 4559 	anybody recommends mixing toolchains or anything.
 4560 
 4561 	* bin/automake.in (handle_languages): Remove conditional modification
 4562 	of 'output_flag' entry for 'c'.
 4563 	(register_language ('name' => 'c')): Add 'output_flag' entry set to '-o'.
 4564 
 4565 2013-05-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4566 
 4567 	NEWS: document recent testsuite fixes (MinGW/MSYS related)
 4568 
 4569 2013-05-30  Peter Rosin  <peda@lysator.liu.se>
 4570 
 4571 	depcomp: avoid trailing backslash in depfile for depmode=msvc7
 4572 
 4573 	When compiling a file without any dependencies (no #includes), the
 4574 	msvc7 depmode (and consequently msvc7msys) generates a depfile
 4575 	Makefile fragment with the last line ending with a backslash. This
 4576 	is less robust than needed.
 4577 
 4578 	Fixes automake bug#14501.
 4579 
 4580 	* lib/depcomp (msvc7): Finish off the depfile Makefile fragment with
 4581 	an empty line.
 4582 	(scriptversion): Update.
 4583 
 4584 2013-05-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4585 
 4586 	Merge branch 'fix-pr14499'
 4587 
 4588 	* fix-pr14499:
 4589 	  tests: avoid a spurious failure when TeX is not installed.
 4590 
 4591 2013-05-29  Peter Rosin  <peda@lysator.liu.se>
 4592 	    Stefano Lattarini  <stefano.lattarini@gmail.com>
 4593 
 4594 	tests: prune some weed in a non-POSIX test
 4595 
 4596 	Fixes automake bug#14494.
 4597 
 4598 	* t/distcheck-pr10470.sh: Make sure the test-driver is installed. Adjust
 4599 	to run_make changes. Adjust grep pattern to match new output.
 4600 
 4601 2013-05-29  Peter Rosin  <peda@lysator.liu.se>
 4602 
 4603 	tests: avoid a spurious failure on MSYS
 4604 
 4605 	Fixes automake bug#14493.
 4606 
 4607 	* t/cxx-demo.sh: Strip CR characters from the program output.
 4608 
 4609 2013-05-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4610 
 4611 	Merge branch 'micro' into maint
 4612 
 4613 	* micro:
 4614 	  t/README: document "run_make", discourage "make -e"
 4615 	  cosmetics: tiny whitespace fixlets
 4616 	  tests: avoid a spurious failure with MSVC
 4617 	  tests: avoid a spurious failure on non-POSIX systems
 4618 
 4619 2013-05-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4620 
 4621 	t/README: document "run_make", discourage "make -e"
 4622 
 4623 2013-05-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4624 
 4625 	cosmetics: tiny whitespace fixlets
 4626 
 4627 2013-05-29  Peter Rosin  <peda@lysator.liu.se>
 4628 
 4629 	tests: avoid a spurious failure with MSVC
 4630 
 4631 	Fixes automake bug#14498.
 4632 
 4633 	* t/yacc-bison-skeleton.sh: inline was not standardized prior to C99.
 4634 
 4635 2013-05-29  Peter Rosin  <peda@lysator.liu.se>
 4636 
 4637 	tests: avoid a spurious failure on non-POSIX systems
 4638 
 4639 	Fixes automake bug#14495.
 4640 
 4641 	* t/lex-header.sh: Make sure unistd.h isn't included.
 4642 
 4643 2013-05-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4644 
 4645 	tests: avoid a spurious failure when TeX is not installed.
 4646 
 4647 	Fixes automake bug#14499.
 4648 
 4649 	* t/txinfo-no-split.sh: If TeX is not available, redefine 'dvi' to be a
 4650 	dummy target, so that "make distcheck" doesn't fail while trying to build
 4651 	the DVI version of the Texinfo manual.
 4652 
 4653 2013-05-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4654 
 4655 	tests: avoid a spurious failure when $CC != cc
 4656 
 4657 	Fixes automake bug#14500.
 4658 
 4659 	* t/subpkg.sh: Improve the grepping of configure output.
 4660 
 4661 2013-05-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4662 
 4663 	Merge branch 'maint'
 4664 
 4665 	* maint:
 4666 	  NEWS: fix typos and grammaros
 4667 	  NEWS: document fix for bug#14441
 4668 	  Automake::Rule: consistently prepend underscore to private variables
 4669 	  Automake::Rule: rename: suffix_rule() -> next_in_suffix_chain()
 4670 	  Automake::Rule: adjust comments and POD according to previous changes
 4671 	  Automake::Rule: make private variables lexically scoped
 4672 	  suffix rules: better distinction between builtin and user-derived
 4673 	  Automake::Rule: expose suffix rules as a function, not a scalar
 4674 	  tests: expose automake bug#14441
 4675 
 4676 2013-05-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4677 
 4678 	Merge branch 'micro' into maint
 4679 
 4680 	* micro:
 4681 	  NEWS: fix typos and grammaros
 4682 	  NEWS: document fix for bug#14441
 4683 	  Automake::Rule: consistently prepend underscore to private variables
 4684 	  Automake::Rule: rename: suffix_rule() -> next_in_suffix_chain()
 4685 	  Automake::Rule: adjust comments and POD according to previous changes
 4686 	  Automake::Rule: make private variables lexically scoped
 4687 	  suffix rules: better distinction between builtin and user-derived
 4688 	  Automake::Rule: expose suffix rules as a function, not a scalar
 4689 	  tests: expose automake bug#14441
 4690 
 4691 2013-05-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4692 
 4693 	NEWS: drop an idiomatic use of latinism 'sic'
 4694 
 4695 	Reported-by: Peter Rosin <peda@lysator.liu.se>
 4696 
 4697 2013-05-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4698 
 4699 	NEWS: fix typos and grammaros
 4700 
 4701 	Reported-by: Peter Rosin <peda@lysator.liu.se>
 4702 
 4703 2013-05-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4704 
 4705 	NEWS: fix Automake versions
 4706 
 4707 	Reported-by: Peter Rosin <peda@lysator.liu.se>
 4708 
 4709 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4710 
 4711 	Merge branch 'suffix-rules-work' into micro
 4712 
 4713 	* suffix-rules-work:
 4714 	  NEWS: document fix for bug#14441
 4715 	  Automake::Rule: consistently prepend underscore to private variables
 4716 	  Automake::Rule: rename: suffix_rule() -> next_in_suffix_chain()
 4717 	  Automake::Rule: adjust comments and POD according to previous changes
 4718 	  Automake::Rule: make private variables lexically scoped
 4719 	  suffix rules: better distinction between builtin and user-derived
 4720 	  Automake::Rule: expose suffix rules as a function, not a scalar
 4721 	  tests: expose automake bug#14441
 4722 
 4723 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4724 
 4725 	Merge branch 'maint'
 4726 
 4727 	* maint:
 4728 	  NEWS: fix typo
 4729 
 4730 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4731 
 4732 	NEWS: fix typo
 4733 
 4734 	Reported-by: Peter Rosin <peda@lysator.liu.se>
 4735 
 4736 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4737 
 4738 	Merge branch 'maint'
 4739 
 4740 	* maint:
 4741 	  NEWS: report recent documentation fix about AM_PROG_MKDIR_P
 4742 	  NEWS: Automake 2.0 will assume "rm -f" without args work
 4743 	  NEWS: on assuming "rm -f" without arguments work
 4744 	  NEWS: fix a couple of typos in older entries
 4745 	  docs: AM_PROG_MKDIR_P: will not be removed in 2.0 release of Automake
 4746 	  docs: AM_PROG_CC_C_O: correct imprecise statements about it
 4747 	  maint: sanity checks in 'check-minimal-autoconf' convenience target
 4748 	  maint: test minimal supported autoconf through convenience target
 4749 	  maint: install minimal supported autoconf through convenience targets
 4750 	  test-lib: typofix in comments
 4751 
 4752 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4753 
 4754 	Merge branch 'micro' into maint
 4755 
 4756 	* micro:
 4757 	  NEWS: report recent documentation fix about AM_PROG_MKDIR_P
 4758 	  NEWS: Automake 2.0 will assume "rm -f" without args work
 4759 	  NEWS: fix a couple of typos in older entries
 4760 
 4761 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4762 
 4763 	NEWS: report recent documentation fix about AM_PROG_MKDIR_P
 4764 
 4765 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4766 
 4767 	NEWS: Automake 2.0 will assume "rm -f" without args work
 4768 
 4769 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4770 
 4771 	NEWS: on assuming "rm -f" without arguments work
 4772 
 4773 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4774 
 4775 	NEWS: fix a couple of typos in older entries
 4776 
 4777 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4778 
 4779 	Merge branch 'micro' into maint
 4780 
 4781 	* micro:
 4782 	  docs: AM_PROG_MKDIR_P: will not be removed in 2.0 release of Automake
 4783 	  maint: sanity checks in 'check-minimal-autoconf' convenience target
 4784 	  maint: test minimal supported autoconf through convenience target
 4785 	  maint: install minimal supported autoconf through convenience targets
 4786 	  test-lib: typofix in comments
 4787 
 4788 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4789 
 4790 	docs: AM_PROG_MKDIR_P: will not be removed in 2.0 release of Automake
 4791 
 4792 	See the 'PLANS/obsolete-removed/am-prog-mkdir-p.txt' file in the 'maint'
 4793 	branch (as of commit v1.13.2-201-gd99e3f3) for details.
 4794 
 4795 	* doc/automake.texi: Adjust.
 4796 
 4797 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4798 
 4799 	docs: AM_PROG_CC_C_O: correct imprecise statements about it
 4800 
 4801 	* doc/automake.texi: Here.
 4802 
 4803 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4804 
 4805 	maint: sanity checks in 'check-minimal-autoconf' convenience target
 4806 
 4807 	* maint.mk (check-minimal-autoconf): Here, check that autoconf seems
 4808 	to be locally installed, and that such local install refers to the
 4809 	correct expected minimal version.
 4810 
 4811 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4812 
 4813 	maint: test minimal supported autoconf through convenience target
 4814 
 4815 	* maint.mk (check-minimal-autoconf): New convenience target.
 4816 
 4817 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4818 
 4819 	maint: install minimal supported autoconf through convenience targets
 4820 
 4821 	To make testing of Automake in conjunction with it easier and more
 4822 	reproducible.  It might be nice to also have convenience targets to
 4823 	actually automatically run the Automake testsuite with such autoconf
 4824 	version, but that is left to later patches.
 4825 
 4826 	* configure.ac: AC_SUBST the definition of $required_autoconf_version.
 4827 	* maint.mk (WGET): Move definition earlier.
 4828 	(gnu-ftp, ac-v, ac-n, ac-p, ac-t, ac-l, ac-d): New auxiliary variables.
 4829 	(fetch-minimal-autoconf, build-minimal-autoconf): New convenience
 4830 	targets.
 4831 	* .gitignore: Update.
 4832 
 4833 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4834 
 4835 	NEWS: document fix for bug#14441
 4836 
 4837 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4838 
 4839 	Automake::Rule: consistently prepend underscore to private variables
 4840 
 4841 	* lib/Automake/Rule.pm (%suffix_rules): Rename ...
 4842 	(%_suffix_rules): ... like this.
 4843 	(%suffix_rules_builtin): Rename ...
 4844 	(%_suffix_rules_builtin): ... like this.
 4845 	(reset, next_in_suffix_chain, register_suffix_rule, suffix_rules_count):
 4846 	Adjust.
 4847 
 4848 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4849 
 4850 	Automake::Rule: rename: suffix_rule() -> next_in_suffix_chain()
 4851 
 4852 	* lib/Automake/Rule.pm (suffix_rule): Rename ...
 4853 	(next_in_suffix_chain): ... like this.
 4854 	(%suffix_rules): Adjust comments.
 4855 	(@EXPORT): Adjust.
 4856 	* automake.in (derive_suffix): Likewise.
 4857 
 4858 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4859 
 4860 	Automake::Rule: adjust comments and POD according to previous changes
 4861 
 4862 	* lib/Automake/Rule.pm: Here, in several places.
 4863 	* automake.in (register_language): And a tiny adjustment here as well.
 4864 
 4865 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4866 
 4867 	Automake::Rule: make private variables lexically scoped
 4868 
 4869 	* lib/Automake/Rule.pm (@_known_extensions_list): This one.
 4870 	(@_suffixes): And this one.
 4871 	(%_rule_dict): And this one.
 4872 
 4873 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4874 
 4875 	suffix rules: better distinction between builtin and user-derived
 4876 
 4877 	Fixes automake bug#14441.
 4878 
 4879 	* lib/Automake/Rule.pm ($_suffix_rules_default): Remove, superseded by ...
 4880 	(%suffix_rules_builtin): ... this lexical variable.
 4881 	($suffix_rules): Remove, superseded by ...
 4882 	(%suffix_rules): ... this lexical variable.
 4883 	(suffix_rules, suffix_rules_count): Adjust.
 4884 	(register_suffix_rule): Update '%suffix_rules_builtin' rather than
 4885 	'%suffix_rules' if the location (as passed by the '$where' argument)
 4886 	is an "internal" one (doesn't come from user-provided Makefile.am).
 4887 	(reset): Simplify resetting of '%suffix_rules' to the default ones
 4888 	accordingly.
 4889 	* t/list-of-tests.mk (XFAIL_TESTS): Drop test 'suffix-custom-pr14441.sh'.
 4890 
 4891 2013-05-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4892 
 4893 	Automake::Rule: expose suffix rules as a function, not a scalar
 4894 
 4895 	This is just a preparatory refactoring in view of future patches.
 4896 	No semantic change is intended.
 4897 
 4898 	* lib/Automake/Rule.pm ($suffix_rules): Turn from a package-level
 4899 	variable to a lexical variable.
 4900 	(suffix_rule): New function, expose the details of $suffix_rules
 4901 	that are actually required by code outside thus modules --- and
 4902 	only those details, no more.
 4903 	(@EXPORT): Adjust.
 4904 	* automake.in (derive_suffix): Likewise.
 4905 
 4906 2013-05-27  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4907 
 4908 	tests: expose automake bug#14441
 4909 
 4910 	* t/suffix-custom-pr14441.sh: New test, still failing.
 4911 	* t/list-of-tests.mk (handwritten_TESTS, XFAIL_TESTS): Add it.
 4912 
 4913 	Helped-by: Felix Salfelder <felix@salfelder.org>
 4914 
 4915 2013-05-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4916 
 4917 	Merge branch 'maint'
 4918 
 4919 	* maint:
 4920 	  tests: remove an "unworthy" FIXME comment
 4921 	  tests: remove one obsolete workaround for long-lifted limitations
 4922 	  tests: couple of minor tweaks in demo tests in C++ and Libtool
 4923 
 4924 2013-05-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4925 
 4926 	test-lib: typofix in comments
 4927 
 4928 	* t/ax/am-test-lib.sh: In here.
 4929 
 4930 2013-05-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4931 
 4932 	Merge branch 'micro' into maint
 4933 
 4934 	* micro:
 4935 	  tests: remove an "unworthy" FIXME comment
 4936 	  tests: remove one obsolete workaround for long-lifted limitations
 4937 	  tests: couple of minor tweaks in demo tests in C++ and Libtool
 4938 
 4939 2013-05-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4940 
 4941 	tests: remove an "unworthy" FIXME comment
 4942 
 4943 	* t/location.sh: Here.
 4944 
 4945 2013-05-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4946 
 4947 	tests: remove one obsolete workaround for long-lifted limitations
 4948 
 4949 	* t/silent-many-languages.sh: Here.
 4950 
 4951 2013-05-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4952 
 4953 	tests: couple of minor tweaks in demo tests in C++ and Libtool
 4954 
 4955 	* t/cxx-lt-demo.sh: Here.
 4956 
 4957 2013-05-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4958 
 4959 	Merge branch 'maint'
 4960 
 4961 	* maint:
 4962 	  NEWS: document deprecation of 'shar' and 'compress' dist formats
 4963 	  m4: rename minuso.m4 -> prog-cc-c-o.m4
 4964 	  tests: adjust or remove some obsolete comments
 4965 	  tests: remove some useless 'unset' of variables
 4966 	  tests: avoid a couple of extra sleep with GNU make
 4967 	  NEWS: document testsuite work for 1.13.3
 4968 	  lint: remove a couple of obsolete syntax checks
 4969 	  lint: cosmetics: use #-comments, not ##-comments
 4970 	  lint: cosmetics: some reordering
 4971 	  lint: recipes of syntax check require GNU grep; ensure it is used
 4972 	  lint: better name for a syntax check
 4973 	  tests: rename $am_make_rc_got -> $am_make_rc
 4974 	  tests: some tests make no sense if "$CC -c -o" doesn't work
 4975 	  AM_PROG_CC_C_O: don't rely on AC_PROG_CC_C_O, re-implement similar logic
 4976 	  compile: rewrite AC_PROG_CC with AM_PROG_CC_C_O contents
 4977 	  tests: ensure $required is not set too late
 4978 
 4979 2013-05-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4980 
 4981 	Merge branch 'micro' into maint
 4982 
 4983 	* micro:
 4984 	  tests: adjust or remove some obsolete comments
 4985 	  tests: remove some useless 'unset' of variables
 4986 
 4987 2013-05-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4988 
 4989 	NEWS: document deprecation of 'shar' and 'compress' dist formats
 4990 
 4991 2013-05-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 4992 
 4993 	Merge branch 'prog-cc-c-o-work' into maint
 4994 
 4995 	* prog-cc-c-o-work:
 4996 	  tests: some tests make no sense if "$CC -c -o" doesn't work
 4997 	  AM_PROG_CC_C_O: don't rely on AC_PROG_CC_C_O, re-implement similar logic
 4998 	  compile: rewrite AC_PROG_CC with AM_PROG_CC_C_O contents
 4999 
 5000 2013-05-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5001 
 5002 	m4: rename minuso.m4 -> prog-cc-c-o.m4
 5003 
 5004 	The new name is much clearer.
 5005 
 5006 	* m4/minuso.m4: Rename ...
 5007 	* m4/prog-cc-c-o.m4: ... like this.
 5008 	* m4/Makefile.in (dist_automake_DATA): Adjust.
 5009 
 5010 2013-05-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5011 
 5012 	tests: adjust or remove some obsolete comments
 5013 
 5014 	* t/ax/am-test-lib.sh: Here.
 5015 	* t/distcheck-configure-flags.sh: And here.
 5016 
 5017 2013-05-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5018 
 5019 	tests: remove some useless 'unset' of variables
 5020 
 5021 	They are redundant, now that we no longer call "make -e".
 5022 
 5023 	* t/javaflags.sh: Don't unset JAVAC.
 5024 	* t/lflags.sh: Don't unset LEX.
 5025 	* t/lflags2.sh: Likewise.
 5026 
 5027 2013-05-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5028 
 5029 	Merge branch 'micro' into maint
 5030 
 5031 	* micro:
 5032 	  tests: avoid a couple of extra sleep with GNU make
 5033 	  NEWS: document testsuite work for 1.13.3
 5034 	  lint: remove a couple of obsolete syntax checks
 5035 	  lint: cosmetics: use #-comments, not ##-comments
 5036 	  lint: cosmetics: some reordering
 5037 	  lint: recipes of syntax check require GNU grep; ensure it is used
 5038 	  lint: better name for a syntax check
 5039 	  tests: rename $am_make_rc_got -> $am_make_rc
 5040 	  tests: ensure $required is not set too late
 5041 
 5042 2013-05-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5043 
 5044 	tests: avoid a couple of extra sleep with GNU make
 5045 
 5046 	They are only required by BSD make in order to avoid spurious failures.
 5047 
 5048 	* t/deleted-am.sh: Here.
 5049 	* t/parallel-tests-recheck-pr11791.sh: And here.
 5050 
 5051 2013-05-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5052 
 5053 	NEWS: document testsuite work for 1.13.3
 5054 
 5055 2013-05-25  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5056 
 5057 	lint: remove a couple of obsolete syntax checks
 5058 
 5059 	* syntax-checks.mk (sc_obsolete_requirements_rules): Remove
 5060 	definition of this variable.
 5061 	(modern-requirement.texi2dvi-o): Likewise.
 5062 	(modern-requirement.makeinfo-html): Likewise.
 5063 	($(sc_obsolete_requirements_rules)): Remove these obsolete
 5064 	syntax checks.
 5065 
 5066 2013-05-25  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5067 
 5068 	lint: cosmetics: use #-comments, not ##-comments
 5069 
 5070 	* syntax-checks.mk: Here.  This makes sense because this file is
 5071 	not meant to be processed by Automake, so the ##-comments are not
 5072 	treated specially.
 5073 
 5074 2013-05-25  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5075 
 5076 	lint: cosmetics: some reordering
 5077 
 5078 	* syntax-checks.mk: Here.
 5079 
 5080 2013-05-25  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5081 
 5082 	lint: recipes of syntax check require GNU grep; ensure it is used
 5083 
 5084 	* syntax-checks.mk (sc_sanity_gnu_grep): New.
 5085 	($(syntax_check_rules)): Depend on it.
 5086 
 5087 2013-05-25  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5088 
 5089 	lint: better name for a syntax check
 5090 
 5091 	* syntax-checks.mk (sc_tests_make_without_am_makeflags): Rename ...
 5092 	(sc_make_without_am_makeflags): ... like this.
 5093 	(syntax_check_rules): Adjust.
 5094 
 5095 2013-05-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5096 
 5097 	tests: rename $am_make_rc_got -> $am_make_rc
 5098 
 5099 	The latter is more natural  More importantly, I keep misspelling
 5100 	the former over and over.
 5101 
 5102 	* t/ax/am-test-lib.sh (run_make): Update.
 5103 	* t/ax/tap-summary-aux.sh: Adjust.
 5104 	* t/ax/testsuite-summary-checks.sh: Likewise.
 5105 	* t/parallel-tests-exit-statuses.sh: Likewise.
 5106 	* t/parallel-tests-extra-programs.sh: Likewise.
 5107 	* t/parallel-tests-fd-redirect-exeext.sh: Likewise.
 5108 	* t/parallel-tests-fd-redirect.sh: Likewise.
 5109 	* t/parallel-tests-fork-bomb.sh: Likewise.
 5110 	* t/parallel-tests-no-spurious-summary.sh: Likewise.
 5111 	* t/parallel-tests-recheck-pr11791.sh: Likewise.
 5112 	* t/parallel-tests-reset-term.sh: Likewise.
 5113 	* t/tap-signal.tap: Likewise.
 5114 	* t/test-driver-acsubst.sh: Likewise.
 5115 	* t/test-driver-cond.sh: Likewise.
 5116 
 5117 2013-05-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5118 
 5119 	tests: some tests make no sense if "$CC -c -o" doesn't work
 5120 
 5121 	So just skip them, to avoid spurious failures when running
 5122 	"make check-no-cc-c-o".
 5123 
 5124 	* t/ax/am-test-lib.sh (require_tool): New requirement '-c-o'.
 5125 	* t/subobj10.sh ($required): Add it.
 5126 	* gen-testsuite-part (%depmodes): Adjust so that tests that
 5127 	use 'makedepend' will be skipped if the compiler is being
 5128 	forced not to grasp "-c -o".
 5129 
 5130 2013-05-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5131 
 5132 	AM_PROG_CC_C_O: don't rely on AC_PROG_CC_C_O, re-implement similar logic
 5133 
 5134 	** Theoretical problems of AC_PROG_CC_C_O:
 5135 
 5136 	  Both cc and $CC are checked to see if they support the '-c' and '-o'
 5137 	  options together.
 5138 	  This behaviour is highly inconsistent with that of the other macros
 5139 	  related to C compiler checks -- which test only $CC.
 5140 	  It can also cause unwarranted uses of the 'compile' script on systems
 5141 	  where the default 'cc' is inferior, but the user is compiling with a
 5142 	  proper, different compiler (e.g., gcc).
 5143 
 5144 	** Practical problems with our previous implementation of C support m4
 5145 	   macros in Automake:
 5146 
 5147 	  - AM_PROG_AR must now be called *before* AC_PROG_CC; this wasn't the
 5148 	    case before, and it turns out there are packages in the wild that
 5149 	    relied on the old behaviour.
 5150 
 5151 	  - The cross-referenced requirements and macro rewrites juggled among
 5152 	    AC_PROG_CC, AC_PROG_CC_C_O and AM_PROG_CC_C_O caused warnings in
 5153 	    autoconf; for example, in our test 't/libobj3.sh', we could see
 5154 	    warnings like these (here slightly tweaked for legibility):
 5155 
 5156 	        configure.ac:5: AC_REQUIRE: `AC_PROG_CC' expanded before required
 5157 	        autoconf/c.m4:567: AC_PROG_CC_C_O is expanded from...
 5158 	        autoconf/c.m4:429: AC_LANG_COMPILER(C) is expanded from...
 5159 	        autoconf/lang.m4:329: AC_LANG_COMPILER_REQUIRE is expanded from...
 5160 	        autoconf/general.m4:2606: AC_COMPILE_IFELSE is expanded from...
 5161 	        m4sugar/m4sh.m4:639: AS_IF is expanded from...
 5162 	        autoconf/general.m4:2031: AC_CACHE_VAL is expanded from...
 5163 	        autoconf/general.m4:2052: AC_CACHE_CHECK is expanded from...
 5164 	        aclocal.m4:70: AM_PROG_AR is expanded from...
 5165 	        configure.ac:5: the top level
 5166 
 5167 	** Fix all of that:
 5168 
 5169 	We fix all of the described issues with a new internal m4 macro
 5170 	_AM_PROG_CC_C_O (inspired to, but not based on, AC_PROG_CC_C_O) that
 5171 	gets tacked on to AC_PROG_CC automatically (this is done in the
 5172 	Automake-generated aclocal.m4) and that takes care of checking and
 5173 	adjusting '$CC' for "-c -o" support.
 5174 
 5175 	The macro AM_PROG_CC_C_O is still present, but is now just a thin
 5176 	wrapper around such Automake-enhanced AC_PROG_CC.
 5177 
 5178 	It is worth noting that the present patch causes three slight
 5179 	*backward-incompatibilities*:
 5180 
 5181 	  1. The name cache variable used by AM_PROG_CC_C_O is no longer
 5182 	     computed (at configure runtime!) from the content of '$CC',
 5183 	     but is statically defined as 'am_cv_prog_cc_c_o'.
 5184 
 5185 	  2. 'cc' is no longer checked by AM_PROG_CC_C_O, only '$CC' is.
 5186 
 5187 	  3. AM_PROG_CC_C_O no longer AC_DEFINE the C preprocessor symbol
 5188 	     'NO_MINUS_C_MINUS_O'.
 5189 
 5190 	Given however that the third change can easily be worked around, that
 5191 	the first two changes can be legitimately seen as bug fixes, and that
 5192 	the new semantics introduced by such changes will simplify the transition
 5193 	to Automake 2.0 (when the 'subdir-objects' will always be enabled
 5194 	unconditionally), we believe they are acceptable to be shipped with
 5195 	Automake 1.14.
 5196 
 5197 	With this patch, we also revert some of the testsuite adjustments done
 5198 	in previous commit v1.13.2-178-g9877109 of 2013-05-24 (compile: rewrite
 5199 	AC_PROG_CC with AM_PROG_CC_C_O contents).  Such adjustments are no longer
 5200 	needed.
 5201 
 5202 	* m4/minuso.m4 (_AM_PROG_CC_C_O): New internal macro, basically and
 5203 	adjusted version of a merge between Autoconf-provided AC_PROG_CC_C_O
 5204 	and our old implementation of AM_PROG_CC_C_O.
 5205 	(AM_PROG_CC_C_O): Redefine as a simple wrapper around AC_PROG_CC.
 5206 	* m4/init.m4 (AC_PROG_CC): Append _AM_PROG_CC_C_O, not AM_PROG_CC_C_O,
 5207 	to the pre-existing expansion of this macro.
 5208 	* m4/ar-lib.m4 (AM_PROG_AR): No longer require it to be expanded after
 5209 	AC_PROG_CC.
 5210 	* t/aclocal-deps.sh: Move AC_PROG_CC invocation after AC_PROG_RANLIB
 5211 	and AM_PROG_AR invocations.  Things should work this way too (as they
 5212 	used to).
 5213 	* t/subobj-clean-lt-pr10697.sh: Likewise.
 5214 	* t/alloca.sh: Move AC_PROG_CC invocation after AM_PROG_AR invocation.
 5215 	* t/condlib.sh: Likewise.
 5216 	* t/aclocal-deps.sh: Move AC_PROG_CC invocation after LT_INIT and
 5217 	AM_PROG_AR invocations.  Make autoconf and autoheader warnings fatal.
 5218 	* t/am-prog-cc-c-o.sh: Adjust to the new semantics, enhance a  little,
 5219 	and reduce code duplication.
 5220 	* t/ccnoco.sh: Make autoconf warnings fatal.
 5221 	* t/subpkg.sh: Likewise.
 5222 	* t/ccnoco-lib.sh: Likewise, and fix a comment.
 5223 	* t/link_cond.sh: Enhance a couple of error messages.
 5224 	* configure.ac: Drop "nullification" of AM_PROG_CC_C_O.
 5225 	* NEWS: Adjust.
 5226 
 5227 2013-05-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5228 
 5229 	compile: rewrite AC_PROG_CC with AM_PROG_CC_C_O contents
 5230 
 5231 	This is a much simpler rewrite than the one we attempted in the past,
 5232 	and that was later removed by commit 'v1.13.1d-137-g32eb770' of
 5233 	2013-05-11 (compile: avoid AC_PROG_CC messy rewrite).
 5234 
 5235 	Not only this change simplifies the code a little, but has the welcome
 5236 	collateral effect of making automatic dependency tracking work better
 5237 	with compilers that doesn't grasp the '-c' and '-o' options together.
 5238 	Issues in that setup have been caught by several failures in the target
 5239 	'check-no-cc-c-o'.
 5240 
 5241 	Unfortunately, this change has less welcome collateral effects:
 5242 
 5243 	  1. AM_PROG_AR must now be called *after* AC_PROG_CC;
 5244 
 5245 	  2. Autoconf emits extra warnings when used with Automake-generated
 5246 	     aclocal.m4.
 5247 
 5248 	These are unacceptable regressions for a release, but since we are
 5249 	going to fix them soon enough in a follow-up patch (surely to be
 5250 	applied before Automake 1.14 is released) we don't worry too much.
 5251 
 5252 	* m4/init.m4: Redefine AC_PROG_CC early, to automatically invoke
 5253 	AM_PROG_CC_C_O as well.  Accordingly, drop now-unneeded "automagical"
 5254 	AM_PROG_CC_C_O expansion at later time (which took place thanks to
 5255 	a AC_CONFIG_COMMANDS_PRE call).
 5256 	* m4/minuso.m4 (AM_PROG_CC_C_O): Ensure the expansion of the body
 5257 	of this macro takes place with C as "current Autoconf language" (use
 5258 	AC_LANG_PUSH/AC_LANG_POP).
 5259 	* m4/ar-lib.m4 (AM_PROG_AR): Likewise.  Also, require this macro to
 5260 	be expanded *after* AC_PROG_CC (so that any rewrite of $CC, if required,
 5261 	has already taken place).
 5262 	* t/add-missing.tap: Adjust to avoid spurious failures.
 5263 	* t/aclocal-deps.sh: Likewise, by having AM_PROG_AR called *after*
 5264 	AC_PROG_CC.
 5265 	* t/subobj-clean-lt-pr10697.sh: Likewise.
 5266 	* t/alloca.sh: Likewise.
 5267 	* t/condlib.sh: Likewise.
 5268 	* t/discover.sh: Likewise.
 5269 	* t/objc-megademo.sh: Likewise.
 5270 	* t/ccnoco.sh: Extend a little.
 5271 	* t/ccnoco-deps.sh: New test.
 5272 	* t/ccnoco-lib.sh: Likewise.
 5273 	* t/ccnoco-lt.sh: Likewise.
 5274 	* t/list-of-tests.mk: Add them.
 5275 
 5276 2013-05-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5277 
 5278 	tests: ensure $required is not set too late
 5279 
 5280 	And do so in a safer way, with a runtime check rather than a brittle
 5281 	static maintainer check.
 5282 
 5283 	* t/ax/test-init.sh: Set the 'required' variable to readonly.
 5284 	* syntax-checks.mk (sc_tests_required_after_defs): Remove.
 5285 	(syntax_check_rules): No longer list it.
 5286 
 5287 2013-05-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5288 
 5289 	Merge branch 'maint'
 5290 
 5291 	* maint:
 5292 	  tests: use append mode to capture parallel make output
 5293 	  tests: new convenience target 'check-parallel'
 5294 
 5295 2013-05-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5296 
 5297 	Merge branch 'micro' into maint
 5298 
 5299 	* micro:
 5300 	  tests: use append mode to capture parallel make output
 5301 	  tests: new convenience target 'check-parallel'
 5302 
 5303 2013-05-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5304 
 5305 	tests: use append mode to capture parallel make output
 5306 
 5307 	This will allow us to run the Automake testsuite forcing all the make
 5308 	invocations in the test cases to run in parallel mode [1], but without
 5309 	hitting the spurious failure reported in automake bug#11413.
 5310 
 5311 	See also:
 5312 	<http://lists.gnu.org/archive/html/bug-make/2013-05/msg00135.html>
 5313 
 5314 	The make invocations in the test cases can be forced to run in parallel
 5315 	mode by, e.g., invoking "make check" like this:
 5316 
 5317 	      make check AM_TESTSUITE_MAKE="make -j16"
 5318 
 5319 	The possible spurious failures hinted above are due to the fact that
 5320 	the redirected output of parallel make can racily lose lines.  For
 5321 	example, if GNU make (3.82) is run with -j10 on a Makefile like this:
 5322 
 5323 	    all = 0 1 2 3 4 5 6 7 8 9
 5324 	    default: $(all)
 5325 	    $(all):
 5326 	            @sleep 0.$$(($RANDOM % 10)); echo $@
 5327 
 5328 	and has its standard output redirected in overwrite-mode to a regular
 5329 	file, it looses a line of that output every 15 runs or so on a Fedora
 5330 	17 system with 64 ppc64 cores.  Redirection in append mode does not
 5331 	suffer of this problem, as explained here:
 5332 	<http://lists.gnu.org/archive/html/automake-patches/2009-03/msg00073.html>
 5333 
 5334 	See also previous commits v1.12-63-g45c1fcd of 2012-05-05 (tests: use
 5335 	append mode to capture parallel make output) and Release-1-10-280-g6426999
 5336 	of 2009-03-10 (Use append mode to capture parallel test output).
 5337 
 5338 	* t/ax/am-test-lib.sh (run_make): Adjust and enhance.
 5339 	* t/lisp8.sh: Take advantage of the enhancement, nd stop doing output
 5340 	redirection for $MAKE by hand.
 5341 	* t/tap-more.sh: Likewise.
 5342 	* t/parallel-tests-concurrency.sh: Likewise.
 5343 
 5344 2013-05-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5345 
 5346 	tests: new convenience target 'check-parallel'
 5347 
 5348 	To force the test scripts to invoke make in parallel mode.  This should
 5349 	enhance coverage of use cases (make concurrency) that are becoming more
 5350 	and more important with today multicore fast machines.
 5351 
 5352 	* Makefile.am (check-parallel): New .PHONY target.
 5353 
 5354 2013-05-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5355 
 5356 	Merge branch 'maint'
 5357 
 5358 	* maint:
 5359 	  tests: avoid few lingering $MAKE redirections
 5360 	  tests: avoid '$MAKE' redirections, use 'run_make' instead
 5361 	  tests: avoid use of redirected 'run_make' invocations
 5362 	  lint: warn against redirected 'run_make' invocations
 5363 	  comments: next GNU make release 4.0, not 3.83
 5364 	  tests: fix a potential spurious failure due to global config.site
 5365 	  HACKING: it's OK to do testsuite refactoring in a micro version
 5366 
 5367 2013-05-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5368 
 5369 	tests: avoid few lingering $MAKE redirections
 5370 
 5371 	These were present in the 'maint' branch, but not in the 'micro' branch.
 5372 	Their occurrences has been found by the 'sc_tests_no_run_make_redirect'
 5373 	maintainer check.
 5374 
 5375 	* t/fort2.sh: Adjust.
 5376 	* t/preproc-demo.sh: Likewise.
 5377 
 5378 2013-05-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5379 
 5380 	Merge branch 'micro' into maint
 5381 
 5382 	* micro:
 5383 	  tests: avoid '$MAKE' redirections, use 'run_make' instead
 5384 	  tests: avoid use of redirected 'run_make' invocations
 5385 	  lint: warn against redirected 'run_make' invocations
 5386 	  comments: next GNU make release 4.0, not 3.83
 5387 	  tests: fix a potential spurious failure due to global config.site
 5388 	  HACKING: it's OK to do testsuite refactoring in a micro version
 5389 
 5390 2013-05-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5391 
 5392 	tests: avoid '$MAKE' redirections, use 'run_make' instead
 5393 
 5394 	The use 'run_make' with  the -E, -O and -M option, it is more
 5395 	idiomatic now.  Also, this way, centralized fixes and improvements
 5396 	done in 'run_make' will automatically propagate through most of
 5397 	the testsuite.
 5398 
 5399 	* syntax-checks.mk (sc_tests_no_run_make_redirect): Also check against
 5400 	'$MAKE' invocations that uses output redirections (and not only against
 5401 	'run_make' invocation that do so).
 5402 	* Several tests: Adjust (and few minor cosmetic changes as well, while
 5403 	at it).
 5404 
 5405 2013-05-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5406 
 5407 	tests: avoid use of redirected 'run_make' invocations
 5408 
 5409 	Instead, properly use the -E, -O, or -M options of run_make.
 5410 	The occurrences of the 'run_make' invocations using output
 5411 	redirection has been found by the recently-introduced maintainer
 5412 	check 'sc_tests_no_run_make_redirect'.
 5413 
 5414 	* t/ax/tap-summary-aux.sh: Adjust.
 5415 	* t/ax/testsuite-summary-checks.sh: Likewise.
 5416 	* Several other tests: Likewise.
 5417 
 5418 2013-05-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5419 
 5420 	lint: warn against redirected 'run_make' invocations
 5421 
 5422 	The testsuite shell function 'run_make()' has the three options
 5423 	-E, -O and -M to perform the various kind of output redirections,
 5424 	so the test scripts should use them instead of redirecting the
 5425 	output "by hand":
 5426 
 5427 	    run_make ARGS >stdout           =>  run_make -O ARGS
 5428 	    run_make ARGS >stderr           =>  run_make -E ARGS
 5429 	    run_make ARGS >stdout 2>stderr  =>  run_make -O -E ARGS
 5430 	    run_make ARGS >output 2>&1      =>  run_make -M ARGS
 5431 
 5432 	See the commit message of v1.13.2-13-gc7cfa09, 2013-05-22
 5433 	(tests: run_make: options to do command redirection) for an
 5434 	explanation of why this is a good idea.
 5435 
 5436 	It's worth noting that the present patch only introduced a new
 5437 	static maintainer check looking against the undesired idioms,
 5438 	but doesn't yet remove usages of such idioms from the testsuite.
 5439 	That will be done by follow-up patch(es).
 5440 
 5441 	* syntax-checks.mk (sc_tests_no_run_make_redirect): New check.
 5442 	(syntax_check_rules): List it.
 5443 
 5444 2013-05-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5445 
 5446 	comments: next GNU make release 4.0, not 3.83
 5447 
 5448 	See: <http://lists.gnu.org/archive/html/bug-make/2013-05/msg00093.html>
 5449 
 5450 	* lib/am/header-vars.am (am__make_running_with_option): Adjust
 5451 	comments.
 5452 
 5453 2013-05-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5454 
 5455 	tests: fix a potential spurious failure due to global config.site
 5456 
 5457 	* t/instdir-no-empty.sh: Here.
 5458 
 5459 2013-05-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5460 
 5461 	HACKING: it's OK to do testsuite refactoring in a micro version
 5462 
 5463 	Reported-by: Peter Rosin <peda@lysator.liu.se>
 5464 
 5465 2013-05-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5466 
 5467 	Merge branch 'maint'
 5468 
 5469 	* maint:
 5470 	  tests: remove some code duplication
 5471 	  cosmetics: remove a couple of extra whitespaces in tests
 5472 	  tests: two minor fixups
 5473 
 5474 2013-05-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5475 
 5476 	Merge branch 'micro' into maint
 5477 
 5478 	* micro:
 5479 	  tests: remove some code duplication
 5480 	  cosmetics: remove a couple of extra whitespaces in tests
 5481 	  tests: two minor fixups
 5482 
 5483 2013-05-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5484 
 5485 	tests: remove some code duplication
 5486 
 5487 	* t/ax/am-test-lib (null_install): New function.
 5488 	* t/instdir-java.sh: Use it instead of copied & pasted code.
 5489 	* t/instdir-lisp.sh: Likewise.
 5490 	* t/instdir-ltlib.sh: Likewise.
 5491 	* t/instdir-prog.sh: Likewise.
 5492 	* t/instdir-python.sh: Likewise.
 5493 	* t/instdir-texi.sh: Likewise.
 5494 	* t/instdir.sh: Likewise.
 5495 	* t/instdir2.sh: Likewise.
 5496 
 5497 2013-05-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5498 
 5499 	cosmetics: remove a couple of extra whitespaces in tests
 5500 
 5501 	* t/instdir.sh: Here.
 5502 	* t/instdir2.sh: And here.
 5503 
 5504 2013-05-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5505 
 5506 	tests: two minor fixups
 5507 
 5508 	* t/make-keepgoing.tap: Here.
 5509 	* t/parallel-tests-exit-statuses.sh: And here.
 5510 
 5511 2013-05-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5512 
 5513 	Merge branch 'maint'
 5514 
 5515 	* maint:
 5516 	  tests: run_make: options to do command redirection
 5517 	  tests: only activate 'unset' alias if required
 5518 	  tests: better idiom to override make macro defs on the cmdline
 5519 	  test-lib: minor style changes
 5520 	  test-lib: fix botched function name in an error message
 5521 
 5522 2013-05-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5523 
 5524 	Merge branch 'micro' into maint
 5525 
 5526 	* micro:
 5527 	  tests: run_make: options to do command redirection
 5528 	  tests: only activate 'unset' alias if required
 5529 	  tests: better idiom to override make macro defs on the cmdline
 5530 	  test-lib: minor style changes
 5531 	  test-lib: fix botched function name in an error message
 5532 
 5533 2013-05-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5534 
 5535 	tests: run_make: options to do command redirection
 5536 
 5537 	Let's improve the API of the 'run_make()' helper shell function by
 5538 	adding three new options:
 5539 
 5540 	  -O   Save the standard output from make on disk, in a regular file
 5541 	       named 'stdout'.
 5542 
 5543 	  -E   Save the standard error from make on disk, in a regular file
 5544 	       named 'stderr'.
 5545 
 5546 	  -M   Save both the standard output and standard error from make on
 5547 	       disk, in a regular file named 'output'. This option supersedes
 5548 	       both the '-O' and '-E' options.
 5549 
 5550 	This new API has two main advantages.
 5551 
 5552 	  1. Its use will allow us to get rid of more cumbersome idioms
 5553 	     like, e.g.,
 5554 
 5555 	       $MAKE check >stdout && { cat stdout; exit 1; }
 5556 	       cat stdout
 5557 
 5558 	     That can now be substituted with a simpler one:
 5559 
 5560 	       run_make -e FAIL -O check
 5561 
 5562 	  2. More importantly, using the new API we will prevent any extra output
 5563 	     from the shell traces of the code in run_make to be redirected along
 5564 	     with the make stderr (where that was redirected).  This problem was
 5565 	     present in usages like, e.g.,
 5566 
 5567 	       run_make TESTS=foo.test check 2>stderr && exit 1
 5568 	       grep 'expected error message' stderr
 5569 
 5570 	     Such usages are now to be rewritten as follows:
 5571 
 5572 	       run_make -e FAIL -E TESTS=foo.test check
 5573 	       grep 'expected error message' stderr
 5574 
 5575 	     ensuring that 'stderr' won't end up containing unrelated stuff.
 5576 
 5577 	Note that we do not convert in bulk the old idioms and the use of
 5578 	redirected 'run_make' invocations with this patch.  We only convert
 5579 	some occurrences, to ensure that the new implementation of 'run_make'
 5580 	is sound enough.  More sweeping conversions will likely be done in
 5581 	follow-up patches.
 5582 
 5583 	* t/ax/am-test-lib.sh (run_make): Enhance and implement the extended API.
 5584 	* t/tap-xfail-tests.sh: Use the new 'run_make' API.
 5585 	* t/test-driver-cond.sh: Likewise.
 5586 	* t/tests-environment-fd-redirect.sh: Likewise.
 5587 	* t/uninstall-fail.sh: Likewise.
 5588 	* t/yacc-dist-nobuild.sh: Likewise.
 5589 
 5590 2013-05-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5591 
 5592 	tests: only activate 'unset' alias if required
 5593 
 5594 	This makes the test logs easier to read for most shells (which do not
 5595 	actually require that alias).  This is especially important now that
 5596 	'unset' is used in the new 'run_make()' function, and that function is
 5597 	likely going to be used more and more in the future.
 5598 
 5599 	* t/ax/test-lib.sh (_am_unset, unset): Only define this function and
 5600 	alias if "unset VAR" returns a non-zero exit status when VAR is already
 5601 	unset.
 5602 
 5603 2013-05-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5604 
 5605 	tests: better idiom to override make macro defs on the cmdline
 5606 
 5607 	We now do so with a new wrapper shell function 'run_make()', which
 5608 	relies on the use of AM_MAKEFLAGS, avoiding the use of the '-e' make
 5609 	option.  The use of that option (that causes the environment variables
 5610 	to take precedence over the macro definitions in the Makefile) has
 5611 	proved to be quite brittle in the past, causing annoying and sometimes
 5612 	problematic spurious failures.  This has worsened in some still
 5613 	unpublished work going on in development branches.  It's time to solve
 5614 	the issue once and for all.
 5615 
 5616 	Note that we do not convert all uses of $MAKE in the testsuite right
 5617 	away; we might do so in follow-up changes, with leisure, to avoid a
 5618 	"patch bomb" effect (this commit is already too much of a bomb itself).
 5619 	What we do in this commit is to get rid of all "$MAKE -e" invocations.
 5620 
 5621 	We admit that the implementation and feature-set of 'run_make()' are
 5622 	far from perfect, but good enough for our current purposes.  We'll
 5623 	improve 'run_make()' if and when the need arises.
 5624 
 5625 	* syntax-checks.mk (sc_tests_no_make_e): New syntax check, guard against
 5626 	the use of "$MAKE -e".
 5627 	(syntax_check_rules): Add it.
 5628 	(sc_tests_overriding_macros_on_cmdline): Adjust.
 5629 	(lint): New, alias for 'maintainer-check', for lazy typists.  Idea
 5630 	backported from the 'maint' branch (Automake 1.13a).
 5631 	* t/ax/am-test-lib.sh (run_make): New function.  Run $MAKE with the
 5632 	given command-line arguments, handling command-line override of variable
 5633 	definitions in a smart way (using AM_MAKEFLAGS if a non-GNU make
 5634 	implementation is detected to be in use).
 5635 	(useless_vpath_rebuild): Adjust to use 'run_make', to avoid a spurious
 5636 	maintainer check failure.
 5637 	(yl_distcheck): Use 'run_make' rather than bare '$MAKE'.
 5638 	(single_quote, append_single_quoted, is_valid_varname): New auxiliary
 5639 	function, used, directly or indirectly, by it.
 5640 	* Many tests: Adjust to avoid the use of "$MAKE -e", and prefer the
 5641 	use of 'run_make' in few other contexts as well, where it makes sense.
 5642 	Other minor fixlets while at it.
 5643 
 5644 2013-05-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5645 
 5646 	test-lib: minor style changes
 5647 
 5648 	* t/ax/am-test-lib.sh (require_tool): Here.
 5649 
 5650 2013-05-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5651 
 5652 	test-lib: fix botched function name in an error message
 5653 
 5654 	* t/ax/am-test-lib.sh (useless_vpath_rebuild): Here.
 5655 
 5656 2013-05-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5657 
 5658 	automake: remove one extra trailing whitespace
 5659 
 5660 	* bin/automake.in (lang_vala_ignore): Here.
 5661 
 5662 2013-05-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5663 
 5664 	automake: cosmetic changes in indentation and line breaks
 5665 
 5666 	* bin/automake.in (handle_LIBOBJS_or_ALLOCA): Here.
 5667 
 5668 2013-05-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5669 
 5670 	Merge branch 'maint'
 5671 
 5672 	* maint:
 5673 	  maintcheck: minor tweaks and fixlets
 5674 
 5675 2013-05-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5676 
 5677 	Merge branch 'micro' into maint
 5678 
 5679 	* micro:
 5680 	  maintcheck: minor tweaks and fixlets
 5681 
 5682 2013-05-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5683 
 5684 	maintcheck: minor tweaks and fixlets
 5685 
 5686 	* syntax-checks.mk: Here.
 5687 
 5688 2013-05-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5689 
 5690 	Merge branch 'maint'
 5691 
 5692 	* maint:
 5693 	  tests: fix a spurious failure on NetBSD 5.1
 5694 	  texi: build version.texi and stamp-vti in srcdir
 5695 	  tests: fix a botched heading comment
 5696 	  tests: fix another spurious with FreeBSD make
 5697 	  tests: fix a spurious failure with FreeBSD make
 5698 
 5699 2013-05-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5700 
 5701 	tests: fix a spurious failure on NetBSD 5.1
 5702 
 5703 	* t/dist-shar.sh ($required): Also require the 'unshar' program.
 5704 	Apparently, NetBSD has a 'shar' program but not the corresponding
 5705 	'unshar' one.
 5706 
 5707 2013-05-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5708 
 5709 	texi: build version.texi and stamp-vti in srcdir
 5710 
 5711 	Do so even when the 'info-in-builddir' option is present, or when
 5712 	the corresponding '*.info' files are listed in $(CLEANFILES) or in
 5713 	$(DISTCLEANFILES).
 5714 
 5715 	This fixes failures in the following tests, when they are run with
 5716 	$MAKE pointing to FreeBSD make:
 5717 
 5718 	    - txinfo-nodist-info.sh
 5719 	    - txinfo23.sh
 5720 	    - txinfo24.sh
 5721 	    - txinfo28.sh
 5722 	    - txinfo25.sh
 5723 
 5724 	BTW, notice that the test 'txinfo-builddir.sh' fails with FreeBSD make
 5725 	as well, but that is due to a known FreeBSD make VPATH issue (the same
 5726 	described in automake bug#7884).  But that is not a regression, since
 5727 	the 'info-in-builddir' option will be new in Automake 1.14.  Moreover,
 5728 	we already warn in the manual that the use of that option can indeed
 5729 	cause problems with VPATH builds done by non-GNU make.
 5730 
 5731 	* bin/automake.in (handle_texinfo_helper): New local variable
 5732 	'$soutdir'.  Use it instead of '$outdir' where appropriate (in
 5733 	particular, in the transform used on file 'texi-vers.am'.
 5734 	* t/txinfo-builddir.sh: Adjust to avoid spurious failures.
 5735 
 5736 2013-05-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5737 
 5738 	tests: fix a botched heading comment
 5739 
 5740 	* t/parallel-tests-recheck-pr11791.sh: Here.
 5741 
 5742 2013-05-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5743 
 5744 	tests: fix another spurious with FreeBSD make
 5745 
 5746 	* t/parallel-tests-recheck-pr11791.sh: Here.
 5747 
 5748 2013-05-18  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5749 
 5750 	tests: fix a spurious failure with FreeBSD make
 5751 
 5752 	Failures due to known VPATH support issues in that make implementation
 5753 	(the same issues that have been causing the long-standing bug#7884).
 5754 
 5755 	* t/ltlibsrc.sh: Adjust.
 5756 
 5757 2013-05-18  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5758 
 5759 	tests: fix a spurious failure with FreeBSD make
 5760 
 5761 	Failures due to known VPATH support issues in that make implementation
 5762 	(the same issues that have been causing the long-standing bug#7884).
 5763 
 5764 	* t/lex-multiple.sh: Adjust.
 5765 
 5766 2013-05-18  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5767 
 5768 	Merge branch 'subdir-objects-work'
 5769 
 5770 	* subdir-objects-work:
 5771 	  depend: reduce code duplication, de-uglify make verbose output
 5772 	  depend: prefer $(...) over `...` in Cygwin-specific recipes
 5773 	  depend: reduce code duplication
 5774 	  depend: a preparatory refactoring
 5775 	  depend: refactor and remove code duplication
 5776 	  Automake::Language: drop unused fields 'compile_flag' and 'output_flag'
 5777 	  depend: assume '-c' compiler flag always means to produce object files
 5778 	  depend: assume we can always pass '-o' to the C compiler
 5779 	  cleanup: having subdir-objects mandatory allow us some simplifications
 5780 	  subdir-objects: enable unconditionally
 5781 
 5782 2013-05-18  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5783 
 5784 	Merge branch 'maint'
 5785 
 5786 	* maint:
 5787 	  lisp: fix a failure with Solaris /usr/xpg4/bin/sh
 5788 	  tests: sanitize 'unset' usages
 5789 	  tests: fix some botched/outdated comments
 5790 	  tests: use perl, not find+rm, to remove temporary directories
 5791 
 5792 2013-05-18  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5793 
 5794 	Merge branch 'micro' into maint
 5795 
 5796 	* micro:
 5797 	  lisp: fix a failure with Solaris /usr/xpg4/bin/sh
 5798 	  tests: sanitize 'unset' usages
 5799 	  tests: fix some botched/outdated comments
 5800 	  tests: use perl, not find+rm, to remove temporary directories
 5801 
 5802 2013-05-18  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5803 
 5804 	lisp: fix a failure with Solaris /usr/xpg4/bin/sh
 5805 
 5806 	* lib/am/lisp.am (.el.elc): By initializing the 'am__dir' properly
 5807 	here.  For most shells, the lacking initialization, while technically
 5808 	incorrect, didn't cause any issue in practice, because in those shells
 5809 	"test -d" returns an exit status of 0. But with /usr/xpg4/bin/sh, the
 5810 	shell complains like this: "test: argument expected", and returns a
 5811 	non-zero exit status.  This caused testsuite failures in several lisp
 5812 	tests.
 5813 	Also, while we are at it, use more proper quoting in the recipe, to
 5814 	ensure a missing initialization to now be caught by more forgiving
 5815 	shells as well.
 5816 	* NEWS: Update.
 5817 
 5818 2013-05-17  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5819 
 5820 	tests: sanitize 'unset' usages
 5821 
 5822 	In some shells (e.g., Solaris 10 /bin/ksh, or NetBSD 5.1 /bin/sh),
 5823 	"unset VAR" returns a non-zero exit status in case the VAR variable
 5824 	is already unset.  This doesn't interact well with our usage of
 5825 	"set -e" in the testsuite.  So far, we've avoided spurious failures
 5826 	by either explicitly ignoring the exit status from unset:
 5827 
 5828 	    unset VAR || :
 5829 
 5830 	or explicitly ensuring that a variable is set, before trying to
 5831 	unset it:
 5832 
 5833 	    VAR=; unset VAR
 5834 
 5835 	But we can do better, by aliasing the 'unset' command to a custom
 5836 	function that will take care of these details for us.  This will
 5837 	avoid us annoying spurious failures in the future, failures that
 5838 	have already bitten us too much times.  For an example, refer to
 5839 	commit 'v1.12.2-88-g5b1dae5' of 2012-08-05 (tests: avoid tons of
 5840 	spurious failures on NetBSD).
 5841 
 5842 	* t/ax/test-lib.sh (_am_unset): New function.
 5843 	(unset): New alias to it.
 5844 	(_am_exit): Adjust comments.
 5845 	* t/ax/am-test-lib.sh: No need to temporary disable the 'errexit'
 5846 	shell flag when unsetting variables that are potentially already
 5847 	unset.
 5848 	(am_process_requirements): Adjust to remove a now-useless
 5849 	workaround related to unset.
 5850 	* t/aclocal-macrodir.tap: Likewise.
 5851 	* t/aclocal-macrodirs.tap: Likewise.
 5852 	* t/auxdir-autodetect.sh: Likewise.
 5853 	* t/ax/am-test-lib.sh: Likewise.
 5854 	* t/ax/test-lib.sh: Likewise.
 5855 	* t/check-tests-in-builddir.sh: Likewise.
 5856 	* t/dist-formats.tap: Likewise.
 5857 	* t/distcheck-configure-flags-am.sh: Likewise.
 5858 	* t/distcheck-configure-flags.sh: Likewise.
 5859 	* t/java-empty-classpath.sh: Likewise.
 5860 	* t/javaflags.sh: Likewise.
 5861 	* t/lflags.sh: Likewise.
 5862 	* t/lflags2.sh: Likewise.
 5863 	* t/lisp-flags.sh: Likewise.
 5864 	* t/lisp6.sh: Likewise.
 5865 	* t/missing-auxfile-stops-makefiles-creation.sh: Likewise.
 5866 	* t/parallel-am.sh: Likewise.
 5867 	* t/parallel-am2.sh: Likewise.
 5868 	* t/parallel-am3.sh: Likewise.
 5869 	* t/parallel-tests-log-override-recheck.sh: Likewise.
 5870 	* t/pkg-config-macros.sh: Likewise.
 5871 	* t/python-missing.sh: Likewise.
 5872 	* t/python-too-old.sh: Likewise.
 5873 	* t/python11.sh: Likewise.
 5874 	* t/self-check-dir.tap: Likewise.
 5875 	* t/self-check-report.sh: Likewise.
 5876 	* t/self-check-seq.tap: Likewise.
 5877 	* t/silent-configsite.sh: Likewise.
 5878 	* t/suffix6c.sh: Likewise.
 5879 	* t/tar-override.sh: Likewise.
 5880 	* t/tests-environment-and-log-compiler.sh: Likewise.
 5881 	* t/vala-configure.sh: Likewise.
 5882 	* t/werror3.sh: Likewise.
 5883 	* t/yflags-cmdline-override.sh: Likewise.
 5884 	* t/yflags.sh: Likewise.
 5885 	* t/yflags2.sh: Likewise.
 5886 
 5887 2013-05-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5888 
 5889 	tests: fix some botched/outdated comments
 5890 
 5891 	* t/ax/test-defs.in: Here.  Also, fix a couple of grammaros while at it.
 5892 
 5893 2013-05-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5894 
 5895 	tests: use perl, not find+rm, to remove temporary directories
 5896 
 5897 	The File::Path::rmtree function from perl, if used right, is
 5898 	more reliable and more portable of our past idiom:
 5899 
 5900 	    find $dirs -type d ! -perm -700 -exec chmod u+rwx {} ';';
 5901 	    rm -rf $$dirs || exit 1
 5902 
 5903 	at least of the face of unreadable dirs/files and other similar
 5904 	permission issues (and we have those in our test directories).
 5905 
 5906 	In fact, this change fixes some spurious failures seen in
 5907 	"make distcheck" on Solaris 10.
 5908 
 5909 	* t/ax/deltree.pl: New.
 5910 	* Makefile.am (EXTRA_DIST): Add it.
 5911 	(clean-local-check): Use it.
 5912 	* t/ax/test-lib.sh (rm_rf_): Use it.
 5913 
 5914 2013-05-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5915 
 5916 	depend: reduce code duplication, de-uglify make verbose output
 5917 
 5918 	Verbose output from compilation recipes had been made more messy by
 5919 	the previous "simplifying" commit v1.13.1d-220-g1fa0c24.  With this
 5920 	patch, not only we reduce some code duplication, but also improve
 5921 	that output again, for the joy of users who dislike silent-rules :-)
 5922 
 5923 	* lib/am/depend2.am: Use the new '%XSOURCE%' transform, instead of
 5924 	copying and pasting the idiom for VPATH rewrite emulation over and
 5925 	over.  Fix some indentation and line-wrapping issues while at it.
 5926 	* bin/automake.in (handle_languages): Implement the new transform.
 5927 	* t/silent-nested-vars.sh: Adjust to avoid spurious failures.
 5928 
 5929 2013-05-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5930 
 5931 	tests: remove remaining exec bits ('master' branch)
 5932 
 5933 	The executable bit gives the impression that the tests are directly
 5934 	runnable, as with "./t/foo.sh", but it has been a while since that
 5935 	was the case.  Today, tests are runnable only through "make check"
 5936 	or "./runtest".
 5937 
 5938 	This change is for the 'master' branch (automake 1.99a), and is a
 5939 	follow-up to commit 'v1.13.2-3-g74017b5', done on the 'micro' branch
 5940 	(automake 1.13.2a) and to commit 'v1.13.2-221-g0985d67', done on the
 5941 	'maint' branch (automake 1.13a).
 5942 
 5943 	* t/am-prog-mkdir-p.sh: Remove executable bit.
 5944 	* t/txinfo-no-split.sh: Likewise.
 5945 
 5946 2013-05-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5947 
 5948 	Merge branch 'maint'
 5949 
 5950 	* maint:
 5951 	  tests: remove remaining exec bits ('maint' branch)
 5952 	  tests: remove exec bit from all of them ('micro' branch)
 5953 	  maint: tests no longer need to have executable bit set
 5954 	  PLANS: subdir-objects: various updates
 5955 
 5956 2013-05-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5957 
 5958 	tests: remove remaining exec bits ('maint' branch)
 5959 
 5960 	The executable bit gives the impression that the tests are directly
 5961 	runnable, as with "./t/foo.sh", but it has been a while since that
 5962 	was the case.  Today, tests are runnable only through "make check"
 5963 	or "./runtest".
 5964 
 5965 	This change is for the 'maint' branch (automake 1.13a), and is a
 5966 	follow-up to commit 'v1.13.2-3-g74017b5', done on the 'micro' branch
 5967 	(automake 1.13.2a).  It will soon be followed by a similar patch
 5968 	for the 'master' branch (automake 1.99a).
 5969 
 5970 	* t/am-prog-cc-c-o.sh: Remove executable bit.
 5971 	* t/ccnoco4.sh: Likewise.
 5972 	* t/dist-shar.sh: Likewise.
 5973 	* t/dist-tarZ.sh: Likewise.
 5974 	* t/lex-multiple.sh: Likewise.
 5975 	* t/preproc-basics.sh: Likewise.
 5976 	* t/preproc-c-compile.sh: Likewise.
 5977 	* t/preproc-demo.sh: Likewise.
 5978 	* t/preproc-errmsg.sh: Likewise.
 5979 	* t/rm-f-probe.sh: Likewise.
 5980 	* t/self-check-cc-no-c-o.sh: Likewise.
 5981 	* t/txinfo-builddir.sh: Likewise.
 5982 	* t/txinfo-nodist-info.sh: Likewise.
 5983 
 5984 2013-05-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5985 
 5986 	Merge branch 'micro' into maint
 5987 
 5988 	* micro:
 5989 	  tests: remove exec bit from all of them ('micro' branch)
 5990 	  maint: tests no longer need to have executable bit set
 5991 
 5992 2013-05-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
 5993 
 5994 	tests: remove exec bit from all of them ('micro' branch)
 5995 
 5996 	It gives the impression that they are directly runnable, as with
 5997 	"./t/foo.sh", but it has been a while since that was the case.  Today,
 5998 	tests are runnable only through "make check" or "./runtest".
 5999 
 6000 	This change is for the 'micro' branch (automake 1.13.2a).  It will
 6001 	soon be followed by similar patches for the 'maint' branch (automake
 6002 	1.13a) and the 'master' branch (automake 1.99a).
 6003 
 6004 	* t/*.sh, t/*.tap: Remove executable bit.
 6005 	* maint.mk (sc_tests_executable): Remove.
 6006 	(syntax_check_rules): Adjust.
 6007 	* gen-testsuite-part: Set permissions of generated tests to
 6008 	'444' (-r--r--r--), rather than 555 (-r-xr-xr-x).
 6009 
 6010 2013-05-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6011 
 6012 	maint: tests no longer need to have executable bit set
 6013 
 6014 	It has been a while since they are runnable only through "make check"
 6015 	or "./runtest", and not directly.
 6016 
 6017 	* HACKING: Adjust.
 6018 	* t/README: Likewise, and updated the surrounding text to get rid of
 6019 	obsolete statements and advices.
 6020 
 6021 2013-05-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6022 
 6023 	PLANS: subdir-objects: various updates
 6024 
 6025 2013-05-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6026 
 6027 	Merge branch 'maint'
 6028 
 6029 	* maint:
 6030 	  post-release: micro version bump (1.13.2a)
 6031 	  release: stable micro release 1.13.2
 6032 	  vala tests: skip in a cross compiler setup
 6033 	  HACKING: miscellaneous fixes, updates and enhancements
 6034 	  NEWS: minor improvements to wording (about new versioning scheme)
 6035 	  THANKS: update Akim's e-mail address
 6036 	  tests: less uses of "make -e"; avoid spurious failures in 'check-cc-no-c-o'
 6037 	  build: be more respectful of user-specified verbosity
 6038 	  check-cc-no-c-o: unify initializations in a single place
 6039 	  check-cc-no-c-o: avoid a spurious failure
 6040 
 6041 2013-05-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6042 
 6043 	Merge branch 'micro' into maint
 6044 
 6045 	* micro:
 6046 	  post-release: micro version bump (1.13.2a)
 6047 	  release: stable micro release 1.13.2
 6048 	  vala tests: skip in a cross compiler setup
 6049 	  HACKING: miscellaneous fixes, updates and enhancements
 6050 	  NEWS: minor improvements to wording (about new versioning scheme)
 6051 
 6052 2013-05-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6053 
 6054 	post-release: micro version bump (1.13.2a)
 6055 
 6056 	* configure.ac, m4/amversion.m4: Bump version: 1.13.2 -> 1.13.2a
 6057 
 6058 2013-05-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6059 
 6060 	release: stable micro release 1.13.2
 6061 
 6062 	* configure.ac (AC_INIT): Bump version number to 1.13.2.
 6063 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 6064 
 6065 2013-05-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6066 
 6067 	vala tests: skip in a cross compiler setup
 6068 
 6069 	The "vala + pkg-config" doesn't interact well with cross-compilation:
 6070 	<https://mail.gnome.org/archives/vala-list/2012-September/msg00094.html>
 6071 
 6072 	That is causing annoying failures in our testsuite when it's run in
 6073 	cross-compilation mode.  It's not worth trying too be overly smart and
 6074 	work around these problems; just skip the affected tests instead.
 6075 
 6076 	* t/ax/am-test-lib.sh (require_tool): A new requirement 'valac', telling
 6077 	that the test needs to compile Vala-generated C files.  In particular,
 6078 	this causes the test to be skipped when using a cross-compiler.
 6079 
 6080 2013-05-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6081 
 6082 	HACKING: miscellaneous fixes, updates and enhancements
 6083 
 6084 2013-05-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6085 
 6086 	NEWS: minor improvements to wording (about new versioning scheme)
 6087 
 6088 2013-05-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6089 
 6090 	THANKS: update Akim's e-mail address
 6091 
 6092 2013-05-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6093 
 6094 	depend: prefer $(...) over `...` in Cygwin-specific recipes
 6095 
 6096 	On Cygwin, we can happily assume only POSIX-conforming shells exist.
 6097 	This commit is useless by itself, but is useful as a preparatory
 6098 	refactoring in view of a future change.
 6099 
 6100 	* lib/am/depend2.am: In recipes for producing '.obj' objects, use
 6101 	the POSIX form $(...) for command substitution, rather than the
 6102 	classic Bourne one `...`.
 6103 
 6104 2013-05-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6105 
 6106 	depend: reduce code duplication
 6107 
 6108 	Just a simplification; no semantic change is intended.
 6109 
 6110 	* lib/am/depend2.am (am__set_depbase): New.  Use it in all the three
 6111 	flavours of recipes (.o, .obj, .lo), instead of duplicating its code.
 6112 
 6113 2013-05-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6114 
 6115 	depend: a preparatory refactoring
 6116 
 6117 	This is just a preparatory change in view of a future refactoring.
 6118 	No semantic change is intended.
 6119 
 6120 	* lib/am/depend2.am: Adjust; for example, using '$@' in all recipes
 6121 	to indicate the output file (rather than recipe-specific transforms
 6122 	like '%OBJ%' and '%LTOBJ%'), and not assuming to know the exact
 6123 	file extension of the output file (e.g., whether it's '.o' or '.lo').
 6124 
 6125 2013-05-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6126 
 6127 	depend: refactor and remove code duplication
 6128 
 6129 	* lib/am/depend2.am: Here.  The obsolescent comments removed from this
 6130 	same file in commit 'v1.13.1-42-g8f06bfb' of 2012-01-09, "depend2.am: fix
 6131 	comments on verbosity of compilation rules", gave the rationale for why
 6132 	that code duplication was there in the first place (rationale that, like
 6133 	those comments, has been obsolete by the silent-rules introduction).
 6134 	* bin/automake.in: Given the refactoring in 'depend2.am', there is no
 6135 	longer need to add extra trailing whitespace to the entries of the
 6136 	'%sourceflags' hash (which are used for the '%SOURCEFLAG%' transform
 6137 	when 'depend2.am' is processed.
 6138 	* t/fort2.sh: Adjust to avoid spurious failures.
 6139 
 6140 2013-05-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6141 
 6142 	Automake::Language: drop unused fields 'compile_flag' and 'output_flag'
 6143 
 6144 	They are unused after the recent flurry of changes.  This is just a
 6145 	simplification; no semantic change is intended.
 6146 
 6147 	* lib/Automake/Language.pm (Class::Struct): Remove unused fields here.
 6148 	* bin/automake.in: And here, in all the 'register_language()' invocations.
 6149 
 6150 2013-05-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6151 
 6152 	depend: assume '-c' compiler flag always means to produce object files
 6153 
 6154 	Rather than fully-fledged executables.  This is the case for all the
 6155 	compilers of all languages supported by Automake.
 6156 
 6157 	This is just a simplification; no semantic changes are intended.
 6158 
 6159 	* bin/automake.in (handle_languages): Drop transform '%-c%'.
 6160 	* lib/am/depend2.am: Adjust to just assume the '%-c%' transform
 6161 	expands to "-c".
 6162 
 6163 2013-05-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6164 
 6165 	depend: assume we can always pass '-o' to the C compiler
 6166 
 6167 	This is the case after the change in commit v1.13.1-56-g34001a9 of
 6168 	2013-01-09 (compile: use 'compile' script when "-c -o" is used with
 6169 	losing compilers).
 6170 
 6171 	This is just a simplification; no semantic changes are intended.
 6172 
 6173 	* bin/automake.in (handle_languages): Drop transform '%-o%', and
 6174 	definitions of '$output_flag'.
 6175 	* lib/am/depend2.am: Adjust to just assume the '?-o?' transform is
 6176 	true, and the '%-o%' transform expands to "-o".
 6177 
 6178 2013-05-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6179 
 6180 	cleanup: having subdir-objects mandatory allow us some simplifications
 6181 
 6182 	Few minor cleanups made possible by earlier changes, plus other minor
 6183 	cleanups triggered in cascade.  No semantic change is intended.
 6184 
 6185 	This is a follow-up on previous commit 'v1.13.1d-214-g3ad07d2', and
 6186 	an adjusted backport of Automake-NG commit 'v1.12.1-315-gc97d41b'
 6187 	of 2012-06-08 ([ng] cleanup: after enabling of subdir-objects
 6188 	unconditionally).
 6189 
 6190 	* bin/automake.in (LANG_IGNORE, LANG_SUBDIR): Remove.
 6191 	(handle_languages): Drop the '%DEPBASE%' transform when processing
 6192 	the '$rule_file'.
 6193 	(register_language ('name' => 'vala', ...)): Add '.vapi' to the entry
 6194 	'extensions', and simplify the entry 'output_extensions' to point to
 6195 	a dummy subroutine (since it wasn't really used anyway).
 6196 	(handle_single_transform): No longer expect the 'lang_*_rewrite'
 6197 	subroutines to return a 'LANG_*' constant, but only a transformed
 6198 	extension, if required.  To decide whether further processing of the
 6199 	source file should be stopped, rely on a new set of 'lang_*_ignore'
 6200 	subroutines, defaulting to a subroutine that returns false.
 6201 	Accordingly, don't special case the handling of '.vapi' files anymore,
 6202 	instead rely on ...
 6203 	(lang_vala_ignore, lang_header_ignore): ... these new subroutines to
 6204 	avoid extra processing of C/C++ headers and Vala '.vapi' headers.
 6205 	(lang_java_rewrite): Remove.
 6206 	Remove an outdated comment.
 6207 	* lib/am/depend2.am: Partial rewrite to reduce code duplication and
 6208 	drop use of the '%DEPBASE%' transform.
 6209 	* t/compile_f_c_cxx.sh: Adjust.
 6210 
 6211 2013-05-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6212 
 6213 	subdir-objects: enable unconditionally
 6214 
 6215 	See automake bug#13378.
 6216 
 6217 	The fact that Automake-generated Makefiles places compiled object files
 6218 	in the current directory by default, also when the corresponding source
 6219 	file is in a subdirectory, is basically an historical accident, due to
 6220 	the fact that the 'subdir-objects' option had only been introduced in
 6221 	April 1999, starting with commit 'user-dep-gen-branchpoint-56-g88b5959',
 6222 	and never made the default, likely to avoid backwards-compatibility
 6223 	issues.
 6224 
 6225 	Since we believe the behaviour enabled by the 'subdir-objects' is the
 6226 	only natural and most useful one, we make it the only only one available,
 6227 	simplifying the Automake implementation and APIs a little in the process.
 6228 
 6229 	This change is basically an adjusted backport of Automake-NG commit
 6230 	'v1.12.1-313-g14fe163' of 2012-06-07, "[ng] subdir-objects: enable
 6231 	unconditionally".
 6232 
 6233 	* NEWS: Update.
 6234 	* doc/automake.texi (Program and Library Variables): The output
 6235 	object files are no longer placed in the current directory by
 6236 	default, but rather in the same directory of the source file.
 6237 	(LIBOBJS): Now the $(LIBOBJS) and $(ALLOCA) variables can also
 6238 	be used outside of the directory where their sources lie.
 6239 	(List of Automake options): Report the 'subdir-objects' option
 6240 	as a no-op, existing only for compatibility with older versions
 6241 	of Automake.
 6242 	Other related minor adjustments.
 6243 	* bin/automake.in (LANG_PROCESS): Remove, it's no longer needed.
 6244 	(handle_languages): Don't test whether option 'subdir-objects'
 6245 	is set (just assume it is), and do not use the '%SUBDIROBJ%'
 6246 	transform when processing '.am' fragments.
 6247 	(lang_sub_obj): Delete, it would just return 'LANG_SUBDIR'
 6248 	unconditionally now.
 6249 	(lang_lex_rewrite): Adjust.  Don't test whether the option
 6250 	'subdir-objects' is set (just assume it is).
 6251 	(lang_yacc_rewrite): Likewise.
 6252 	(handle_single_transform): Likewise.  Remove an obsolete
 6253 	comment.  Add a proper "FIXME" comments about a fragment
 6254 	of code that might have become dead code now.
 6255 	(handle_LIBOBJS_or_ALLOCA): Simplify assuming that the option
 6256 	'subdir-objects' is always set.  Accordingly, there's no need
 6257 	to warn anymore if '$(LIBOBJS)' or '$(ALLOCA)' are used outside
 6258 	the '$config_libobj_dir' directory (as specified by autoconf
 6259 	macro 'AC_CONFIG_LIBOBJ_DIR').
 6260 	* lib/am/depend2.am: Assume the '?SUBDIROBJ?' Automake time
 6261 	conditional is always true, and remove its uses accordingly.
 6262 	* t/compile_f_c_cxx.sh: Adjust.
 6263 	* t/cscope.tap: Likewise.
 6264 	* t/depcomp8a.sh: Likewise.
 6265 	* t/depcomp8b.sh: Likewise.
 6266 	* t/libtool3.sh: Likewise.
 6267 	* t/ltlibsrc.sh: Likewise.
 6268 	* t/pr401.sh: Likewise.
 6269 	* t/pr401b.sh: Likewise.
 6270 	* t/pr401c.sh: Likewise.
 6271 	* t/subobj.sh: Likewise.
 6272 	* t/lex-line.sh: Likewise.
 6273 	* t/yacc-line.sh: Likewise.
 6274 	* t/yacc5.sh: Likewise.
 6275 	* t/vala-libs.sh: Likewise.
 6276 	* t/fort4.sh: Likewise, and extend a bit.
 6277 	* t/fort5.sh: Likewise.
 6278 	* t/gcj.sh: Likewise.
 6279 	* t/subpkg.sh: Likewise.
 6280 	* t/subpkg-yacc.sh: Likewise.
 6281 	* t/xsource.sh: Likewise.
 6282 	* t/libobj20a.sh: Remove as obsolete.
 6283 	* t/libobj20b.sh: Adjust heading comments.
 6284 	* t/libobj20c.sh: Likewise.
 6285 	* t/subobj4.sh: Remove as obsolete.
 6286 	* t/sourcefile-in-subdir.sh: Likewise.
 6287 
 6288 2013-05-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6289 
 6290 	tests: less uses of "make -e"; avoid spurious failures in 'check-cc-no-c-o'
 6291 
 6292 	That is, when the testsuite is run using a fake C compiler that doesn't
 6293 	grasp the '-c' and '-o' options together.
 6294 
 6295 	* t/instdir-prog.sh: Adjust.
 6296 	* t/instdir-ltlib.sh: Likewise.
 6297 	* t/python-virtualenv.sh: Likewise.
 6298 
 6299 2013-05-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6300 
 6301 	build: be more respectful of user-specified verbosity
 6302 
 6303 	* t/Makefile.in (check-cc-no-c-o, check-no-trailing-backslash-in-recipes,
 6304 	installcheck-testsuite, perf): Here.
 6305 
 6306 2013-05-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6307 
 6308 	check-cc-no-c-o: unify initializations in a single place
 6309 
 6310 	* t/ax/test-defs.in: That is, by setting CC and GNU_CC here, in accord
 6311 	with the value of the variable 'AM_TESTSUITE_SIMULATING_NO_CC_C_O'.
 6312 	* t/Makefile.in (check-cc-no-c-o) : No need to reset CC and GNU_CC any
 6313 	longer in the recursive "make check" invocation.
 6314 
 6315 2013-05-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6316 
 6317 	check-cc-no-c-o: avoid a spurious failure
 6318 
 6319 	* t/am-prog-cc-c-o.sh: In this test, by relying on the knowledge
 6320 	that we are running under the aegis of the 'check-cc-no-c-o'
 6321 	maintainer-specific target, knowledge given us by ...
 6322 	* t/Makefile.in (check-cc-no-c-o) : ... the new environment
 6323 	variable 'AM_TESTSUITE_SIMULATING_NO_CC_C_O', set to a value
 6324 	of "yes" by this rule.
 6325 	* t/ax/test-defs.in: Initialize the new variable to "no" by
 6326 	default, and add an explanatory comment.
 6327 
 6328 2013-05-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6329 
 6330 	Merge branch 'maint'
 6331 
 6332 	* maint:
 6333 	  build: fixup for building in a VPATH setup
 6334 
 6335 2013-05-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6336 
 6337 	build: fixup for building in a VPATH setup
 6338 
 6339 	* bin/Makefile.inc (%D%/automake, %D%/aclocal): Make sure that the
 6340 	directory where the targets scripts are going to be built exists,
 6341 	before trying to create said scripts.
 6342 
 6343 2013-05-13  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6344 
 6345 	Merge branch 'maint'
 6346 
 6347 	* maint:
 6348 	  Use AC_DEFUN_ONCE to define AM_PROG_CC_C_O
 6349 	  compile: avoid AC_PROG_CC messy rewrite
 6350 	  options: tiny simplification in dealing with incompatible versions
 6351 
 6352 2013-05-11  Nick Bowler  <nbowler@elliptictech.com>  (tiny change)
 6353 
 6354 	Use AC_DEFUN_ONCE to define AM_PROG_CC_C_O
 6355 
 6356 	If AM_PROG_CC_C_O is expanded multiple times, and the compiler does not
 6357 	support -c and -o together, each expansion of the macro will prepend
 6358 	the compile script to CC.  This can result in the compile script
 6359 	invoking the compile script, which at best pointless and silly.
 6360 	Fortunately, there does not appear to be any serious problems as the
 6361 	first compile invocation strips out -o options, causing subsequent
 6362 	invocations of the script to merely exec their arguments.
 6363 
 6364 	Other than fixing the above, this should not normally cause any changes
 6365 	to the resulting configure script, except in the (hopefully rare) case
 6366 	where AM_PROG_CC_C_O is directly expanded (i.e., *not* using AC_REQUIRE)
 6367 	in the body of a macro defined with AC_DEFUN.  In that case, the use of
 6368 	AC_DEFUN_ONCE may cause the expansion of AM_PROG_CC_C_O to appear
 6369 	earlier in the configure script.
 6370 
 6371 	* m4/minuso.m4: Change the definition of AM_PROG_CC_C_O to use
 6372 	AC_DEFUN_ONCE, avoiding problems caused by multiple expansions.
 6373 
 6374 2013-05-11  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6375 
 6376 	compile: avoid AC_PROG_CC messy rewrite
 6377 
 6378 	Instead, add an hook to AC_OUTPUT to have AM_PROG_CC_C_O invoked
 6379 	automatically.
 6380 
 6381 	See also the long-winded discussion about automake bug#13378.
 6382 
 6383 	* m4/minuso.m4 (AM_PROG_CC_C_O): Bring back the old implementation,
 6384 	from commit v1.13.1-55-g1ab8fb6.
 6385 	* m4/init.m4 (AC_PROG_CC): Remove this horrible, hacky re-write.
 6386 	* (AM_INIT_AUTOMAKE): Arrange for AM_PROG_CC_C_O to be called if
 6387 	necessary.
 6388 	* t/am-prog-cc-c-o.sh: Adjust to avoid spurious failure.
 6389 	* t/subobj.sh: Likewise.
 6390 
 6391 	Suggested-by: Nick Bowler <nbowler@elliptictech.com>
 6392 
 6393 2013-05-11  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6394 
 6395 	options: tiny simplification in dealing with incompatible versions
 6396 
 6397 	* lib/Automake/Options.pm (_process_option_list): Here, when an
 6398 	incompatible version number option is detected, there's no need
 6399 	to call error() with the "uniq_scope => US_GLOBAL" switch.
 6400 	In fact, if the same incompatible version number is specified in
 6401 	AUTOMAKE_OPTIONS in both (say) 'Makefile.am' and 'sub/Makefile.am',
 6402 	we want each such erroneous usage reported separately, rather than
 6403 	just the first time it is encountered (as we'd expect to happen
 6404 	when "uniq_scope => US_GLOBAL" is used).
 6405 	Ideally, this change should have been folded into the similar
 6406 	commit 'v1.13.1d-129-gf7ef16f', but we noticed that too late.
 6407 	Oh well.
 6408 
 6409 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6410 
 6411 	options: consistently use return statuses to report errors (more)
 6412 
 6413 	This is a follow-up to commit 'v1.13.1d-128-g6e486c5', needed to deal
 6414 	with errors that are only present in the 'master' branch (while the
 6415 	reported commit was on the 'maint' branch).
 6416 
 6417 	* lib/Automake/Options.pm (_process_option_list): Adjust, where dealing
 6418 	with the removed options 'dist-shar' and 'dist-tarZ'.
 6419 
 6420 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6421 
 6422 	Merge branch 'maint'
 6423 
 6424 	* maint:
 6425 	  options: try to report as much errors as possible
 6426 	  warns: don't tell AM_PROG_MKDIR_P is going to be removed
 6427 	  refactor: fix few "inverted boolean" usages
 6428 	  options: better name for an internal function
 6429 	  options: more consistency in use of return statuses to report errors
 6430 	  options: tiny simplification in dealing with erroneous opts
 6431 	  options: consistently use return statuses to report errors
 6432 	  options: re-enable some sanity checks
 6433 	  THANKS: update Eric Blake's e-mail address
 6434 	  NEWS: typofix
 6435 
 6436 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6437 
 6438 	options: try to report as much errors as possible
 6439 
 6440 	For example, if two invalid options are used in AUTOMAKE_OPTIONS, don't
 6441 	report just the first one, but both of them.
 6442 
 6443 	* lib/Automake/Options.pm (_process_option_list): Do so by avoiding
 6444 	early returns in here.
 6445 
 6446 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6447 
 6448 	Merge branch 'branch-1.13.2' into maint
 6449 
 6450 	* branch-1.13.2:
 6451 	  warns: don't tell AM_PROG_MKDIR_P is going to be removed
 6452 
 6453 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6454 
 6455 	warns: don't tell AM_PROG_MKDIR_P is going to be removed
 6456 
 6457 	That is no longer true.  For a more extended rationale, see file
 6458 	'PLANS/obsolete-removed/am-prog-mkdir-p.txt' in the maint branch
 6459 	(as of commit v1.13.1d-132-g90ec3fe).
 6460 
 6461 	* automake.in (scan_autoconf_traces): So adjust the warning message
 6462 	here.
 6463 
 6464 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6465 
 6466 	refactor: fix few "inverted boolean" usages
 6467 
 6468 	In some subroutines, we used a return value of 0 to indicate success,
 6469 	and a return status of 1 to indicate failure.  That was not very
 6470 	consistent with the perl interpretation of 0 as a false value and 1 as
 6471 	a true value.  So we now invert the meaning of the exit statuses.
 6472 
 6473 	* lib/Automake/Options.pm (_process_option_list): Here.
 6474 	(process_global_option_list, process_option_list): And by reflex,
 6475 	here as well.
 6476 	* bin/automake.in (handle_options): And here.
 6477 	(generate_makefile, scan_autoconf_traces): Adjust.
 6478 
 6479 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6480 
 6481 	options: better name for an internal function
 6482 
 6483 	* lib/Automake/Options.pm (_option_must_be_from_configure): Rename ...
 6484 	(_option_is_from_configure): ... like this.
 6485 
 6486 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6487 
 6488 	options: more consistency in use of return statuses to report errors
 6489 
 6490 	* lib/Automake/Options.pm (_option_must_be_from_configure): By giving
 6491 	a proper return status here.
 6492 	(_process_option_list): And using it here.
 6493 
 6494 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6495 
 6496 	options: tiny simplification in dealing with erroneous opts
 6497 
 6498 	* lib/Automake/Options.pm (_process_option_list): Here, when an
 6499 	invalid option is detected, there's no need to call &error with
 6500 	the "uniq_scope => US_GLOBAL" switch.  In fact, if the same
 6501 	erroneous option is specified in AUTOMAKE_OPTIONS in both (say)
 6502 	'Makefile.am' and 'sub/Makefile.am', we want each such erroneous
 6503 	usage reported separately, rather than just the first time it is
 6504 	encountered (as happens when "uniq_scope => US_GLOBAL" is used).
 6505 
 6506 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6507 
 6508 	options: consistently use return statuses to report errors
 6509 
 6510 	* lib/Automake/Options.pm (_process_option_list): Here.
 6511 	(process_option_list, process_global_option_list): Remove
 6512 	redundant use of 'return'.
 6513 
 6514 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6515 
 6516 	options: re-enable some sanity checks
 6517 
 6518 	They had been unwittingly disabled by a slightly incorrect
 6519 	code ordering.
 6520 
 6521 	* lib/Automake/Options.pm (process_option_list): Here.
 6522 	(process_global_option_list): And here.
 6523 
 6524 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6525 
 6526 	THANKS: update Eric Blake's e-mail address
 6527 
 6528 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6529 
 6530 	NEWS: typofix
 6531 
 6532 	Reported-by: Eric Blake <eblake@redhat.com>
 6533 
 6534 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6535 
 6536 	dist: remove support for shar and tarZ formats
 6537 
 6538 	See also discussion about automake wishlist bug#13324.
 6539 
 6540 	* lib/Automake/Options.pm: Give fatal errors (rather than warnings) if
 6541 	the 'dist-shar' or 'dist-tarZ' options are used.
 6542 	* lib/distdir.am: Remove the 'dist-tarZ' and 'dist-shar' targets, and
 6543 	references to the '.tar.Z' and '.shar' archives.
 6544 	* bin/automake.in (preprocess_file): Remove 'COMPRESS' and 'SHAR'
 6545 	transforms.
 6546 	(handle_dist): Remove lingering references to 'dist-tarZ' and 'dist-shar'
 6547 	options.
 6548 	* doc/automake.texi: Adjust, removing references to the removed targets
 6549 	and distribution formats.
 6550 	* t/dist-shar.sh: Adjust to expect fatal errors rather than warnings.
 6551 	* t/dist-tarZ.sh: Likewise.
 6552 
 6553 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6554 
 6555 	Merge branch 'maint'
 6556 
 6557 	* maint:
 6558 	  news: document new 'subdir-objects' warning
 6559 	  PLANS: one minor fixlet (mostly cosmetic)
 6560 	  PLANS: we have already dropped support for split info files in master
 6561 	  NEWS: fix a reference to Automake 1.14 where Automake 2.0 was intended
 6562 	  PLANS: fix reference to non-existent 'next' branch
 6563 	  PLANS: fix botched version reference
 6564 	  maintcheck: fix two references to old location of aclocal and automake
 6565 	  dist: deprecated shar and tar+compress formats
 6566 
 6567 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6568 
 6569 	news: document new 'subdir-objects' warning
 6570 
 6571 	* NEWS: Automake 1.14 will warn if a subdir source file is
 6572 	specified but the 'subdir-objects' option is not given.  This
 6573 	is done to smooth the transition to Automake 2.0, which will
 6574 	unconditionally assume the behaviour now given only with the
 6575 	'subdir-objects' option.
 6576 
 6577 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6578 
 6579 	PLANS: one minor fixlet (mostly cosmetic)
 6580 
 6581 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6582 
 6583 	PLANS: we have already dropped support for split info files in master
 6584 
 6585 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6586 
 6587 	NEWS: fix a reference to Automake 1.14 where Automake 2.0 was intended
 6588 
 6589 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6590 
 6591 	PLANS: fix reference to non-existent 'next' branch
 6592 
 6593 	* PLANS/obsolete-removed/configure.in.txt: Here.  We should refer
 6594 	to the 'master' branch instead.
 6595 
 6596 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6597 
 6598 	PLANS: fix botched version reference
 6599 
 6600 	* PLANS/rm-f-without-args.txt: Here.  The probe checking that "rm -f"
 6601 	without arguments works will be introduced in Automake 1.14, not in
 6602 	Automake 1.13.2.
 6603 
 6604 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6605 
 6606 	maintcheck: fix two references to old location of aclocal and automake
 6607 
 6608 	* maintainer/syntax-checks.mk (sc_perl_at_substs): Here: it should
 6609 	refer to 'bin/automake' and 'bin/aclocal', not 'automake' and
 6610 	'alocal'.
 6611 
 6612 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6613 
 6614 	dist: deprecated shar and tar+compress formats
 6615 
 6616 	See also discussion about automake wishlist bug#13324.
 6617 
 6618 	* lib/Automake/Options.pm: Give proper warnings in the 'obsolete'
 6619 	category if the 'dist-shar' or 'dist-tarZ' options are used.
 6620 	* lib/distdir.am: When the 'dist-tarZ' or 'dist-shar' targets are
 6621 	invoked, make them give a non-fatal warning.
 6622 	* doc/automake.texi: Report the new deprecations.
 6623 	* t/dist-shar.sh: New test.
 6624 	* t/dist-tarZ.sh: Likewise.
 6625 	* t/lzma.sh: While at it, rename ...
 6626 	* t/dist-lzma.sh: ... like this, and tweak it to keep more in
 6627 	sync with the new tests.
 6628 	* t/dist-formats.tap: Remove references to deprecated formats.
 6629 	* t/list-of-tests.mk: Adjust.
 6630 
 6631 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6632 
 6633 	Merge branch 'maint'
 6634 
 6635 	* maint:
 6636 	  automake: typofix in comments: s/AC_CONFIG_HEADER/AC_CONFIG_HEADERS/
 6637 	  am: prefer a shorter idiom where possible
 6638 	  cosmetics: fix few typos, grammaros and missing whitespace
 6639 	  fixup: remove an obsolete comment
 6640 	  docs: we still don't have the promised better Java interface
 6641 	  build: move automake and aclocal in 'bin' subdir
 6642 	  build: break up monolithic Makefile.am in subdir-specific fragments
 6643 
 6644 	+ Extra non-trivial edits:
 6645 
 6646 	* m4/Makefile.inc (dist_automake_ac_DATA): Drop lead-dot.m4 and
 6647 	mkdirp.m4.
 6648 	* lib/Automake/Makefile.inc (dist_perllib_DATA): Drop Configure_ac.pm.
 6649 
 6650 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6651 
 6652 	Merge branch 'dog-feeding' into maint
 6653 
 6654 	* dog-feeding:
 6655 	  build: move automake and aclocal in 'bin' subdir
 6656 	  build: break up monolithic Makefile.am in subdir-specific fragments
 6657 
 6658 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6659 
 6660 	Merge branch 'branch-1.13.2' into maint
 6661 
 6662 	* branch-1.13.2:
 6663 	  automake: typofix in comments: s/AC_CONFIG_HEADER/AC_CONFIG_HEADERS/
 6664 
 6665 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6666 
 6667 	automake: typofix in comments: s/AC_CONFIG_HEADER/AC_CONFIG_HEADERS/
 6668 
 6669 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6670 
 6671 	am: prefer a shorter idiom where possible
 6672 
 6673 	That is, prefer:
 6674 
 6675 	    test -f FILE || do_action
 6676 
 6677 	over:
 6678 
 6679 	    if test ! -f FILE; then do_action; else :; fi
 6680 
 6681 	* lib/am/remake-hdr.am (%CONFIG_H%): Here.
 6682 
 6683 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6684 
 6685 	Merge branch 'branch-1.13.2' into maint
 6686 
 6687 	* branch-1.13.2:
 6688 	  cosmetics: fix few typos, grammaros and missing whitespace
 6689 	  fixup: remove an obsolete comment
 6690 	  docs: we still don't have the promised better Java interface
 6691 
 6692 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6693 
 6694 	cosmetics: fix few typos, grammaros and missing whitespace
 6695 
 6696 	* lib/am/*.am: In comments in some of these files.
 6697 
 6698 2013-05-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6699 
 6700 	fixup: remove an obsolete comment
 6701 
 6702 	* lib/am/header-vars.am (am__make_running_with_option): Here.
 6703 
 6704 2013-05-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6705 
 6706 	docs: we still don't have the promised better Java interface
 6707 
 6708 	Reported by Michael Zucchi:
 6709 	<http://lists.gnu.org/archive/html/automake/2013-05/threads.html>
 6710 
 6711 	See also automake bug#9088.
 6712 
 6713 	* doc/automake.texi (Java): Adjust and clarify.
 6714 	* THANKS: Update.
 6715 
 6716 	Reported-by: Michael Zucchi <notzed@gmail.com>
 6717 
 6718 2013-05-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6719 
 6720 	Merge branch 'maint'
 6721 
 6722 	* maint:
 6723 	  maint: re-run "make update-copyright" ...
 6724 
 6725 2013-05-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6726 
 6727 	maint: re-run "make update-copyright" ...
 6728 
 6729 	* t/lex-multiple.sh: ... which updates the copyright years
 6730 	of this test (they were somehow not bumped in the past).
 6731 
 6732 2013-05-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6733 
 6734 	Merge branch 'maint'
 6735 
 6736 	* maint:
 6737 	  tests: avoid spurious failure with older flex (2.5.4)
 6738 	  announcement: cater to more flexible NEWS format
 6739 
 6740 2013-05-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6741 
 6742 	tests: avoid spurious failure with older flex (2.5.4)
 6743 
 6744 	That old version is unfortunately still relevant, being the one
 6745 	installed on NetBSD 5.1.
 6746 
 6747 	* t/lex-multiple.sh: Use the '-o' option rather than the longer
 6748 	equivalent '--outfile'.  The latter is not supported by older
 6749 	versions of flex (e.g., flex 2.5.4).
 6750 
 6751 2013-05-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6752 
 6753 	Merge branch 'branch-1.13.2' into maint
 6754 
 6755 	* branch-1.13.2:
 6756 	  announcement: cater to more flexible NEWS format
 6757 
 6758 2013-05-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6759 
 6760 	announcement: cater to more flexible NEWS format
 6761 
 6762 	* maint.mk (announcement): Here, be prepared to handle the case
 6763 	in which the first section of the NEWS file is dedicated to report
 6764 	future backward-incompatibilities and/or other warnings.
 6765 
 6766 2013-05-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6767 
 6768 	build: move automake and aclocal in 'bin' subdir
 6769 
 6770 	* automake.in: Rename ...
 6771 	* bin/automake.in: ... like this.
 6772 	* aclocal.in: Rename ...
 6773 	* bin/aclocal.in: ... like this.
 6774 	* Makefile.am: Move parts that dealt with the building/distribution
 6775 	of aclocal and Automake ..
 6776 	* bin/Makefile.inc): ... in this new included fragment.  Adjust as
 6777 	needed, and make deliberate use of the '%D%' substitution.
 6778 	* lib/gen-perl-protos: Move ...
 6779 	* bin/gen-perl-protos: ... here.
 6780 	* bootstrap.sh, configure.ac, maintainer/rename-tests,
 6781 	t/wrap/aclocal.in, t/wrap/automake.in, doc/Makefile.inc,
 6782 	t/ax/tap-setup.sh, .gitignore: Adjust.
 6783 	* maintainer/syntax-checks.mk: Likewise, and enhance a little.
 6784 
 6785 2013-05-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6786 
 6787 	build: break up monolithic Makefile.am in subdir-specific fragments
 6788 
 6789 	This is convenient to do, now that we have improved "relative directory"
 6790 	support with the '%reladir%' (a.k.a. '%D%') and '%canon_reladir%' (a.k.a.
 6791 	'%C%') Automake-time substitutions for included makefile fragments.
 6792 
 6793 	This move also satisfy our philosophy of using new Automake features in
 6794 	our own build system, as a way of facilitating early discovery of possible
 6795 	bugs or interface warts.
 6796 
 6797 	* Makefile.am: Break up ...
 6798 	* doc/Makefile.inc, lib/Automake/Makefile.inc, lib/Makefile.inc,
 6799 	lib/am/Makefile.inc, m4/Makefile.inc, t/Makefile.inc): ... in this
 6800 	new included fragments.  Adjust as needed, and make deliberate use
 6801 	of the '%D%' substitution.
 6802 	* contrib/t/local.am: Rename ...
 6803 	* contrib/t/Makefile.inc: ... like this.
 6804 
 6805 2013-05-06  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6806 
 6807 	docs: fix warnings given by Texinfo 5.x
 6808 
 6809 	These warnings:
 6810 
 6811 	  doc/automake.texi:3885: warning: node next `Public Macros' in menu \
 6812 	    `Private Macros' and in sectioning `Obsolete Macros' differ
 6813 	  doc/automake.texi:4050: warning: node `Private Macros' is next for \
 6814 	    `Obsolete Macros' in sectioning but not in menu
 6815 	  doc/automake.texi:4050: warning: node prev `Obsolete Macros' in menu \
 6816 	    `Private Macros' and in sectioning `Public Macros' differ
 6817 	  doc/automake.texi:4090: warning: node `Obsolete Macros' is next for \
 6818 	    `Private Macros' in menu but not in sectioning
 6819 	  doc/automake.texi:4090: warning: node prev `Private Macros' in menu \
 6820 	    `Public Macros' and in sectioning `Obsolete Macros' differ
 6821 
 6822 	* doc/automake.texi: Adjust.
 6823 
 6824 2013-05-04  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6825 
 6826 	fixlet: remove leftover comment fragment
 6827 
 6828 	* configure.ac: Here.
 6829 
 6830 2013-05-04  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6831 
 6832 	master branch: we are going to become Automake 2.0
 6833 
 6834 	* configure.ac (AC_INIT): So adjust the beta version in here, from
 6835 	1.13a to 1.99a.
 6836 	* m4/amversion.m4: Regenerate.
 6837 
 6838 2013-05-04  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6839 
 6840 	Merge branch 'maint'
 6841 
 6842 	* maint:
 6843 	  maint branch: we are going to become Automake 1.14
 6844 	  maint: version bump after beta release 1.13.1d
 6845 	  release: beta release 1.13.1d (will become 1.13.2)
 6846 	  NEWS: document more robust handling/recognition of make options
 6847 	  tests: typofixes in comments in t/preproc-c-compile.sh
 6848 	  tests: remove bashism from a test
 6849 
 6850 2013-05-04  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6851 
 6852 	maint branch: we are going to become Automake 1.14
 6853 
 6854 	* configure.ac (AC_INIT): So adjust beta version in here, from
 6855 	1.13.2a to 1.13a.
 6856 	* m4/amversion.m4: Regenerate.
 6857 
 6858 2013-05-04  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6859 
 6860 	Merge branch 'branch-1.13.2' into maint
 6861 
 6862 	* branch-1.13.2:
 6863 	  maint: version bump after beta release 1.13.1d
 6864 	  release: beta release 1.13.1d (will become 1.13.2)
 6865 	  NEWS: document more robust handling/recognition of make options
 6866 
 6867 2013-05-04  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6868 
 6869 	maint: version bump after beta release 1.13.1d
 6870 
 6871 	* configure.ac (AC_INIT): Bump version number to 1.13.1e.
 6872 	* m4/amversion.m4: Likewise (autoupdated by "make bootstrap").
 6873 
 6874 2013-05-04  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6875 
 6876 	release: beta release 1.13.1d (will become 1.13.2)
 6877 
 6878 	* configure.ac (AC_INIT): Bump version number to 1.13.1d.
 6879 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 6880 
 6881 2013-05-04  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6882 
 6883 	NEWS: document more robust handling/recognition of make options
 6884 
 6885 2013-05-04  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6886 
 6887 	tests: typofixes in comments in t/preproc-c-compile.sh
 6888 
 6889 2013-05-04  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6890 
 6891 	tests: remove bashism from a test
 6892 
 6893 	* t/preproc-c-compile.sh (Makefile.am): Use "test foo = bar", not the
 6894 	bash-specific "test foo == bar".
 6895 
 6896 2013-05-04  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6897 
 6898 	Merge branch 'maint'
 6899 
 6900 	* maint:
 6901 	  maint: targets and recipes to simplify testing on real-world packages
 6902 	  build: preparatory refactoring
 6903 	  build: tiny reduction in code duplication
 6904 	  make flags analysis: handle more options with args
 6905 	  make flags analysis: use simpler variable names
 6906 	  make flags analysis: whitespace changes
 6907 	  make flags analysis: embed in a subshell
 6908 	  make flags analysis: be more robust
 6909 	  make flags analysis: cater to GNU make 3.83 (still unreleased as of now)
 6910 	  tests: expose weaknesses in make flags analysis
 6911 	  tests: improve debugging output in checks on make flags analysis
 6912 	  make flags analysis: refactor, to reduce code duplication
 6913 	  tests: avoid one tricky use of "make -e"
 6914 	  cosmetics: remove few trailing whitespace occurrences
 6915 	  typofix: in NEWS (extra whitespace)
 6916 	  tests: avoid a spurious error with Solaris make
 6917 	  subdirs: don't return false positives for the '-k' option's presence
 6918 	  header-vars: recognize more make flags ('-k' in particular)
 6919 	  header-vars: simplify how make flags are determined
 6920 	  tests: remove dead code from t/make-dryrun.tap
 6921 	  header-vars: new variable $(am__running_with_option)
 6922 	  tests: expose bug#12554 (false positives for presence of '-k' make option)
 6923 
 6924 2013-05-04  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6925 
 6926 	Merge branch 'branch-1.13.2' into maint
 6927 
 6928 	* branch-1.13.2:
 6929 	  maint: targets and recipes to simplify testing on real-world packages
 6930 	  build: preparatory refactoring
 6931 	  build: tiny reduction in code duplication
 6932 	  make flags analysis: handle more options with args
 6933 	  make flags analysis: use simpler variable names
 6934 	  make flags analysis: whitespace changes
 6935 	  make flags analysis: embed in a subshell
 6936 	  make flags analysis: be more robust
 6937 	  make flags analysis: cater to GNU make 3.83 (still unreleased as of now)
 6938 	  tests: expose weaknesses in make flags analysis
 6939 	  tests: improve debugging output in checks on make flags analysis
 6940 	  make flags analysis: refactor, to reduce code duplication
 6941 	  tests: avoid one tricky use of "make -e"
 6942 	  tests: avoid a spurious error with Solaris make
 6943 	  subdirs: don't return false positives for the '-k' option's presence
 6944 	  header-vars: recognize more make flags ('-k' in particular)
 6945 	  header-vars: simplify how make flags are determined
 6946 	  tests: remove dead code from t/make-dryrun.tap
 6947 	  header-vars: new variable $(am__running_with_option)
 6948 	  tests: expose bug#12554 (false positives for presence of '-k' make option)
 6949 
 6950 2013-05-04  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6951 
 6952 	Merge 'better-makeflags-recognition' and 'testing-work' into branch-1.13.2
 6953 
 6954 	* better-makeflags-recognition:
 6955 	  make flags analysis: handle more options with args
 6956 	  make flags analysis: use simpler variable names
 6957 	  make flags analysis: whitespace changes
 6958 	  make flags analysis: embed in a subshell
 6959 	  make flags analysis: be more robust
 6960 	  make flags analysis: cater to GNU make 3.83 (still unreleased as of now)
 6961 	  tests: expose weaknesses in make flags analysis
 6962 	  tests: improve debugging output in checks on make flags analysis
 6963 	  make flags analysis: refactor, to reduce code duplication
 6964 
 6965 	* testing-work:
 6966 	  maint: targets and recipes to simplify testing on real-world packages
 6967 	  build: preparatory refactoring
 6968 	  build: tiny reduction in code duplication
 6969 
 6970 2013-05-04  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6971 
 6972 	maint: targets and recipes to simplify testing on real-world packages
 6973 
 6974 	We introduce a new section in the maintainer-specific makefile that
 6975 	contains recipes to test the build system of some well-known GNU
 6976 	packages with the current development version of Automake.  Not the
 6977 	cleanest way to do so, but good enough for the moment.  We'll revisit
 6978 	the matter after the 1.13.2 release (which we now hope will happen
 6979 	soon).
 6980 
 6981 	* maint.mk (git-sv-host): New.
 6982 	(SV_GIT_CF, SV_GIT_AC, SV_GIT_GL): Use it to reduce code duplication.
 6983 	(ALL_PACKAGES, FEW_PACKAGES): New, lists of GNU packages to try out.
 6984 	(ttp-check, ttp-check-all): New targets, do the checking with said
 6985 	packages.
 6986 	(ttp): New, alias for 'ttp-check'.
 6987 	(ttp-all): New, alias for 'ttp-check-all'.
 6988 
 6989 2013-05-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 6990 
 6991 	build: preparatory refactoring
 6992 
 6993 	No semantic change intended.  This change will be required
 6994 	by a later patch.
 6995 
 6996 	* Makefile.am (setup_autotools_paths): New.
 6997 	($(srcdir)/doc/amhello-1.0.tar.gz): Use it.
 6998 
 6999 2013-05-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7000 
 7001 	build: tiny reduction in code duplication
 7002 
 7003 	The code was only duplicated two times, but we are soon going to
 7004 	need a third occurrence, and that would be one to much.
 7005 
 7006 	* Makefile.am (extend_path): New.
 7007 	(update_mans): Use it instead of copying & pasting its contents.
 7008 	($(srcdir)/doc/amhello-1.0.tar.gz): Likewise, and minor related
 7009 	adjustments.
 7010 
 7011 2013-05-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7012 
 7013 	make flags analysis: handle more options with args
 7014 
 7015 	That is, not only -I, but also -O (for upcoming GNU make 3.83),
 7016 	-l (GNU make), -d, -E, -D, -m (BSD make), -J, -T (NetBSD make).
 7017 
 7018 	* lib/am/header-vars.am (am__make_running_with_option): Extend
 7019 	and adjust, both code and comments.
 7020 
 7021 2013-05-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7022 
 7023 	make flags analysis: use simpler variable names
 7024 
 7025 	* lib/am/header-vars.am (am__make_running_with_option): Here.  Now
 7026 	that we expect to be run in a subshell, we don't have to worry about
 7027 	being namespace-safe.  And '$foo' is much more pleasant to read than
 7028 	'$am__foo' -- and pleasant code tends to be more correct.
 7029 	(am__make_dryrun, am__make_keepgoing): Adjust.
 7030 
 7031 2013-05-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7032 
 7033 	make flags analysis: whitespace changes
 7034 
 7035 	* lib/am/header-vars.am (am__make_running_with_option): Here.
 7036 	No semantic change is intended.
 7037 
 7038 2013-05-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7039 
 7040 	make flags analysis: embed in a subshell
 7041 
 7042 	So that we won't have to worry about leaking temporary variables,
 7043 	and similar stuff.
 7044 
 7045 	* lib/am/header-vars.am (am__make_dryrun, am__make_keepgoing): Here.
 7046 	(am__make_running_with_option): Minor adjustments.
 7047 
 7048 2013-05-02  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7049 
 7050 	make flags analysis: be more robust
 7051 
 7052 	In particular, catering to option-with-argument bundled together
 7053 	with options-without-arguments, as in:
 7054 
 7055 	   # With GNU make 3.82:
 7056 	   $ make -f- <<<'all:; echo "$$MFLAGS"' -s -k -I none
 7057 	   -skiI none
 7058 
 7059 	   # With development version of GNU make (Git commit b5ea49b):
 7060 	   $ make -f- <<<'all:; @echo "$$MFLAGS"' -I none
 7061 	   -skiI none
 7062 
 7063 	This fixes some lingering failures in the testsuite; precisely, in
 7064 	tests 'make-dryrun.tap' and 'make-keepgoing.tap'.
 7065 
 7066 	* t/header-vars.am (am__make_running_with_option): Enhance.
 7067 
 7068 2013-05-02  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7069 
 7070 	make flags analysis: cater to GNU make 3.83 (still unreleased as of now)
 7071 
 7072 	The current development version of GNU make (that is planned to become
 7073 	GNU make 3.83, sooner or later) has changed the format its $(MFLAGS)
 7074 	variable slightly, removing the space between an option and its argument:
 7075 
 7076 	   # With GNU make 3.82, compiled from official tarball:
 7077 	   $ make -f- <<<'all:; @echo "$$MFLAGS"' -I none
 7078 	   -I none
 7079 
 7080 	   # With development version of GNU make (Git commit b5ea49b):
 7081 	   $ make -f- <<<'all:; @echo "$$MFLAGS"' -I none
 7082 	   -Inone
 7083 
 7084 	This was done on purpose, in order to support more easily the new
 7085 	option '-O', which takes an optional argument; see:
 7086 
 7087 	  <http://lists.gnu.org/archive/html/bug-make/2013-05/msg00001.html>
 7088 
 7089 	So Just adapt to the new format as well.
 7090 
 7091 	* t/header-vars.am (am__make_running_with_option): Adjust to cater
 7092 	to the new GNU make behaviour.
 7093 
 7094 2013-05-01  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7095 
 7096 	tests: expose weaknesses in make flags analysis
 7097 
 7098 	This introduces some failures in the testsuite, that will be fixed
 7099 	soon enough by follow-up patches.
 7100 
 7101 	* t/make-dryrun.tap: Enhance and extend.
 7102 	* t/make-keepgoing.tap: Likewise.
 7103 
 7104 2013-05-01  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7105 
 7106 	tests: improve debugging output in checks on make flags analysis
 7107 
 7108 	* t/make-dryrun.tap: Here.
 7109 	* t/make-keepgoing.tap: And here.
 7110 
 7111 2013-05-01  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7112 
 7113 	make flags analysis: refactor, to reduce code duplication
 7114 
 7115 	And make it a little more reliable.
 7116 
 7117 	* t/header-vars.am (am__make_running_with_option): Here.
 7118 
 7119 2013-05-01  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7120 
 7121 	tests: avoid one tricky use of "make -e"
 7122 
 7123 	Which was causing a spurious failure on FreeBSD.  Not particularly
 7124 	surprising, given how brittle "make -e" is in general ...
 7125 
 7126 	* t/cxx-lt-demo.sh: Instead of forcing $(CC) to be 'false' by
 7127 	exporting "CC=false" in the environment and then passing the '-e'
 7128 	option to make, do so by passing "CC=false" on the make command
 7129 	line, both directly and using AM_MAKEFLAGS.
 7130 
 7131 2013-04-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7132 
 7133 	Merge branch 'fix-pr12554' into branch-1.13.2
 7134 
 7135 	* fix-pr12554:
 7136 	  tests: avoid a spurious error with Solaris make
 7137 	  subdirs: don't return false positives for the '-k' option's presence
 7138 	  header-vars: recognize more make flags ('-k' in particular)
 7139 	  header-vars: simplify how make flags are determined
 7140 	  tests: remove dead code from t/make-dryrun.tap
 7141 	  header-vars: new variable $(am__running_with_option)
 7142 	  tests: expose bug#12554 (false positives for presence of '-k' make option)
 7143 
 7144 2013-04-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7145 
 7146 	Merge branch 'branch-1.13.2' into maint
 7147 
 7148 	* branch-1.13.2:
 7149 	  cosmetics: remove few trailing whitespace occurrences
 7150 	  typofix: in NEWS (extra whitespace)
 7151 
 7152 2013-04-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7153 
 7154 	cosmetics: remove few trailing whitespace occurrences
 7155 
 7156 2013-04-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7157 
 7158 	typofix: in NEWS (extra whitespace)
 7159 
 7160 2013-04-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7161 
 7162 	compat: substitute '@mkdir_p@' as '$(MKDIR_P)' unconditionally
 7163 
 7164 	We had started to do so only when the obsolete AM_PROG_MKDIR_P variable
 7165 	was used, but that has recently turned out not to be not good enough,
 7166 	because Gettext 0.18.2 (the latest one at the moment of writing) is using
 7167 	the modern macro AC_PROG_MKDIR_P in its '*.m4' files, but at the same time
 7168 	is still using the obsolete make variable '$(mkdir_p)' in its Makefile
 7169 	fragments.  This has been causing failures in real-world packages
 7170 	bootstrapped with latest Gettext (0.18.2) and with bleeding-edge automake.
 7171 
 7172 	See for example:
 7173 
 7174 	  <http://lists.gnu.org/archive/html/bug-tar/2013-04/msg00009.html>
 7175 
 7176 	And the worst things is that those failures have been basically silent
 7177 	and unnoticeable up until the final "make install" --- that is, after
 7178 	the affected package had been bootstrapped, configured and built.  Yikes!
 7179 
 7180 	This issue has been reported to the Gettext developers:
 7181 
 7182 	  <http://lists.gnu.org/archive/html/bug-gettext/2013-04/msg00044.html>
 7183 
 7184 	But even if they will quickly fix it, Gettext 0.18.2 will remain
 7185 	affected.
 7186 
 7187 	The fix we can implement in Automake is small and unobtrusive enough,
 7188 	definitely worth having in order to save the users and developers from
 7189 	having to deal with this hard-to-spot issue.
 7190 
 7191 	* m4/init.m4 (AM_INIT_AUTOMAKE): AC_SUBST '@mkdir_p@' to '$(MKDIR_P)'.
 7192 	* t/am-prog-mkdir-p.sh: Enhance.
 7193 
 7194 2013-04-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7195 
 7196 	Merge branch 'maint'
 7197 
 7198 	* maint:
 7199 	  sync: update files from upstream with "make fetch"
 7200 	  maintcheck: remove outdated whitelisting
 7201 	  tar: format 'ustar' cannot support UID/GID longer than 21 bits
 7202 	  docs: issues with configure substitutions in TESTS
 7203 	  tests: avoid possible autotools caching issues (automake bug#13832)
 7204 	  tests: rename some with more descriptive names
 7205 	  docs: add myself and Ralf Wildenhues as authors
 7206 	  authors: add myself
 7207 	  dry-run: don't get confused by '-I' option
 7208 	  tests: avoid a spurious failure with the Korn Shell
 7209 	  dry-run: with GNU make, prefer $(MFLAGS) over $(MAKEFLAGS)
 7210 	  header vars: can determine whether we are running under GNU make
 7211 	  NEWS: improve wording for automake bug#13514 fix
 7212 	  NEWS: document fix for automake bug#13514
 7213 
 7214 2013-04-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7215 
 7216 	Merge branch 'branch-1.13.2' into maint
 7217 
 7218 	* branch-1.13.2:
 7219 	  sync: update files from upstream with "make fetch"
 7220 	  maintcheck: remove outdated whitelisting
 7221 	  tar: format 'ustar' cannot support UID/GID longer than 21 bits
 7222 
 7223 2013-04-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7224 
 7225 	Merge branch 'fix-pr13588-pax-hangs' into branch-1.13.2
 7226 
 7227 	* fix-pr13588-pax-hangs:
 7228 	  tar: format 'ustar' cannot support UID/GID longer than 21 bits
 7229 
 7230 2013-04-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7231 
 7232 	tests: avoid a spurious error with Solaris make
 7233 
 7234 	* t/make-keepgoing.tap (Makefile.am): Here.
 7235 
 7236 2013-04-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7237 
 7238 	subdirs: don't return false positives for the '-k' option's presence
 7239 
 7240 	This change fixes automake bug#12554.
 7241 
 7242 	The old implementation of the code descending into $(SUBDIRS)
 7243 	entries used the following snippet to decide whether make is running
 7244 	with the '-k' a.k.a. '--keep-going' option, and thus whether a failure
 7245 	in a subdirectory should prevent the descent in the following ones:
 7246 
 7247 	    fail= failcom='exit 1'; \
 7248 	    for f in x $$MAKEFLAGS; do \
 7249 	      case $$f in \
 7250 	        *=* | --[!k]*);; \
 7251 	        *k*) failcom='fail=yes';; \
 7252 	      esac; \
 7253 	    done
 7254 
 7255 	It's clear that the second pattern in the 'case' construct could possibly
 7256 	match false positives, for examples in these two cases:
 7257 
 7258 	    make check TESTS="x.test k.test"
 7259 	    make -I /usr/local/kool-fragments
 7260 
 7261 	which are somewhat unusual, but not invalid.  So we need a more resilient
 7262 	implementation, as we did for the detection of the '-n' flag.
 7263 
 7264 	This implementation is now provided by the new private macro
 7265 	'$(am__make_keepgoing)' (introduced in recent commits); so we can
 7266 	just us that to fix the bug.
 7267 
 7268 	* lib/am/subdirs.am ($(am__recursive_targets)): Use '$(am__make_keepgoing)'
 7269 	instead of ad-hoc and more brittle checks.
 7270 	* t/list-of-tests.mk (XFAIL_TESTS): Remove the now-passing test case
 7271 	't/subdir-keep-going-pr12554.sh'.
 7272 
 7273 	Reported-by: Michael Daniels <mdaniels@rim.com>
 7274 
 7275 2013-04-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7276 
 7277 	header-vars: recognize more make flags ('-k' in particular)
 7278 
 7279 	* lib/am/header-vars.am (am__running_with_option): Here.
 7280 	Few improvements to comments, while at it.
 7281 	(am__make_keepgoing): New, tell whther make is being runt with
 7282 	the '-k' option.
 7283 	* t/make-keepgoing.tap: New test.
 7284 	* t/list-of-tests.mk: Add it.
 7285 	* t/make-dryrun.tap: Minor edits to keep it more in sync with
 7286 	the new test.
 7287 
 7288 2013-04-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7289 
 7290 	sync: update files from upstream with "make fetch"
 7291 
 7292 	* lib/config.guess: Likewise.
 7293 	* lib/config.sub: Likewise.
 7294 
 7295 2013-04-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7296 
 7297 	maintcheck: remove outdated whitelisting
 7298 
 7299 	* syntax-checks.mk (sc_tests_overriding_macros_on_cmdline): Here.
 7300 	The test 'make-dryrun.sh' has been since long rewritten as the TAP
 7301 	test 'make-dryrun.tap', and no longer spuriously triggers this
 7302 	maintainer check.
 7303 
 7304 2013-04-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7305 
 7306 	header-vars: simplify how make flags are determined
 7307 
 7308 	Actually, son far only the '-n' option ("dry mode") was detected,
 7309 	but this change will allow us to soon detect more options.
 7310 
 7311 	* lib/am/header-vars.am (am__running_with_option): Even when $MAKEFLAGS
 7312 	appears to contain definition of variables with embedded whitespace,
 7313 	use simple textual pre-processing over $MAKEFLAGS rather than tricky
 7314 	recursive invocations of make to determine whether the '-n' option was
 7315 	given.  This is enough to correctly handle all the tricky usages covered
 7316 	in the testsuite.
 7317 	* t/nodep.sh: Adjust to avoid a spurious failure.
 7318 
 7319 2013-04-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7320 
 7321 	tests: remove dead code from t/make-dryrun.tap
 7322 
 7323 2013-04-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7324 
 7325 	header-vars: new variable $(am__running_with_option)
 7326 
 7327 	This is a preparatory refactoring, needed by later patches.
 7328 	No semantic change is intended.
 7329 
 7330 	* lib/am/header-vars.am (am__running_with_option): New, contains
 7331 	shell code that determines whether the current make instance is
 7332 	running with a given one-letter option (e.g., -k, -n) that takes
 7333 	no argument.  Actually, the only supported option at the moment
 7334 	is '-n' (support for '-k' will be added soon).
 7335 	(am__make_dryrun): Rewrite as a thin wrapper around
 7336 	'$(am__make_running_with_option)'.
 7337 
 7338 2013-04-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7339 
 7340 	tests: expose bug#12554 (false positives for presence of '-k' make option)
 7341 
 7342 	The current implementation of the code descending into $(SUBDIRS)
 7343 	entries uses the following snippet to decide whether make is running
 7344 	with the '-k' a.k.a. '--keep-going' option, and thus whether a failure
 7345 	in a subdirectory should prevent the descent in the following ones:
 7346 
 7347 	    fail= failcom='exit 1'; \
 7348 	    for f in x $$MAKEFLAGS; do \
 7349 	      case $$f in \
 7350 	        *=* | --[!k]*);; \
 7351 	        *k*) failcom='fail=yes';; \
 7352 	      esac; \
 7353 	    done
 7354 
 7355 	It's clear that the second pattern in the 'case' construct can possibly
 7356 	match false positives, for examples in these two cases:
 7357 
 7358 	    make check TESTS="x.test k.test"
 7359 	    make -I /usr/local/kool-fragments
 7360 
 7361 	which are somewhat unusual, but not invalid.  So we need a more resilient
 7362 	implementation, as we did for the detection of the '-n' flag.
 7363 
 7364 	But alas, such an implementation seems quite tricky to obtain in portable
 7365 	make.  So for the moment we content ourselves with exposing the bug, with
 7366 	the hope of being able to fix soon enough.
 7367 
 7368 	* t/subdir-keep-going-pr12554.sh: New test.
 7369 	* t/list-of-tests.mk (handwritten_TESTS, XFAIL_TESTS): Add it.
 7370 	* THANKS: Update
 7371 
 7372 	Reported-by: Michael Daniels <mdaniels@rim.com>
 7373 
 7374 2013-04-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7375 
 7376 	tar: format 'ustar' cannot support UID/GID longer than 21 bits
 7377 
 7378 	See automake bug#8343 and bug#13588.
 7379 
 7380 	POSIX 1988 'ustar' format is defined with *fixed-size* fields. There
 7381 	is notably a 21 bits limit (2097151) for the UID and the GID.
 7382 
 7383 	Tom Rini tom_rini@mentor.com says (in bug#8343):
 7384 
 7385 	    When the user has a UID or GID that is larger than the ustar format
 7386 	    supports, pax does not error out gracefully in some cases (FC13).
 7387 
 7388 	Marc Herbert <marc.herbert@intel.com> adds (in bug#8343):
 7389 
 7390 	    When "configure" is run by a user with an UID bigger than 21 bits,
 7391 	    BSD pax 3.4 aborts when trying to create the 'conftest.tar' test
 7392 	    archive and leaves an empty or corrupted conftest.tar file behind.
 7393 	    In the next step, pax tries to extract this incomplete or corrupted
 7394 	    archive and this *** hangs the whole ./configure script ***.
 7395 
 7396 	    Note: GNU cpio 2.9 pretends to pass the test but it is a LIE: it
 7397 	    silently truncates any big UID to its lower 21 bits. I don't know
 7398 	    what can be the consequences of this lie.
 7399 
 7400 	    I think there is currently a design issue in automake/m4/tar.m4
 7401 	    considering that a ustar archive should should *never* succeed when
 7402 	    ./configure is run from a big user ID.
 7403 
 7404 	Months later, Petr Hracek <phracek@redhat.com> reports a similar issue
 7405 	(in bug#13588) for Fedora 17:
 7406 
 7407 	    I am trying to solve problem in case a user is created with big
 7408 	    UID and during configuration pax hangs with message
 7409 
 7410 	        ATTENTION! pax archive volume change required.
 7411 	        Ready for archive volume: 1
 7412 	        Input archive name or "." to quit pax.
 7413 	        Archive name >
 7414 
 7415 	    and needs user interaction.
 7416 
 7417 	    Reference: <https://bugzilla.redhat.com/show_bug.cgi?id=843376>
 7418 
 7419 	Time to fix this issue, on the line of a preliminary patch provided by
 7420 	Petr Hracek in bug#13588.  The final patch ended up being remarkably
 7421 	different from that original proposition, though.
 7422 
 7423 	* m4/tar.m4 (_AM_PROG_TAR): If the UID or GID of the current user is
 7424 	too high (> 2097151), the 'ustar' format cannot work.  Adjust checks
 7425 	accordingly.  Some related code reordering and clean-up.
 7426 	* t/tar-ustar-id-too-high.sh: New test.
 7427 	* t/list-of-tests.mk: Add it.
 7428 	* t/tar.sh: While at it, tweak and enhance a little.
 7429 	* t/tar2.sh: Likewise.
 7430 	* t/tar3.sh: Likewise.
 7431 	* t/tar-override.sh: Likewise.
 7432 	* NEWS: Update.
 7433 	* THANKS: Likewise.
 7434 
 7435 	Helped-by: Pavel Raiskup <praiskup@redhat.com>
 7436 	Helped-by: Petr Hracek <phracek@redhat.com>
 7437 	Helped-by: Marc Herbert <marc.herbert@intel.com>
 7438 	Helped-by: Tom Rini <tom_rini@mentor.com>
 7439 
 7440 2013-04-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7441 
 7442 	Merge branch 'branch-1.13.2' into maint
 7443 
 7444 	* branch-1.13.2:
 7445 	  docs: issues with configure substitutions in TESTS
 7446 	  tests: avoid possible autotools caching issues (automake bug#13832)
 7447 	  docs: add myself and Ralf Wildenhues as authors
 7448 	  authors: add myself
 7449 	  dry-run: don't get confused by '-I' option
 7450 	  tests: avoid a spurious failure with the Korn Shell
 7451 	  dry-run: with GNU make, prefer $(MFLAGS) over $(MAKEFLAGS)
 7452 	  header vars: can determine whether we are running under GNU make
 7453 	  NEWS: improve wording for automake bug#13514 fix
 7454 	  NEWS: document fix for automake bug#13514
 7455 
 7456 2013-04-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7457 
 7458 	Merge few bug-fixing branches into branch-1.13.2
 7459 
 7460 	* fix-part-pr13832:
 7461 	  tests: avoid possible autotools caching issues (automake bug#13832)
 7462 
 7463 	* fix-pr13760:
 7464 	  dry-run: don't get confused by '-I' option
 7465 	  dry-run: with GNU make, prefer $(MFLAGS) over $(MAKEFLAGS)
 7466 	  header vars: can determine whether we are running under GNU make
 7467 
 7468 	* fix-doc-pr14019:
 7469 	  docs: issues with configure substitutions in TESTS
 7470 
 7471 	* news-wording-improve:
 7472 	  NEWS: improve wording for automake bug#13514 fix
 7473 
 7474 2013-04-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7475 
 7476 	docs: issues with configure substitutions in TESTS
 7477 
 7478 	Motivated by automake bug#14019.
 7479 
 7480 	* doc/automake.texi: Currently, when the parallel test harness is in use,
 7481 	configure substitutions in TESTS definitions can only work if they expand
 7482 	to tests that ends with a suffix listed in TEST_EXTENSIONS.  Document this
 7483 	limitation.
 7484 
 7485 2013-04-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7486 
 7487 	tests: avoid possible autotools caching issues (automake bug#13832)
 7488 
 7489 	* t/ar-lib4.sh: Here, by removing the autom4te cache before editing
 7490 	configure.ac and re-running the autotools.
 7491 
 7492 2013-04-25  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7493 
 7494 	tests: rename some with more descriptive names
 7495 
 7496 	* t/tar3.sh: Rename ...
 7497 	* t/tar-opts-errors.sh: ... like this.
 7498 	* t/tar2.sh: Rename...
 7499 	* t/tar-pax.sh: ... like this.
 7500 	* t/tar.sh: Rename ...
 7501 	* t/tar-ustar.sh: ... like this.
 7502 	* t/list-of-tests.mk: Adjust.
 7503 
 7504 2013-04-25  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7505 
 7506 	docs: add myself and Ralf Wildenhues as authors
 7507 
 7508 	* doc/automake.texi: Here.
 7509 
 7510 2013-04-25  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7511 
 7512 	authors: add myself
 7513 
 7514 2013-04-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7515 
 7516 	dry-run: don't get confused by '-I' option
 7517 
 7518 	Fixes automake bug#13760 for non-GNU make implementations that still
 7519 	support the option '-I'.  So far, the only such make implementation
 7520 	are FreeBSD (8.x) make and NetBSD (5.x) make.
 7521 
 7522 	* lib/am/header-vars.am (am__make_dryrun): If a non-GNU make is being
 7523 	used, try to handle the '-I' option in $MAKEFLAGS correctly.  For GNU
 7524 	make, that is already done by the proper use of the $MFLAGS variable.
 7525 
 7526 2013-04-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7527 
 7528 	tests: avoid a spurious failure with the Korn Shell
 7529 
 7530 	* t/list-flags.sh: Ensure the EMACS variable is set before unsetting it.
 7531 
 7532 2013-04-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7533 
 7534 	dry-run: with GNU make, prefer $(MFLAGS) over $(MAKEFLAGS)
 7535 
 7536 	Fixes automake bug#13760 for GNU make.
 7537 
 7538 	* lib/am/header-vars.am (am__make_dryrun): If GNU make is being used, rely
 7539 	on the contents of the $(MFLAGS) variable rather than of the $(MAKEFLAGS)
 7540 	to decide whther make is being executed in "dry run" mode.  Not only this
 7541 	makes the code possibly faster and less brittle, but also fixes automake
 7542 	bug#13760 (at least when GNU make is in use).
 7543 	* t/make-dryrun.tap: Adjust: some tests that were xfailing now pass.
 7544 
 7545 2013-04-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7546 
 7547 	header vars: can determine whether we are running under GNU make
 7548 
 7549 	This is mostly a preparatory patch in view of future changes.
 7550 
 7551 	* lib/am/header-vars.am (am__is_gnu_make): New, contains shell code that
 7552 	determines whether we are running under GNU make.
 7553 	* t/make-is-gnu.sh: New test.
 7554 	* t/list-of-tests.mk: Add it.
 7555 
 7556 2013-04-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7557 
 7558 	NEWS: improve wording for automake bug#13514 fix
 7559 
 7560 	Helped-by: Jack Kelly <jack@jackkelly.name>
 7561 
 7562 2013-04-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7563 
 7564 	NEWS: document fix for automake bug#13514
 7565 
 7566 	It was fixed by commit v1.13.1b-5-g7a3d7ce.
 7567 
 7568 2013-04-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7569 
 7570 	Merge branch 'maint'
 7571 
 7572 	* maint:
 7573 	  INSTALL: regen
 7574 	  sync: update files from upstream with "make fetch"
 7575 	  tests: tweaks and minimal improvements to t/insthook.sh
 7576 	  typofix: in comments in t/extra2.sh
 7577 	  tests: fix botched cross-reference in a heading comment
 7578 
 7579 2013-04-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7580 
 7581 	Merge branch 'branch-1.13.2' into maint
 7582 
 7583 	* branch-1.13.2:
 7584 	  INSTALL: regen
 7585 	  sync: update files from upstream with "make fetch"
 7586 	  tests: tweaks and minimal improvements to t/insthook.sh
 7587 
 7588 2013-04-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7589 
 7590 	INSTALL: regen
 7591 
 7592 2013-04-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7593 
 7594 	sync: update files from upstream with "make fetch"
 7595 
 7596 	* lib/INSTALL: Update.
 7597 	* lib/config.guess: Likewise.
 7598 	* lib/config.sub: Likewise.
 7599 	* lib/gendocs.sh: Likewise.
 7600 	* lib/gendocs_template: Likewise.
 7601 	* lib/gitlog-to-changelog: Likewise.
 7602 	* lib/gnupload: Likewise.
 7603 	* lib/texinfo.tex: Likewise.
 7604 	* lib/update-copyright: Likewise.
 7605 
 7606 2013-04-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7607 
 7608 	tests: tweaks and minimal improvements to t/insthook.sh
 7609 
 7610 2013-04-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7611 
 7612 	typofix: in comments in t/extra2.sh
 7613 
 7614 2013-04-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7615 
 7616 	tests: fix botched cross-reference in a heading comment
 7617 
 7618 	* t/extra-dist-wildcards.sh: Here.
 7619 
 7620 2013-04-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7621 
 7622 	compat: reinstate support for obsolete $(INCLUDES) variable
 7623 
 7624 	Since INCLUDES and AM_CPPFLAGS can be both used at the same time in a
 7625 	single project, and interact in a non-obvious way, switching away from
 7626 	INCLUDES is, unfortunately, not simply a matter of doing a mindless
 7627 	INCLUDES -> AM_CPPFLAGS substitution.  In fact, the transition from a
 7628 	"INCLUDES + CPPFLAGS" setup to a "AM_CPPFLAGS only" setup could cause
 7629 	subtle bugs to sneak into the build systems of the client packages
 7630 	doing the transition.  Forcing this risk upon our users is definitely
 7631 	not worth the small enhancement in consistency and cleanliness that
 7632 	the removal of INCLUDES would give us.
 7633 
 7634 	For a more extended rationale and related discussions, see:
 7635 	<http://lists.gnu.org/archive/html/automake/2013-02/msg00001.html>
 7636 
 7637 	This change basically reverts commit v1.13-8-g880595d.
 7638 
 7639 	* automake.in: Re-introduce support for INCLUDES (its use will cause
 7640 	warnings in the 'obsolete' category rather than fatal errors).
 7641 	* doc/automake.texi: Adjust to re-add mention of INCLUDES.
 7642 	* t/includes-no-more.sh: Rename ...
 7643 	* t/includes-deprecation.sh: ... like this, and adjust.
 7644 	* t/list-of-tests.mk: Update.
 7645 	* t/warnopts.sh: Minor fixlet while at it.
 7646 
 7647 	Suggested-by: Peter Rosin <peda@lysator.liu.se>
 7648 
 7649 2013-04-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7650 
 7651 	Merge branch 'maint'
 7652 
 7653 	* maint:
 7654 	  automake: refactoring: factor out common cpp-like flags
 7655 
 7656 2013-04-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7657 
 7658 	automake: refactoring: factor out common cpp-like flags
 7659 
 7660 	* automake.in (@cpplike_flags): In this new variable...
 7661 	(C, C++, Objective C, Objective C++, Unified Parallel C, Preprocessed
 7662 	Assembler, Preprocessed Fortran, Preprocessed Fortran 77): ... to be
 7663 	used by registration (with the 'register_language' subroutine) of these
 7664 	languages.
 7665 
 7666 	This is a refactoring meant to simplify future changes; no semantic
 7667 	change is intended.
 7668 
 7669 2013-04-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7670 
 7671 	Merge branch 'maint'
 7672 
 7673 	* maint:
 7674 	  NEWS (mint): reflect new Automake versioning scheme
 7675 
 7676 2013-04-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7677 
 7678 	NEWS (mint): reflect new Automake versioning scheme
 7679 
 7680 	The next minor Automake version will be 1.14, and *not* 1.13.2 -- that
 7681 	will be the next bug-fixing version.
 7682 
 7683 2013-04-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7684 
 7685 	NEWS (master): reflect new Automake versioning scheme
 7686 
 7687 	The next minor Automake version will be 1.14 (not 1.13.2 -- that will
 7688 	be the next bug-fixing version).  And the next major Automake version
 7689 	will be 2.0.
 7690 
 7691 2013-04-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7692 
 7693 	Merge branch 'maint'
 7694 
 7695 	* maint:
 7696 	  Add missing '$' for variable expansion in depout.m4
 7697 	  maintcheck: avoid spurious failure
 7698 	  perl: perl subroutine prototypes are problematic, don't use them
 7699 	  fixup: one stray reference to older versioning scheme
 7700 
 7701 2013-04-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7702 
 7703 	Merge branch 'branch-1.13.2' into maint
 7704 
 7705 	* branch-1.13.2:
 7706 	  Add missing '$' for variable expansion in depout.m4
 7707 	  fixup: one stray reference to older versioning scheme
 7708 
 7709 2013-04-19  Gavin Smith  <gavinsmith0123@gmail.com>  (tiny change)
 7710 
 7711 	Add missing '$' for variable expansion in depout.m4
 7712 
 7713 	* m4/depout.m4: "am__include" appeared where "$am__include" was meant.
 7714 
 7715 	Reference:
 7716 	<http://lists.gnu.org/archive/html/automake-patches/2013-04/msg00000.html>
 7717 
 7718 2013-04-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7719 
 7720 	Merge branch 'preproc' into maint
 7721 
 7722 	* preproc:
 7723 	  maintcheck: avoid spurious failure
 7724 
 7725 2013-04-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7726 
 7727 	maintcheck: avoid spurious failure
 7728 
 7729 	* t/preproc-errmsg.sh: Here, breaking up a sed command to avoid spuriously
 7730 	triggering a failure in the 'sc_tests_logs_duplicate_prefixes' maintainer
 7731 	check.
 7732 
 7733 2013-04-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7734 
 7735 	Merge branch 'no-perl-protos' into maint
 7736 
 7737 	* no-perl-protos:
 7738 	  perl: perl subroutine prototypes are problematic, don't use them
 7739 
 7740 2013-04-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7741 
 7742 	perl: perl subroutine prototypes are problematic, don't use them
 7743 
 7744 	Basically, in perl, "subroutine prototypes" are not prototypes at all;
 7745 	rather, they are a trick to allow user-defined subroutines that behave
 7746 	like perl built-in functions.  For example, prototyped subroutines can
 7747 	be called without parentheses, and can impose context on their arguments.
 7748 
 7749 	Such semantics can be useful in some selected situations, but might also
 7750 	easily cause unexpected and harmful behaviours and side effects if we
 7751 	try to use perl prototypes as we would use C prototypes.
 7752 
 7753 	See the excellent article "Far More than Everything You've Ever Wanted
 7754 	to Know about Prototypes in Perl" by Tom Christiansen for more detailed
 7755 	information:
 7756 
 7757 	    <http://www.perlmonks.org/?node_id=861966>
 7758 	    <http://web.archive.org/web/20080421062920/\
 7759 	     library.n0i.net/programming/perl/articles/fm_prototypes>
 7760 
 7761 	It is important to note that modern perl allows a non-predeclared
 7762 	subroutine to be called without the '&' character, as long as its
 7763 	call uses proper parentheses:
 7764 
 7765 	    foo 'str', 2;   # will trigger errors if foo is not predeclared
 7766 	    foo('str', 2);  # ok even if foo is not predeclared
 7767 	    &foo('str', 2); # ditto; but the '&' is old-style and redundant
 7768 
 7769 	Note also that the prototype indicating "no argument":
 7770 
 7771 	    sub func() { ... }
 7772 
 7773 	can actually be useful, and has no discernible downsides, so we'll
 7774 	keep using it where it makes sense.
 7775 
 7776 	Also, in few, selected cases, we *want* to have subroutines behave like
 7777 	perl builtins (e.g., we want the 'append_exeext' function to be able
 7778 	to take a code block as first argument).  In such cases, we will of
 7779 	course continue to make use of perl subroutine prototypes.
 7780 
 7781 	Let's finally see an example that might clarify the kind of problems the
 7782 	use of subroutine prototypes in perl can cause.  This is just scratching
 7783 	the surface; there are several other aspects, typically subtler and more
 7784 	dangerous, that are not touched here.
 7785 
 7786 	If you have the prototyped subroutine definition:
 7787 
 7788 	    sub foo ($@)
 7789 	    {
 7790 	        my $s = shift;
 7791 	        print "SCALAR: $s\n";
 7792 	        print "ARRAY: @_\n";
 7793 	    }
 7794 
 7795 	and call 'foo' in code like:
 7796 
 7797 	    @list = (-1, 0, 1);
 7798 	    foo(@list);
 7799 
 7800 	you won't get a compile-time nor a runtime error (as a naive interpretation
 7801 	of the "prototype" characterization would let you think).  Rather, the
 7802 	prototype will cause the array '@list' will be coerced into scalar context
 7803 	before being passed too 'foo', which means that its *length* (3) will be
 7804 	passed to 'foo' as first argument; and since no further arguments are
 7805 	present after '@list', that *void* will be coerced to an empty list before
 7806 	being passed to 'foo'.
 7807 
 7808 	So code above will have the result of printing:
 7809 
 7810 	  SCALAR: 3
 7811 	  ARRAY:
 7812 
 7813 	Quite tricky, and definitely a behaviour we don't want to rely on.
 7814 
 7815 	* automake.in: Delete most subroutine prototypes.  Fix few of the
 7816 	remaining ones.  Related minor simplifications and adjustments.
 7817 	* lib/gen-perl-protos: Adjust.
 7818 
 7819 2013-03-05  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7820 
 7821 	Merge branch 'maint'
 7822 
 7823 	* maint:
 7824 	  preproc: enhance and extend tests
 7825 	  preproc: add support for relative names in included fragments
 7826 
 7827 2013-03-05  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7828 
 7829 	Merge branch 'preproc' into maint
 7830 
 7831 	* preproc:
 7832 	  preproc: enhance and extend tests
 7833 	  preproc: add support for relative names in included fragments
 7834 
 7835 2013-02-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7836 
 7837 	fixup: one stray reference to older versioning scheme
 7838 
 7839 	* lib/Automake/Options.pm: Here, in a FIXME comment: reference
 7840 	"Automake 3.0" rather than "Automake 1.15".
 7841 
 7842 2013-02-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7843 
 7844 	Merge branch 'maint' into master
 7845 
 7846 	* maint:
 7847 	  maint: more adjustments to the new versioning scheme
 7848 	  aclocal: fix for more-than-once specified directories
 7849 	  aclocal: just warn if the primary local m4 dir doesn't exist (don't error)
 7850 	  coverage: expose automake bug#13760
 7851 	  tests: refactor/enhance tests about make dry-run mode
 7852 	  maint: describe new versioning and branching scheme, and adjust to it
 7853 	  cosmetics: fix some "docstring-like" comments in automake
 7854 
 7855 2013-02-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7856 
 7857 	Merge branch 'branch-1.13.2' into maint
 7858 
 7859 	* branch-1.13.2:
 7860 	  aclocal: fix for more-than-once specified directories
 7861 	  aclocal: just warn if the primary local m4 dir doesn't exist (don't error)
 7862 
 7863 2013-02-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7864 
 7865 	maint: more adjustments to the new versioning scheme
 7866 
 7867 	This is a follow-up to commit 'v1.13.1b-11-g97aaf12'.
 7868 
 7869 	* automake.in: Adjust a comment.
 7870 	* PLANS: Adjust several files in here.
 7871 
 7872 2013-02-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7873 
 7874 	Merge branch 'fix-pr13514' into branch-1.13.2
 7875 
 7876 	* fix-pr13514:
 7877 	  aclocal: fix for more-than-once specified directories
 7878 	  aclocal: just warn if the primary local m4 dir doesn't exist (don't error)
 7879 
 7880 2013-02-21  Pavel Raiskup  <praiskup@redhat.com>
 7881 
 7882 	aclocal: fix for more-than-once specified directories
 7883 
 7884 	Related to automake bug#13514.
 7885 
 7886 	Do not consider directories for extra m4 files multiple times in
 7887 	'aclocal'.  Doing so caused problems on older packages that specify
 7888 
 7889 	    configure.ac:  AC_CONFIG_MACRO_DIRS([m4])
 7890 	    Makefile.am:   ACLOCAL_AMFLAGS = -I m4
 7891 
 7892 	if the 'm4' directory does not exist when aclocal is called the first
 7893 	time by autoreconf.
 7894 
 7895 	See:
 7896 	<http://lists.gnu.org/archive/html/bug-automake/2013-01/msg00115.html>
 7897 
 7898 	* aclocal.in (scan_m4_files): Remove duplicates in @user_includes.
 7899 	* t/aclocal-macrodir.tap: Extend.
 7900 	* t/aclocal-macrodirs.tap: Likewise.
 7901 
 7902 2013-02-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7903 
 7904 	Merge branch 'docstrings-tweaks' into maint
 7905 
 7906 	* docstrings-tweaks:
 7907 	  cosmetics: fix some "docstring-like" comments in automake
 7908 
 7909 2013-02-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7910 
 7911 	Merge branch 'branch-1.13.2' into maint
 7912 
 7913 	* branch-1.13.2:
 7914 	  coverage: expose automake bug#13760
 7915 	  tests: refactor/enhance tests about make dry-run mode
 7916 	  maint: describe new versioning and branching scheme, and adjust to it
 7917 
 7918 2013-02-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7919 
 7920 	Merge branch 'new-versioning-scheme' into branch-1.13.2
 7921 
 7922 	* new-versioning-scheme:
 7923 	  maint: describe new versioning and branching scheme, and adjust to it
 7924 
 7925 2013-02-20  Pavel Raiskup  <praiskup@redhat.com>
 7926 
 7927 	aclocal: just warn if the primary local m4 dir doesn't exist (don't error)
 7928 
 7929 	Related to automake bug#13514.
 7930 
 7931 	Every package which does not need to have the local m4 macro
 7932 	directory pre-existing in the version control system (because
 7933 	e.g., it does not have nor need any private m4 macros) would
 7934 	fail during the "autoreconf -vfi" phase if AC_CONFIG_MACRO_DIRS([m4])
 7935 	is specified in configure.ac (it could be to instruct tools like
 7936 	'autopoint' and 'libtoolize' to use 'm4' as the local directory
 7937 	where to install definitions of their m4 macros, and to instruct
 7938 	aclocal to look into it).  The failure would go like this:
 7939 
 7940 	  autoreconf: Entering directory `.'
 7941 	  autoreconf: running: aclocal --force
 7942 	  aclocal: error: couldn't open directory 'm4': No such file or directory
 7943 	  autoreconf: aclocal failed with exit status: 1
 7944 
 7945 	The problem is that when 'aclocal' is run for the first time during
 7946 	'autoreconf', the directory 'm4' does not exist yet.  It will be
 7947 	created by e.g., 'libtoolize' or 'autopoint' later on.  During the
 7948 	second 'aclocal' run, the 'm4' directory exists and aclocal does not
 7949 	complain.
 7950 
 7951 	To work around this issue, we degrade the error to a simple warning.
 7952 	The warning is still quite useful when aclocal is run by hand - so
 7953 	we are not removing completely.
 7954 
 7955 	See also:
 7956 	<http://lists.gnu.org/archive/html/bug-automake/2013-01/msg00115.html>
 7957 	<http://lists.gnu.org/archive/html/automake-patches/2010-02/msg00030.html>
 7958 	<http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=565663>
 7959 	<https://bugzilla.redhat.com/show_bug.cgi?id=901333>
 7960 
 7961 	* aclocal.in (SCAN_M4_DIRS_SILENT, SCAN_M4_DIRS_WARN)
 7962 	(SCAN_M4_DIRS_ERROR): New constants.
 7963 	(scan_m4_dirs): Change the second parameter name to $ERR_LEVEL to
 7964 	better reflect new semantic. Use new constants.
 7965 	(scan_m4_files): Adjust to reflect the new 'scan_m4_dirs' semantics.
 7966 	* t/aclocal-macrodir.tap: Adjust.
 7967 	* t/aclocal-macrodirs.tap: Likewise.
 7968 	* THANKS: Update.
 7969 	* NEWS: Likewise.
 7970 
 7971 	Suggested-by: Ben Pfaff <blp@cs.stanford.edu>
 7972 
 7973 2013-02-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7974 
 7975 	coverage: expose automake bug#13760
 7976 
 7977 	* t/make-dryrun.tap: Here.
 7978 	* THANKS: Update with the name of the bug reporter.
 7979 
 7980 2013-02-20  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7981 
 7982 	tests: refactor/enhance tests about make dry-run mode
 7983 
 7984 	* t/make-dryrun.tap: Here.
 7985 
 7986 2013-02-17  Stefano Lattarini  <stefano.lattarini@gmail.com>
 7987 
 7988 	maint: describe new versioning and branching scheme, and adjust to it
 7989 
 7990 	See discussion about automake bug#13578 for more details and background.
 7991 
 7992 	Basically, for the versioning scheme:
 7993 
 7994 	  - micro versions only for bug and regression fixing;
 7995 	  - minor versions for new backward-compatible features, and new
 7996 	    non-fatal deprecations;
 7997 	  - major versions for backward-incompatibilities, complex new
 7998 	    features, and major refactoring.
 7999 
 8000 	And for the git branching scheme:
 8001 
 8002 	  + branch 'next' is for the upcoming major version;
 8003 	  + branch 'master' is now for the upcoming minor version;
 8004 	  + branch 'maint' is for the upcoming micro (bug-fixing) version;
 8005 	  + the merging hierarchy is: 'maint' -> 'master' -> 'next'.
 8006 
 8007 	* HACKING (Automake versioning and compatibility scheme): New.
 8008 	(Working with git): Adjust.
 8009 	* NEWS: Update and fix.
 8010 	* aclocal.in: Adjust some "FIXME" messages.
 8011 	* automake.in: Likewise.
 8012 	* m4/mkdirp.m4: Likewise.
 8013 	* t/aclocal-acdir.sh: Likewise.
 8014 	* t/aclocal-macrodir.tap: Likewise.
 8015 	* t/aclocal-macrodirs.tap: Likewise.
 8016 	* lib/Automake/Options.pm: Likewise.
 8017 	* m4/internal/ac-config-macro-dirs.m4: Likewise.
 8018 
 8019 2013-02-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8020 
 8021 	cosmetics: fix some "docstring-like" comments in automake
 8022 
 8023 	* automake.in: Here.  And remove some redundant ones.
 8024 
 8025 2013-02-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8026 
 8027 	Merge branch 'maint'
 8028 
 8029 	* maint:
 8030 	  style: call perl functions 'like_this()', not '&like_this()'
 8031 	  maint: use more perl subroutines prototypes in the automake script
 8032 	  build: auto-generate perl subroutines prototypes for automake and aclocal
 8033 	  refactor: rip module Automake::Language out of automake script
 8034 
 8035 2013-02-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8036 
 8037 	Merge branch 'perl-sub-protos' into maint
 8038 
 8039 	* perl-sub-protos:
 8040 	  style: call perl functions 'like_this()', not '&like_this()'
 8041 	  maint: use more perl subroutines prototypes in the automake script
 8042 	  build: auto-generate perl subroutines prototypes for automake and aclocal
 8043 	  refactor: rip module Automake::Language out of automake script
 8044 
 8045 2013-02-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8046 
 8047 	Merge branch 'maint'
 8048 
 8049 	* maint:
 8050 	  typofix: in comments in 't/ax/test-lib.sh'
 8051 	  tests on TAP: don't run the driver with perl unconditionally
 8052 	  typofix: in comments in 'automake.in'
 8053 	  tests: remove obsolete (and now wrong) comment
 8054 	  typofix: in diagnostic in test 't/tap-realtime.sh'
 8055 	  automake: fix reference to relevant tests in comments
 8056 	  NEWS: we no longer plan to drop $(INCLUDES) support in next major version
 8057 
 8058 2013-02-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8059 
 8060 	Merge branch 'branch-1.13.2' into maint
 8061 
 8062 	* branch-1.13.2:
 8063 	  typofix: in comments in 't/ax/test-lib.sh'
 8064 	  tests on TAP: don't run the driver with perl unconditionally
 8065 	  typofix: in comments in 'automake.in'
 8066 	  tests: remove obsolete (and now wrong) comment
 8067 	  typofix: in diagnostic in test 't/tap-realtime.sh'
 8068 	  automake: fix reference to relevant tests in comments
 8069 	  NEWS: we no longer plan to drop $(INCLUDES) support in next major version
 8070 
 8071 2013-02-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8072 
 8073 	typofix: in comments in 't/ax/test-lib.sh'
 8074 
 8075 2013-02-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8076 
 8077 	style: call perl functions 'like_this()', not '&like_this()'
 8078 
 8079 	We can do so now that our build rules auto-generate a list of
 8080 	prototypes for all functions ins our scripts.
 8081 
 8082 	* automake.in: Adjust throughout.
 8083 	* HACKING: Adjust advises.
 8084 
 8085 2013-02-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8086 
 8087 	preproc: enhance and extend tests
 8088 
 8089 	* t/preproc-demo.sh: New test, a "demo" of how the new pre-processing
 8090 	feature could be used in a real-world package.
 8091 	* t/preproc-errmsg.sh: New test, check that error messages remain
 8092 	useful when the new pre-processing features are involved.
 8093 	* t/preproc-reldir.sh: Split up ...
 8094 	* t/preproc-basics.sh, t/preproc-c-compile.sh: ... into these two
 8095 	tests, with some refactorings, clean-up and enhancements.
 8096 	* t/list-of-tests.mk: Adjust.
 8097 
 8098 2013-02-14  Peter Rosin  <peda@lysator.liu.se>
 8099 	    Stefano Lattarini  <stefano.lattarini@gmail.com>
 8100 
 8101 	preproc: add support for relative names in included fragments
 8102 
 8103 	The rationale for this change is that it is annoying to have
 8104 	to repeat the directory name when including a Makefile fragment.
 8105 	For deep directory structures these repeats can generate a lot
 8106 	of bloat.  It also hinders reuse and easy directory restructuring
 8107 	if all Makefile fragments have to know exactly where they live.
 8108 
 8109 	Suggested by Bob Friesenhahn, and later discussed in bug#13524.
 8110 
 8111 	In the course of discussion, the following notations were rejected:
 8112 	&{reldir}& - to hard to type, {reldir} - interferes with ${reldir},
 8113 	{am_reldir} - short form {D} interferes with ${D}, @am_reldir@ - short
 8114 	form @D@ interferes with AC_SUBST([D]) as well as invading the
 8115 	config.status turf. Other notations were also suggested...
 8116 
 8117 	* automake.in (read_am_file): Add third argument specifying the
 8118 	relative directory of this Makefile fragment compared to the
 8119 	main Makefile.  Replace %reldir% and %canon_reldir% in the
 8120 	fragment with this relative directory (with slashes etc, or
 8121 	canonicalized).
 8122 	(read_main_am_file): Adjust.
 8123 	* t/preproc-reldir.sh: New test.
 8124 	* t/list-of-tests.mk: Augment.
 8125 	* doc/automake.texi (Include): Document the new feature.
 8126 	NEWS: Add new feature.
 8127 
 8128 2013-02-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8129 
 8130 	tests on TAP: don't run the driver with perl unconditionally
 8131 
 8132 	* t/ax/tap-setup.sh: When a 'Makefile.am' was pre-existent in the
 8133 	test directory at the moment the client test script sourced this
 8134 	file, said 'Makefile.am' was tweaked to provide it with a proper
 8135 	definition of TEST_LOG_DRIVER.  However, there was an error in this
 8136 	automatic definition, since it caused the TAP test driver to be
 8137 	unconditionally invoked with perl.  This wasn't an issue in most
 8138 	situations, since perl is smart enough to re-execute a given script
 8139 	with the proper interpreter if it sees a she-bang line that doesn't
 8140 	seem to point to perl itself.  Still, there is no reason to do
 8141 	something blatantly wrong even if our tools correct the dumb mistake
 8142 	for us.  So fix the TEST_LOG_DRIVER definition.
 8143 
 8144 2013-02-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8145 
 8146 	typofix: in comments in 'automake.in'
 8147 
 8148 2013-02-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8149 
 8150 	tests: remove obsolete (and now wrong) comment
 8151 
 8152 	* t/ax/am-test-lib.sh: Here.
 8153 
 8154 2013-02-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8155 
 8156 	typofix: in diagnostic in test 't/tap-realtime.sh'
 8157 
 8158 2013-02-13  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8159 
 8160 	maint: use more perl subroutines prototypes in the automake script
 8161 
 8162 	* automake.in: Throughout this file.  Note that these new prototypes
 8163 	are not much useful, since many subroutine calls still use the old
 8164 	'&foo' form; but we'll take care of that in later patches.
 8165 	* lib/Automake/Language.pm (target_hook): Call the '_target_hook'
 8166 	of the given language in a more modern form, avoiding '&'.
 8167 
 8168 2013-02-13  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8169 
 8170 	build: auto-generate perl subroutines prototypes for automake and aclocal
 8171 
 8172 	This will allow us to avoid either using the '&foo' invocation form when
 8173 	invoking a subroutine before its definition, or having to maintain the
 8174 	list of prototypes by hand (with the risk of having it become incomplete
 8175 	or fall out-of-sync when future edits to the automake and aclocal scripts
 8176 	are done).
 8177 
 8178 	* Makefile.am (automake, aclocal): Automatically generate a list of
 8179 	prototypes by looking at the subroutines definitions.
 8180 	* bootstrap.sh: Likewise, when generating the temporary automake and
 8181 	aclocal scripts used for bootstrapping.
 8182 	* automake.in: Add a placeholder that will be tracked by the new recipes
 8183 	and substituted with the computed prototypes.  Remove existing prototypes,
 8184 	that are now superfluous. Some adjustments required by the new, more
 8185 	comprehensive prototypes declarations.
 8186 	* aclocal.in: Likewise.
 8187 	* maintainer/syntax-checks.mk (sc_diff_automake, sc_diff_aclocal): Adjust.
 8188 
 8189 2013-02-04  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8190 
 8191 	refactor: rip module Automake::Language out of automake script
 8192 
 8193 	This is just a preparatory patch in view of future changes.
 8194 
 8195 	* lib/Automake/Language.pm: New module, ripped out from ...
 8196 	* automake.in: ... here.  Related adjustments.
 8197 	* Makefile.am (dist_perllib_DATA): List the new module.
 8198 
 8199 2013-02-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8200 
 8201 	automake: fix reference to relevant tests in comments
 8202 
 8203 	Some of those tests had been renamed in recent or less-recent changes.
 8204 
 8205 2013-02-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8206 
 8207 	NEWS: we no longer plan to drop $(INCLUDES) support in next major version
 8208 
 8209 	For a rationale and related discussion, see:
 8210 	<http://lists.gnu.org/archive/html/automake/2013-02/msg00001.html>
 8211 
 8212 	And note that support for INCLUDES has not been re-introduced in the
 8213 	master branch yet, at the moment of writing; but we plan to definitely
 8214 	do so before the next major release.
 8215 
 8216 2013-01-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8217 
 8218 	NEWS: IRIX is still supported; only SGI depmode no longer is
 8219 
 8220 	This change is for the master branch (see the similar commit
 8221 	'v1.13.1b-3-g97f04a5' on the maint branch).
 8222 
 8223 	The only IRIX-specific support that is actually removed in master (and
 8224 	thus in the next major Automake version) is the depcomp support for the
 8225 	SGI compiler.  That means that automatic dependency tracking will no
 8226 	longer work with that compiler, but "normal" compilation should still
 8227 	work, at least until the compiler is supported by Autoconf.
 8228 
 8229 	So there is no point in alarming our users by stating in the NEWS file
 8230 	that "support for IRIX and the SGI compilers has been removed"; after
 8231 	all, while we don't test on nor particularly care about IRIX anymore,
 8232 	that doesn't mean we are deliberately breaking it, and the likelihood
 8233 	of an intended breakage there is very low.
 8234 
 8235 	See:
 8236 	<http://lists.gnu.org/archive/html/automake-patches/2013-01/msg00164.html>
 8237 	<http://lists.gnu.org/archive/html/automake-patches/2013-01/msg00165.html>
 8238 
 8239 	* NEWS: Adjust.  Improve explanation of our rationale for dropping
 8240 	support for the SGI compilers.
 8241 
 8242 	Suggested-by: Peter Rosin <peda@lysator.liu.se>
 8243 
 8244 2013-01-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8245 
 8246 	Merge branch 'maint'
 8247 
 8248 	* maint:
 8249 	  NEWS: IRIX is still supported; only SGI C compiler depcomp no longer is
 8250 	  maint: version bump after beta release 1.13.1b
 8251 	  NEWS: typofix
 8252 	  release: beta release 1.13.1b (will become 1.13.2)
 8253 
 8254 2013-01-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8255 
 8256 	Merge branch 'branch-1.13.2' into maint
 8257 
 8258 	* branch-1.13.2:
 8259 	  NEWS: IRIX is still supported; only SGI C compiler depcomp no longer is
 8260 	  maint: version bump after beta release 1.13.1b
 8261 	  NEWS: typofix
 8262 	  release: beta release 1.13.1b (will become 1.13.2)
 8263 
 8264 2013-01-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8265 
 8266 	NEWS: IRIX is still supported; only SGI C compiler depcomp no longer is
 8267 
 8268 	This change is for the maint branch.
 8269 
 8270 	The only IRIX-specific support that is going to actually be removed in
 8271 	the next major Automake versions is the depcomp support for the SGI
 8272 	compiler.  That means that automatic dependency tracking will no
 8273 	longer work with that compiler, but "normal" compilation should still
 8274 	work, at least until the compiler is supported by Autoconf.
 8275 
 8276 	So there is no point in alarming our users by stating in the NEWS file
 8277 	that "support for IRIX and the SGI compilers is going to be removed";
 8278 	after all, while we don't test on nor particularly care about IRIX
 8279 	anymore, that doesn't mean we are deliberately breaking it, and the
 8280 	likelihood of an intended breakage there is very low.
 8281 
 8282 	See:
 8283 	<http://lists.gnu.org/archive/html/automake-patches/2013-01/msg00164.html>
 8284 	<http://lists.gnu.org/archive/html/automake-patches/2013-01/msg00165.html>
 8285 
 8286 	* NEWS (Future backward-incompatibilities: Adjust.  Improve explanation
 8287 	of our rationale for dropping SGI support. Fix an unrelated typo while
 8288 	at it.
 8289 
 8290 	Suggested-by: Peter Rosin <peda@lysator.liu.se>
 8291 
 8292 2013-01-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8293 
 8294 	coverage: split info files are not installed nor distributed
 8295 
 8296 	See automake bugs #12320 and #13351.
 8297 
 8298 	* t/txinfo-no-extra-dist.sh: Enhance.
 8299 
 8300 2013-01-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8301 
 8302 	maint: version bump after beta release 1.13.1b
 8303 
 8304 	* configure.ac (AC_INIT): Bump version number to 1.13.1c.
 8305 	* m4/amversion.m4: Likewise (autoupdated by "make bootstrap").
 8306 
 8307 2013-01-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8308 
 8309 	NEWS: typofix
 8310 
 8311 2013-01-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8312 
 8313 	release: beta release 1.13.1b (will become 1.13.2)
 8314 
 8315 	* configure.ac (AC_INIT): Bump version number to 1.13.1b.
 8316 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 8317 
 8318 2013-01-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8319 
 8320 	Merge branch 'maint'
 8321 
 8322 	* maint:
 8323 	  docs: '.txi' and '.texinfo' extensions are deprecated
 8324 	  NEWS: document recent documentation improvements
 8325 	  docs: more precise cross reference
 8326 	  docs: 'dist-shar' and 'dist-tarZ' are obsolescent today
 8327 	  docs: improve documentation of 'dist-*' targets slightly
 8328 	  docs: make even clearer 'dist-gzip' is the default.
 8329 	  docs: document 'dist-xz' together with the other 'dist-*' options
 8330 	  docs: 'no-define' option and AM_INIT_AUTOMAKE three-args usage: fixlets
 8331 	  warn: correct broken hyperlink in warning message
 8332 
 8333 2013-01-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8334 
 8335 	Merge branch 'branch-1.13.2' into maint
 8336 
 8337 	* branch-1.13.2:
 8338 	  docs: '.txi' and '.texinfo' extensions are deprecated
 8339 	  NEWS: document recent documentation improvements
 8340 	  docs: more precise cross reference
 8341 	  docs: 'dist-shar' and 'dist-tarZ' are obsolescent today
 8342 	  docs: improve documentation of 'dist-*' targets slightly
 8343 	  docs: make even clearer 'dist-gzip' is the default.
 8344 	  docs: document 'dist-xz' together with the other 'dist-*' options
 8345 	  docs: 'no-define' option and AM_INIT_AUTOMAKE three-args usage: fixlets
 8346 	  warn: correct broken hyperlink in warning message
 8347 
 8348 2013-01-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8349 
 8350 	docs: '.txi' and '.texinfo' extensions are deprecated
 8351 
 8352 	And Automake will warn about them (since commit 'v1.13.1-6-ge1ed314').
 8353 
 8354 	* doc/automake.texi (Texinfo): Adjust.
 8355 
 8356 2013-01-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8357 
 8358 	NEWS: document recent documentation improvements
 8359 
 8360 2013-01-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8361 
 8362 	docs: more precise cross reference
 8363 
 8364 	See automake bug#13520.
 8365 
 8366 	* doc/automake.texi (The Types of Distributions): Here,
 8367 	cross-reference "List of Automake options" rather then
 8368 	the more generic node "Options".  Improve wording while
 8369 	at it.
 8370 
 8371 2013-01-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8372 
 8373 	docs: 'dist-shar' and 'dist-tarZ' are obsolescent today
 8374 
 8375 	Both the options and the formats; and they might be deprecated
 8376 	and removed in future automake versions (see discussion on
 8377 	automake bug#13324).
 8378 
 8379 	In any case, it's better if the documentation starts advising
 8380 	against their use right now.
 8381 
 8382 	* doc/automake.texi (The Types of Distributions): Here.
 8383 	(List of Automake options): And here.
 8384 
 8385 2013-01-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8386 
 8387 	docs: improve documentation of 'dist-*' targets slightly
 8388 
 8389 	* doc/automake.texi (The Types of Distributions): Here.
 8390 
 8391 2013-01-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8392 
 8393 	docs: make even clearer 'dist-gzip' is the default.
 8394 
 8395 	See automake bug#13520.
 8396 
 8397 	* doc/automake.texi (The Types of Distributions): Here.
 8398 
 8399 2013-01-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8400 
 8401 	docs: document 'dist-xz' together with the other 'dist-*' options
 8402 
 8403 	See automake bug#13520.
 8404 
 8405 	* doc/automake.texi (List of Automake options): Here.
 8406 
 8407 2013-01-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8408 
 8409 	docs: 'no-define' option and AM_INIT_AUTOMAKE three-args usage: fixlets
 8410 
 8411 	See automake bug#13519.
 8412 
 8413 	* doc/automake.texi: Adjust.
 8414 
 8415 2013-01-22  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8416 
 8417 	warn: correct broken hyperlink in warning message
 8418 
 8419 	* automake.in (scan_autoconf_traces): Here, when an use of the
 8420 	deprecated two- and three-arguments forms of AM_INIT_AUTOMAKE
 8421 	is detected.
 8422 
 8423 	Fixes automake bug#13519.
 8424 
 8425 2013-01-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8426 
 8427 	Merge branch 'maint'
 8428 
 8429 	* maint:
 8430 	  tests: more information about Lex and Yacc programs
 8431 	  lint: fix spurious failure for 'sc_rm_minus_f' syntax check
 8432 	  maint: bump version 1.13.1a -> 1.13.2a
 8433 	  maint: update copyright in files generated by automake and aclocal
 8434 	  tests: avoid a spurious failure when running inside Emacs
 8435 	  tests: make two new test executable
 8436 	  m4: rename an m4 file to a more appropriate name
 8437 	  NEWS: update w.r.t. recent documentation fixes
 8438 	  compat: reinstate AM_CONFIG_HEADER and AM_PROG_CC_STDC
 8439 	  docs: parallel-tests is no longer experimental
 8440 	  docs: serial-tests are not deprecated, just discouraged
 8441 	  plans: we are not going to remove AM_PROG_MKDIR_P in Automake 1.14
 8442 	  NEWS: we are not going to remove AM_PROG_MKDIR_P in Automake 1.14
 8443 	  init.m4: add probe to check "rm -f" without args work
 8444 
 8445 2013-01-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8446 
 8447 	tests: more information about Lex and Yacc programs
 8448 
 8449 	* t/get-sysconf.sh: Try to also get the version of '$LEX' and '$YACC'.
 8450 	This will help debugging of user-reported problems.
 8451 
 8452 2013-01-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8453 
 8454 	lint: fix spurious failure for 'sc_rm_minus_f' syntax check
 8455 
 8456 	* maintainer/syntax-checks.mk (sc_rm_minus_f): Also exempt
 8457 	file 't/rm-f-probe.sh'.
 8458 
 8459 2013-01-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8460 
 8461 	maint: bump version 1.13.1a -> 1.13.2a
 8462 
 8463 	The 1.13.2 bug-fixing release will ship from the 'branch-1.13.2' git
 8464 	branch, not from the 'maint' one, since the latter contains changes
 8465 	that are non-trivial and hasn't cooked enough yet.  The 'maint' branch
 8466 	will give rise to the 1.13.3 release instead, eventually.  Adjust the
 8467 	version number to match.
 8468 
 8469 	* configure.ac (AC_INIT): Bump version number to 1.13.2b.
 8470 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 8471 
 8472 2013-01-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8473 
 8474 	Merge branch 'branch-1.13.2' into maint
 8475 
 8476 	That branch is for the "emergency" bug-fixing release 1.13.2.
 8477 
 8478 	* branch-1.13.2:
 8479 	  maint: update copyright in files generated by automake and aclocal
 8480 	  tests: avoid a spurious failure when running inside Emacs
 8481 	  tests: make two new test executable
 8482 	  m4: rename an m4 file to a more appropriate name
 8483 	  NEWS: update w.r.t. recent documentation fixes
 8484 	  compat: reinstate AM_CONFIG_HEADER and AM_PROG_CC_STDC
 8485 	  docs: parallel-tests is no longer experimental
 8486 	  docs: serial-tests are not deprecated, just discouraged
 8487 	  NEWS: we are not going to remove AM_PROG_MKDIR_P in Automake 1.14
 8488 
 8489 2013-01-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8490 
 8491 	maint: update copyright in files generated by automake and aclocal
 8492 
 8493 	* bootstrap.sh ($RELEASE_YEAR): Bump top 2013.
 8494 	* configure.ac ($RELEASE_YEAR): Likewise.
 8495 
 8496 2013-01-21  Thien-Thi Nguyen  <ttn@gnuvola.org>  (tiny change)
 8497 
 8498 	tests: avoid a spurious failure when running inside Emacs
 8499 
 8500 	Some versions of Emacs set the environment variable 'EMACS' to 't'
 8501 	for child processes.  Thus, when running from inside Emacs, "$(MAKE) -e"
 8502 	erroneously allows the 't' to override the one in the Makefile.
 8503 
 8504 	* t/lisp-flags.sh: Unset var 'EMACS', fixing the issue.
 8505 
 8506 2013-01-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8507 
 8508 	tests: make two new test executable
 8509 
 8510 	* t/backslash-issues.sh: This.
 8511 	* t/extra-data.sh: And this.
 8512 
 8513 	Issue revealed by the 'sc_tests_executable' maintainer check.
 8514 
 8515 2013-01-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8516 
 8517 	m4: rename an m4 file to a more appropriate name
 8518 
 8519 	* m4/obsolete-err.m4: Rename ...
 8520 	* m4/obsolete.m4: ... like this.
 8521 	* Makefile.am (dist_automake_ac_DATA): Adjust.
 8522 	* t/ansi2knr-no-more.sh: Likewise.
 8523 
 8524 2013-01-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8525 
 8526 	NEWS: update w.r.t. recent documentation fixes
 8527 
 8528 2013-01-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8529 
 8530 	compat: reinstate AM_CONFIG_HEADER and AM_PROG_CC_STDC
 8531 
 8532 	Make them give runtime warnings in the obsolete category, but apart
 8533 	from that, make them behave as they did in Automake 1.12.x and earlier.
 8534 
 8535 	While removing those macros seemed quite harmless, because it didn't put
 8536 	a real burden on the developers (requiring them just to do a quick edit
 8537 	to configure.ac), it turned out to place an unsustainable burden (or at
 8538 	least, a burden perceived as such) on distro packagers who use the latest
 8539 	Automake to bootstrap existing packages.  Many of those packages, while
 8540 	having likely updated to AC_CONFIG_HEADERS in their development version,
 8541 	still used AM_CONFIG_HEADER in their existing released versions, and the
 8542 	removal of this macro would have thus forced the Fedora packagers to
 8543 	patch all of them.  References:
 8544 
 8545 	  <http://www.mail-archive.com/devel@lists.fedoraproject.org/msg52840.html>
 8546 	  <http://www.spinics.net/lists/fedora-devel/msg175922.html>
 8547 	  <http://blog.flameeyes.eu/2013/01/autotools-mythbuster-automake-pains>
 8548 
 8549 	In addition, the Fedora packagers have already decided to patch their
 8550 	Automake 1.13.1 to reinstate the AM_CONFIG_HEADER and AM_PROG_CC_STDC
 8551 	macros (plus other macros that I don't believe it's worth worrying about):
 8552 
 8553 	  <http://www.spinics.net/lists/fedora-devel/msg176098.html>
 8554 	  <http://www.mail-archive.com/devel@lists.fedoraproject.org/msg53030.html>
 8555 	  <http://pkgs.fedoraproject.org/cgit/automake.git/commit/?id=ffe6bc39>
 8556 
 8557 	So, rather than having one more incompatibility floating around, we
 8558 	better mirror that change (or, actually, its relevant parts) in the
 8559 	upstream.
 8560 
 8561 	* m4/obsolete-err.m4 (AM_CONFIG_HEADER, AM_PROG_CC_STDC): Revert to the
 8562 	older semantics, plus a runtime warning in the 'obsolete' category.
 8563 	* t/backcompat6.sh: Use AM_CONFIG_HEADER once again.
 8564 	* t/am-config-header-no-more.sh: Rename ...
 8565 	* t/am-config-header.sh: ... like this, and adjust.
 8566 	* t/am-prog-cc-stdc-no-more.sh: Rename ...
 8567 	* t/am-prog-cc-stdc.sh: ... like this, and adjust.
 8568 	* t/list-of-tests.mk: Adjust.
 8569 	* NEWS: Update.
 8570 
 8571 2013-01-18  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8572 
 8573 	Merge branch 'rm-f-probe' into maint
 8574 
 8575 	* rm-f-probe:
 8576 	  init.m4: add probe to check "rm -f" without args work
 8577 
 8578 2013-01-18  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8579 
 8580 	docs: parallel-tests is no longer experimental
 8581 
 8582 	So don't declare it as such in the documentation.
 8583 
 8584 	Reported by Brandon Black:
 8585 	<http://lists.gnu.org/archive/html/automake/2013-01/msg00052.html>
 8586 
 8587 	* doc/automake.texi: Adjust.
 8588 	* THANKS: Update.
 8589 
 8590 2013-01-18  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8591 
 8592 	docs: serial-tests are not deprecated, just discouraged
 8593 
 8594 	We don't plan to remove support for them, nor to have the serial-tests
 8595 	option give any kind of runtime warning, so don't alarm the users
 8596 	still using serial tests with pointless "deprecation" or "obsolescence"
 8597 	warnings.
 8598 
 8599 	Fixes automake bug#13478.
 8600 
 8601 	See also:
 8602 	<http://lists.gnu.org/archive/html/automake/2013-01/msg00058.html>
 8603 
 8604 	* doc/automake.texi: Adjust.
 8605 	* THANKS: Update.
 8606 
 8607 2013-01-18  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8608 
 8609 	NEWS (master): we are not going to remove AM_PROG_MKDIR_P in 1.14
 8610 
 8611 	See commit v1.13.1-109-g030ecb4 of 2013-01-16, "compat: restore
 8612 	AM_PROG_MKDIR, again", for the rationale.
 8613 
 8614 2013-01-18  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8615 
 8616 	plans: we are not going to remove AM_PROG_MKDIR_P in Automake 1.14
 8617 
 8618 	See commit v1.13.1-109-g030ecb4 of 2013-01-16, "compat: restore
 8619 	AM_PROG_MKDIR, again", for the rationale; that rationale is now
 8620 	also copied...
 8621 
 8622 	* PLANS/obsolete-removed/am-prog-mkdir-p.txt: ... here.
 8623 
 8624 2013-01-18  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8625 
 8626 	NEWS: we are not going to remove AM_PROG_MKDIR_P in Automake 1.14
 8627 
 8628 	See commit v1.13.1-109-g030ecb4 of 2013-01-16, "compat: restore
 8629 	AM_PROG_MKDIR, again", for the rationale.
 8630 
 8631 2013-01-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8632 
 8633 	Merge branch 'texi-no-split-info'
 8634 
 8635 	* texi-no-split-info:
 8636 	  texi: no longer support split info files
 8637 
 8638 2013-01-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8639 
 8640 	Merge branch 'maint'
 8641 
 8642 	* maint:
 8643 	  subdir-objects: complain if it isn't enabled
 8644 	  plans: update w.r.t. latest changes
 8645 
 8646 2013-01-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8647 
 8648 	compat: restore AM_PROG_MKDIR, again
 8649 
 8650 	OK, this is getting ridiculous, but we cannot remove this macro yet
 8651 	(and, yes, the fault for this mess lies entirely on me; let's not
 8652 	dwell on that, thank you very much).
 8653 
 8654 	Gettext (so far the greatest "offender" in the use of AM_PROG_MKDIR), in
 8655 	its latest release 0.18.2, has removed all the uses of that macro still
 8656 	present in its code base.  So I thought we could finally and safely
 8657 	remove it.  Wrong.  If a package's 'configure.ac' contains a call like:
 8658 
 8659 	   AM_GNU_GETTEXT_VERSION([0.18])
 8660 
 8661 	then the 'autopoint' script will bring the data files from the Gettext
 8662 	release *1.18* into the package's tree -- yes, even even if the developer
 8663 	has installed *and is using* Gettext 1.18.2!  Now, these data files
 8664 	comprise m4 files (that will be seen by subsequent aclocal and autoconf
 8665 	calls), and of course, the pre-0.18.2 version of some of these files
 8666 	still contains occurrences of AM_PROG_MKDIR_P -- so Automake 1.13 errors
 8667 	out, and we lose.  This has already happened in practice:
 8668 
 8669 	  <http://lists.gnu.org/archive/html/bug-grep/2013-01/msg00003.html>
 8670 
 8671 	Moreover, while we might see it as not unreasonable to ask a developer
 8672 	using Automake 1.14 to also update Gettext to 1.18.2, that would not
 8673 	be enough; in order for gettext to use the correct data files, our
 8674 	developer would have to update his configure.ac to read:
 8675 
 8676 	   AM_GNU_GETTEXT_VERSION([0.18.2])
 8677 
 8678 	thus requiring *all* of his co-developers to install Gettext 1.18.2,
 8679 	even if they are still using, say, Automake 1.13.  Bad.
 8680 
 8681 	So we re-instate this macro as a simple alias for AC_PROG_MKDIR (plus
 8682 	a non-fatal runtime warning in the 'obsolete' category), and drop any
 8683 	plan to remove it (see how much good those plans have done us so far).
 8684 
 8685 	Note that NEWS is not yet adjusted, since we'll have to adjust it in
 8686 	maint before (to minimize spurious merge conflicts).
 8687 
 8688 	* doc/automake.texi: Update.
 8689 	* PLANS/obsolete-removed/am-prog-mkdir-p.txt: Likewise.
 8690 	* t/gettext-macros.sh: Adjust.
 8691 	* t/am-prog-mkdir-p.sh: New test.
 8692 	* t/mkdir_p.sh: Remove, folded into the new one.
 8693 	* t/am-prog-mkdir-p-no-more: Remove as superseded.
 8694 	* t/list-of-tests.mk: Adjust.
 8695 	* t/obsolete-err.m4: Re-instate AM_PROG_MKDIR_P as a working
 8696 	alias for AC_PROG_MKDIR_P (albeit giving runtime warnings, and
 8697 	calling AC_SUBST on 'mkdir_p' too).
 8698 	* m4/init.m4 (AM_INIT_AUTOMAKE): No longer call AC_SUBST for
 8699 	'mkdir_p', as that is once again AM_PROG_MKDIR_P's business.
 8700 
 8701 2013-01-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8702 
 8703 	init.m4: add probe to check "rm -f" without args work
 8704 
 8705 	See automake bug#10828.
 8706 
 8707 	POSIX will say in a future version that running "rm -f" with no argument
 8708 	is OK: <http://austingroupbugs.net/view.php?id=542>).
 8709 
 8710 	We want to be able to make that assumption in our Makefile recipes.
 8711 	So we introduce an aggressive probe to check that the usage we want is
 8712 	actually supported "in the wild" to an acceptable degree.
 8713 
 8714 	* m4/init.m4 (AM_INIT_AUTOMAKE): Implement the probe.  To make any issue
 8715 	more visible, cause the running configure to be aborted by default if
 8716 	the 'rm' program in use doesn't match our expectations; the user can
 8717 	still override this though, by setting the ACCEPT_INFERIOR_RM_PROGRAM
 8718 	environment variable to "yes".
 8719 
 8720 	* t/spy-rm.tap: Update heading comments.
 8721 	* t/rm-f-probe.sh: New test.
 8722 	* t/list-of-tests.mk: Add it.
 8723 	* PLANS/rm-f-without-args.txt: Adjust.
 8724 
 8725 2013-01-13  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8726 
 8727 	subdir-objects: complain if it isn't enabled
 8728 
 8729 	Since the next major automake version will make the behaviour so far
 8730 	only activated with the 'subdir-object' option mandatory, it's better
 8731 	if we start warning users not using that option.
 8732 
 8733 	As suggested by Peter Johansson, we strive to avoid the warning when
 8734 	it would be irrelevant, i.e., if all source files sit in "current"
 8735 	directory.
 8736 
 8737 	See automake bug#13378.
 8738 
 8739 	* automake.in (handle_single_transform): Print the warning when
 8740 	necessary.
 8741 	* t/subobj.sh: Enhance.
 8742 	* t/ax/depcomp.sh: Adjust.
 8743 	* t/cscope.tap: Likewise.
 8744 	* t/depcomp8a.sh: Likewise.
 8745 	* t/depcomp8b.sh: Likewise.
 8746 	* t/ext2.sh: Likewise.
 8747 	* t/extra-portability.sh: Likewise.
 8748 	* t/fort2.sh: Likewise.
 8749 	* t/fort4.sh: Likewise.
 8750 	* t/fort5.sh: Likewise.
 8751 	* t/lex-line.sh: Likewise.
 8752 	* t/libtool3.sh: Likewise.
 8753 	* t/ltinstloc.sh: Likewise.
 8754 	* t/ltlibsrc.sh: Likewise.
 8755 	* t/ltorder.sh: Likewise.
 8756 	* t/parallel-tests-suffix-prog.sh: Likewise.
 8757 	* t/sourcefile-in-subdir.sh: Likewise.
 8758 	* t/specflg9.sh: Likewise.
 8759 	* t/subobj4.sh: Likewise.
 8760 	* t/subobj7.sh: Likewise.
 8761 	* t/subpkg-yacc.sh: Likewise.
 8762 	* t/subpkg.sh: Likewise.
 8763 	* t/suffix-custom-subobj-and-specflg.sh: Likewise.
 8764 	* t/vala-libs.sh: Likewise.
 8765 	* t/vala-non-recursive-setup.sh: Likewise.
 8766 	* t/yacc-grepping2.sh: Likewise.
 8767 	* t/yacc-line.sh: Likewise.
 8768 
 8769 2013-01-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8770 
 8771 	plans: update w.r.t. latest changes
 8772 
 8773 2013-01-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8774 
 8775 	Merge branch 'maint'
 8776 
 8777 	* maint:
 8778 	  ywrap: remove an obsolete FIXME comment
 8779 	  ywrap: style fixes (no semantic change intended)
 8780 	  convenience: "make lint" as an alias for "make maintainer-check"
 8781 	  docs: typofix in manual
 8782 	  coverage: using multiple lexers in a single program
 8783 
 8784 2013-01-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8785 
 8786 	ywrap: remove an obsolete FIXME comment
 8787 
 8788 	If it were still relevant, somebody would have complained by now.
 8789 
 8790 2013-01-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8791 
 8792 	ywrap: style fixes (no semantic change intended)
 8793 
 8794 2013-01-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8795 
 8796 	convenience: "make lint" as an alias for "make maintainer-check"
 8797 
 8798 	* maintainer/syntax-checks.mk (lint): Here.  I'm a lazy typist ...
 8799 
 8800 2013-01-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8801 
 8802 	docs: typofix in manual
 8803 
 8804 	* doc/automake.texi (Yacc and Lex): Here, don't write "automake -i"
 8805 	where "automake -a" is actually intended.  Re-wrap some text while
 8806 	at it.
 8807 
 8808 2013-01-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8809 
 8810 	coverage: using multiple lexers in a single program
 8811 
 8812 	Using Flex and Automake built-in support for lex, that is possible.
 8813 	A little tricky, but not difficult.
 8814 
 8815 	See:
 8816 	<http://lists.gnu.org/archive/html/automake/2010-10/msg00081.html>
 8817 	<http://lists.gnu.org/archive/html/automake/2009-03/msg00061.html>
 8818 
 8819 	* t/lex-multiple.sh: New test.
 8820 	* t/list-of-tests.mk: Add it.
 8821 
 8822 2013-01-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8823 
 8824 	Merge branch 'maint'
 8825 
 8826 	* maint:
 8827 	  tests: remove most uses of the AM_PROG_CC_C_O obsolete macro
 8828 	  coverage: obsolete macro AM_PROG_CC_C_O should cause no warning nor errors
 8829 	  INSTALL: update copyright years
 8830 	  ithreads: use runtime (not configure time) detection of perl threads
 8831 	  copyright: add few missing copyright notices
 8832 	  maint: files in PLANS are to be exempted from copyright notice
 8833 	  maint: consistently honor the UPDATE_COPYRIGHT_YEAR environment variable
 8834 	  copyright: update some copyright years
 8835 
 8836 2013-01-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8837 
 8838 	tests: remove most uses of the AM_PROG_CC_C_O obsolete macro
 8839 
 8840 	Our NEWS file says its use will no longer be required in Automake
 8841 	1.13, so better make sure that is actually the case.
 8842 
 8843 	* Several tests: Adjust.
 8844 
 8845 2013-01-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8846 
 8847 	coverage: obsolete macro AM_PROG_CC_C_O should cause no warning nor errors
 8848 
 8849 	Suggested by Eric Blake.
 8850 
 8851 	* t/am-prog-cc-c-o.sh: New test.
 8852 	* t/list-of-tests.mk: Add it.
 8853 
 8854 2013-01-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8855 
 8856 	INSTALL: update copyright years
 8857 
 8858 2013-01-12  Mike Frysinger  <vapier@gentoo.org>  (tiny change)
 8859 
 8860 	ithreads: use runtime (not configure time) detection of perl threads
 8861 
 8862 	I can't imagine the runtime checks being a big runtime penalty, so there
 8863 	shouldn't be a need to do the checks at configure check and hardcode the
 8864 	result in the generated automake.
 8865 
 8866 	With the current system, it means if you change your perl config (build
 8867 	perl w/threads, build automake, build perl w/out threads), or deploy a
 8868 	compiled automake package on a different system (build had threads, but
 8869 	deployed system does not), you get errors when trying to run automake.
 8870 
 8871 	So take the logic from configure.ac and move it to the one place where
 8872 	PERL_THREADS is used (lib/Automake/Config.in) and do the version/config
 8873 	checking at runtime.
 8874 
 8875 	* bootstrap.sh (PERL_THREADS): Delete assignment and use in sed.
 8876 	* configure.ac (am_cv_prog_PERL_ithreads, PERL_THREADS): Delete all code
 8877 	related to these two variables.
 8878 	* lib/Automake/Config.in (perl_threads): Initialize to 0, and only set to
 8879 	1 if the perl version is at least 5.007_002, and useithreads is in Config.
 8880 
 8881 2013-01-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8882 
 8883 	copyright: add few missing copyright notices
 8884 
 8885 	Issue revealed by warnings from "make update-copyright".
 8886 
 8887 	* maintainer/am-ft: Add copyright notice.
 8888 	* maintainer/am-xft: Likewise.
 8889 	* maintainer/rename-tests: Likewise.
 8890 
 8891 2013-01-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8892 
 8893 	maint: files in PLANS are to be exempted from copyright notice
 8894 
 8895 	* maintainer/maint.mk (update-copyright): Adjust.
 8896 
 8897 2013-01-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8898 
 8899 	maint: consistently honor the UPDATE_COPYRIGHT_YEAR environment variable
 8900 
 8901 	* maintainer/maint.mk (update-copyright): Here.  The 'lib/update-copyright'
 8902 	already honoured it, but some parts of our recipe didn't.  This has caused
 8903 	the incomplete copyright bump that was fixed by the previous patch.
 8904 
 8905 2013-01-12  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8906 
 8907 	copyright: update some copyright years
 8908 
 8909 	With "make update-copyright".  Apparently they were missed in the
 8910 	last bump.
 8911 
 8912 	* bootstrap.sh, configure.ac, t/txinfo-builddir.sh: In these files.
 8913 
 8914 2013-01-11  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8915 
 8916 	Merge branch 'maint'
 8917 
 8918 	* maint:
 8919 	  compile: use 'compile' script when "-c -o" is used with losing compilers
 8920 	  HACKING: suggest more checks before releasing
 8921 	  tests: can fake a compiler not grasping "-c -o" -- globally in all tests
 8922 	  sync: update files from upstream with "make fetch"
 8923 	  typofix: in comments in GNUmakefile
 8924 	  Rename 'maint/' -> 'maintainer/', for Git's sake
 8925 	  HACKING: minor typofix
 8926 	  HACKING: bug-tracker, the PLANS directory, and how to plan "big" changes
 8927 	  HACKING: rewindable branches should live in the 'experimental/*' namespace
 8928 	  HACKING: fixlets about git branch rewinding policy
 8929 	  HACKING: commit messages are not to follow GCS ChangeLog rules too strongly
 8930 	  HACKING: "detailed explanation" in commit messages is almost mandatory
 8931 	  HACKING: we use "merge --log" even when merging master
 8932 	  HACKING: typofix
 8933 	  depend2.am: fix comments on verbosity of compilation rules
 8934 	  depend2.am: improve comments a little
 8935 	  plans: automake 1.14 is to assume "rm -f" with no args is OK
 8936 	  plans: we want to active subdir-objects unconditionally in automake 1.14
 8937 
 8938 2013-01-11  Stefano Lattarini  <stefano.lattarini@gmail.com>
 8939 
 8940 	compile: use 'compile' script when "-c -o" is used with losing compilers
 8941 
 8942 	Do so seen when only source files in the "current" directory are present.
 8943 
 8944 	This commit is part of a series of related changes addressing automake
 8945 	bug#13378 (see also the plan 'PLANS/subdir-objects.txt').
 8946 
 8947 	Before this change, Automake-generated C compilation rules mistakenly
 8948 	passed the "-c -o" options combination unconditionally (even to losing
 8949 	compiler) when the 'subdir-objects' was used but sources were only
 8950 	present in the top-level directory.  Issue spotted by Nick Bowler:
 8951 
 8952 	  <http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13378#35>
 8953 	  <http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13378#44>
 8954 
 8955 	We fix this by having Automake redefine AC_PROG_CC to take over the role
 8956 	of AM_PROG_CC_C_O and to require the 'compile' script unconditionally
 8957 	(albeit that will continue to be invoked only when inferior compilers
 8958 	are detected).
 8959 
 8960 	Among other things, this means AM_PROG_CC_C_O explicitly is no longer
 8961 	required; that macro is still supported for backward-compatibility, but
 8962 	calling it is basically a no-op now.
 8963 
 8964 	This change has some pros and some cons (obviously, we believe the former
 8965 	outweighs the latter).  Here are the most relevant ones:
 8966 
 8967 	+ Pros 1:
 8968 	  Some logic in the Automake script has been simplified.
 8969 	+ Pros 2:
 8970 	  That simplification has automatically fixed an actual bug (see
 8971 	  Nick's mails referenced above; admittedly, that was present only in
 8972 	  corner-case situations, but still); the test 't/ccnoco4.sh', which
 8973 	  demonstrated the bug and has been failing so far, now passes.
 8974 	+ Pros 3:
 8975 	  Things works more "automagically" now (no need to manually add the
 8976 	  AM_PROG_CC_C_O macro to configure.ac anymore).
 8977 
 8978 	* Cons 1:
 8979 	  The 'compile' script will be required in all projects using C
 8980 	  compilation; this will only be a problem for packages not using
 8981 	  '--add-missing'.  However, such packages are definitely more rare
 8982 	  than the ones using '--add-missing', and adjusting them will be
 8983 	  trivial -- just copy the compile script over from the new Automake
 8984 	  installation.
 8985 	* Cons 2:
 8986 	  The copy & paste of autoconf internals hack this change has introduced
 8987 	  in our "rewrite" of AC_PROG_CC is really an egregious abomination.  It
 8988 	  can only be justified with the fact that we expect future versions of
 8989 	  autoconf to implement the semantics we need directly in AC_PROG_CC, so
 8990 	  that we'll be able to leverage that (since Automake 1.14 will require
 8991 	  the latest Autoconf version released).
 8992 
 8993 	Now, the detailed list of file-by-file changes ...
 8994 
 8995 	* automake.in ($seen_cc_c_o): Remove this global variable.
 8996 	(scan_autoconf_traces): Don't set it, and do not trace the
 8997 	'AM_PROG_CC_C_O' m4 macro.
 8998 	(lang_c_rewrite): Remove, no longer needed.
 8999 	* doc/automake.texi: Adjust expected "autoreconf --install" output
 9000 	in the amhello example.  Remove statements about the need for the
 9001 	AM_PROG_CC_C_O macro.  Report it is obsolete now.
 9002 	* m4/init.m4: Re-write AC_PROG_CC to append checks about whether the
 9003 	C compiler supports "-c -o" together.  These checks have basically
 9004 	been ripped out (with adaptations) from the 'AC_PROG_CC_C_O' macro
 9005 	of Autoconf and ...
 9006 	* m4/minuso.m4 (AM_PROG_CC_C_O): ... this macro of ours, which has
 9007 	thus basically become a no-op.
 9008 	* t/ax/am-test-lib.sh (am_setup_testdir): Also copy the 'compile'
 9009 	script in the test directory; if we don't do so, every test using
 9010 	AC_PROG_CC should call automake with the "--add-missing" option, or
 9011 	copy the 'compile' script itself.
 9012 	* t/cond11.sh: No need to create a dummy 'compile' script: that is
 9013 	already brought in by 'am_setup_testdir()', that is automatically
 9014 	invoked when 'test-lib.sh' is sourced.
 9015 	* t/add-missing.tap: Adjust: we expect the 'compile' script to be
 9016 	required by a mere AC_PROG_CC call now.
 9017 	* t/dist-auxdir-many-subdirs.sh: Likewise.
 9018 	* t/specflg6.sh: Likewise.
 9019 	* t/subobj4.sh: Likewise.
 9020 	* t/cxx-lt-demo.sh: Likewise, and update comments to match.
 9021 	* t/distcom2.sh: Enhance a little.
 9022 	* t/dollarvar2.sh: Adjust.
 9023 	* t/extra-portability.sh: Likewise.
 9024 	* t/libobj19.sh: Likewise.
 9025 	* t/per-target-flags.sh: Likewise.
 9026 	* t/repeated-options.sh: Likewise.
 9027 	* t/subobj.sh: Likewise, and enhance a little.
 9028 	* t/ccnoco2.sh: Remove as obsolete.
 9029 	* t/list-of-tests.mk (handwritten_TESTS): Adjust.
 9030 	(XFAIL_TESTS): Remove 't/ccnoco4.sh'.
 9031 	* NEWS: Update.
 9032 
 9033 2013-01-11  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9034 
 9035 	HACKING: suggest more checks before releasing
 9036 
 9037 	In particular, "make check-no-trailing-backslash-in-recipes",
 9038 	"make check-cc-no-c-o" and "make maintainer-check" should also
 9039 	be run.
 9040 
 9041 2013-01-11  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9042 
 9043 	tests: can fake a compiler not grasping "-c -o" -- globally in all tests
 9044 
 9045 	The ability to easily do so will be quite important in upcoming changes
 9046 	about C compilation handling and semantics of the 'subdir-objects'
 9047 	option.  Refer to the extensive discussion about automake bug#13378 for
 9048 	more details: <http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13378>.
 9049 
 9050 	See also commit 'v1.13.1-34-g744cd57' of 2013-01-08, "coverage: compile
 9051 	rules used "-c -o" also with losing compilers".
 9052 
 9053 	* t/ax/cc-no-c-o.in: New, a "C compiler" that chokes when the '-c' and
 9054 	'-o' options are passed together to it on the command line.
 9055 	* Makefile.am (t/ax/cc-no-c-o): Generate this script from it.
 9056 	(noinst_SCRIPTS, CLEANFILES): Add it.
 9057 	(EXTRA_DIST): Add 't/ax/cc-no-c-o.in'.
 9058 	(check-cc-no-c-o): New target, runs the whole testsuite with 'cc-no-c-o'
 9059 	as the C compiler (bot GNU and non-GNU).
 9060 	* .gitignore: Update.
 9061 	* t/ccnoco.sh: Use the new script instead of duplicating it.
 9062 	* t/ccnoco3.sh: Likewise.
 9063 	* t/ccnoco4.sh: Likewise.
 9064 	* t/self-check-cc-no-c-o.sh: New testsuite self-check.
 9065 	* t/list-of-tests.mk: Adjust.
 9066 
 9067 2013-01-11  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9068 
 9069 	sync: update files from upstream with "make fetch"
 9070 
 9071 	* lib/INSTALL: Update.
 9072 	* lib/config.guess: Likewise.
 9073 	* lib/config.sub: Likewise.
 9074 	* lib/gendocs_template: Likewise.
 9075 	* lib/gitlog-to-changelog: Likewise.
 9076 	* lib/gnupload: Likewise.
 9077 	* lib/texinfo.tex: Likewise.
 9078 	* lib/update-copyright: Likewise.
 9079 
 9080 2013-01-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9081 
 9082 	typofix: in comments in GNUmakefile
 9083 
 9084 2013-01-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9085 
 9086 	Rename 'maint/' -> 'maintainer/', for Git's sake
 9087 
 9088 	Otherwise, Git gets confused by the fact that a directory ('maint')
 9089 	is named like a branch, and forces me to tweak the command line to
 9090 	resolve the ambiguity for it.
 9091 
 9092 	* maint/: Rename ...
 9093 	* maintainer/: ... like this.
 9094 	* Makefile.am, GNUmakefile: Adjust.
 9095 
 9096 2013-01-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9097 
 9098 	HACKING: minor typofix
 9099 
 9100 2013-01-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9101 
 9102 	HACKING: bug-tracker, the PLANS directory, and how to plan "big" changes
 9103 
 9104 2013-01-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9105 
 9106 	HACKING: rewindable branches should live in the 'experimental/*' namespace
 9107 
 9108 2013-01-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9109 
 9110 	HACKING: fixlets about git branch rewinding policy
 9111 
 9112 2013-01-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9113 
 9114 	HACKING: commit messages are not to follow GCS ChangeLog rules too strongly
 9115 
 9116 2013-01-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9117 
 9118 	HACKING: "detailed explanation" in commit messages is almost mandatory
 9119 
 9120 2013-01-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9121 
 9122 	HACKING: we use "merge --log" even when merging master
 9123 
 9124 2013-01-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9125 
 9126 	HACKING: typofix
 9127 
 9128 2013-01-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9129 
 9130 	depend2.am: fix comments on verbosity of compilation rules
 9131 
 9132 	The situation and decisions described on those comments have become
 9133 	quite outdated since the introduction of the silent-rules support.
 9134 	Today, the general idea is to have nice, terse output if silent rules
 9135 	are enabled, and complete, faithful, very verbose output if they are
 9136 	not -- without trying to "massage" this verbose output in a more
 9137 	pleasant form if that would cause complication in the affected code.
 9138 
 9139 	So it's better to just drop the obsolescent comments.
 9140 
 9141 	Note that we don't start simplifying the existing rules according
 9142 	to this new philosophy; that will only be done when touching some
 9143 	existing code (for the 'depend2.am' code, that will probably happen
 9144 	on the master branch).
 9145 
 9146 2013-01-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9147 
 9148 	depend2.am: improve comments a little
 9149 
 9150 	* lib/am/depend2.am: The "fastdep" mode is supported not only for
 9151 	gcc 3.x, but for gcc 3.x or later, in particular, for all gcc in
 9152 	the 4.x series (at the time of writing, the latest release is 4.72).
 9153 	Adjust the comments to match, and re-wrap them while at it.
 9154 
 9155 2013-01-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9156 
 9157 	plans: automake 1.14 is to assume "rm -f" with no args is OK
 9158 
 9159 	See automake bug#10828.
 9160 
 9161 	* PLANS/rm-f-without-args.txt: New.
 9162 
 9163 2013-01-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9164 
 9165 	plans: we want to active subdir-objects unconditionally in automake 1.14
 9166 
 9167 	See automake bug#13378.
 9168 
 9169 	* PLANS/subdir-objects.txt: New.
 9170 	* t/ccnoco4.sh: Improve heading comments a little.
 9171 
 9172 2013-01-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9173 
 9174 	Merge branch 'maint'
 9175 
 9176 	* maint:
 9177 	  tests: adjust stale references to old test names
 9178 	  tests: rename the last aclocal test with dumb name
 9179 	  tests: fix an old botched change to an aclocal test
 9180 	  tests: fix some botched inter-test references in heading comments
 9181 	  coverage: compile rules used "-c -o" also with losing compilers
 9182 	  texi: remove extra verbosity in creation of dirstamp directory
 9183 	  coverage: user can avoid distributing '.info' pages
 9184 	  plans: add some on-going plans (already registered on the bug tracker)
 9185 	  docs: mention dist-hook help for EXTRA_DIST
 9186 	  texi: remove workaround for older Texinfo (4.1)
 9187 	  NEWS: improve wordings in entry deprecating suffix-less info files
 9188 	  plans: add the "PLANS" directory
 9189 
 9190 2013-01-08  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9191 
 9192 	tests: adjust stale references to old test names
 9193 
 9194 	* t/remake-renamed-m4-macro-and-file.sh: Adjust to reflect to old
 9195 	"acloca22 -> t/aclocal-deleted-header.sh" test rename.
 9196 	* t/aclocal-pr450.sh (configure.ac): Use '$me' in the AC_INIT call,
 9197 	instead of hard-coding the old name of this test, i.e., "acloca19".
 9198 
 9199 2013-01-08  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9200 
 9201 	tests: rename the last aclocal test with dumb name
 9202 
 9203 	* t/acloca10.sh: Rename ...
 9204 	* t/aclocal-I-install.sh: ... to this saner and slightly more
 9205 	self-explanatory name.
 9206 	* t/list-of-tests.mk: Adjust.
 9207 
 9208 2013-01-08  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9209 
 9210 	tests: fix an old botched change to an aclocal test
 9211 
 9212 	* t/acloca10.sh (configure.ac): Here, invoke the m4 macro 'MACRO2'
 9213 	before the macro 'MACRO1' (the related test 't/aclocal-I-order-2.sh'
 9214 	does the opposite).  This reverts a botched edit done (by myself,
 9215 	oops) in commit 'v1.11-1335-gefdc3e1' of 2011-09-11, "tests: minor
 9216 	optimizations/simplifications in some aclocal tests", and makes the
 9217 	behaviour of the test match once again what is stated in the
 9218 	heading comments.  While at it, improve those same heading comments
 9219 	a little.
 9220 
 9221 2013-01-08  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9222 
 9223 	tests: fix some botched inter-test references in heading comments
 9224 
 9225 	* t/aclocal-I-order-2.sh: Here.
 9226 	* t/aclocal-I-order-2.sh: And here.
 9227 
 9228 2013-01-08  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9229 
 9230 	coverage: compile rules used "-c -o" also with losing compilers
 9231 
 9232 	If the 'subdir-objects' option is used, Automake-generated rules for
 9233 	C compilation pass both the "-c" and "-o" options to the C compiler,
 9234 	*unconditionally*.  There are some compilers that choke on such an
 9235 	usage, but the AM_PROG_CC_C_O macro takes care of them (it does so by
 9236 	redefining $CC to use the Automake-provided 'compile' wrapper script
 9237 	automatically, if a losing compiler is detected at configure runtime).
 9238 
 9239 	Unfortunately, in case the 'subdir-objects' option is specified in a
 9240 	Makefile.am, but all the source files resided anyway in the top-level
 9241 	directory (relative to the Makefile.am), Automake do *not* complain
 9242 	if AM_PROG_CC_C_O wasn't invoked in 'configure.ac' -- all the while
 9243 	still passing "-c -o" to the compiler invocations.  This could cause
 9244 	compilation failures with losing compilers if the user forget to call
 9245 	AM_PROG_CC_C_O in 'configure.ac' (and Automake would not warn him of
 9246 	the issue).
 9247 
 9248 	Expose this bug in the testsuite.
 9249 
 9250 	Issue identified by Nick Bowler in the discussion on automake bug#13378:
 9251 	<http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13378#35>
 9252 	<http://debbugs.gnu.org/cgi/bugreport.cgi?bug=13378#44>
 9253 
 9254 	* t/ccnoco4.sh: New test.
 9255 	* t/list-of-tests.mk (XFAIL_TESTS, handwritten_TESTS): List it.
 9256 
 9257 2013-01-07  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9258 
 9259 	texi: no longer support split info files
 9260 
 9261 	See automake bug#13351 and bug#12320.
 9262 
 9263 	Automake-generated recipes have for a long time supported "split"
 9264 	info files:
 9265 
 9266 	  http://www.gnu.org/software/texinfo/manual/texinfo/texinfo.html#Tag-and-Split-Files
 9267 
 9268 	When I asked the rationale for this feature:
 9269 
 9270 	  http://lists.gnu.org/archive/html/texinfo-devel/2012-08/msg00015.html
 9271 
 9272 	Karl Berry confirmed that the reason for its existence was indeed
 9273 	"efficiency, especially memory size":
 9274 
 9275 	  http://lists.gnu.org/archive/html/texinfo-devel/2012-08/msg00024.html
 9276 
 9277 	He also added that "The Elisp manual is one of the largest ones around.
 9278 	Looks like it would be maybe 3.5mb as one file."  Not in any way big by
 9279 	modern standards.
 9280 
 9281 	OTOH, it appears that the use of split info files (at least in the way
 9282 	they have been handled by Automake-generated rules for a long time) can
 9283 	cause real problems in some (admittedly quite corner-case) situations:
 9284 
 9285 	  http://thread.gmane.org/gmane.comp.parsers.bison.bugs/3963
 9286 	  http://debbugs.gnu.org/cgi/bugreport.cgi?bug=12320
 9287 
 9288 	So we now follow suit with Automake-NG (see commit v1.12.2-901-gdd603e2,
 9289 	<http://lists.gnu.org/archive/html/automake-ng/2012-08/msg00147.html>)
 9290 	and have Automake-generated makefiles pass the '--no-split' option
 9291 	unconditionally to makeinfo invocations.
 9292 
 9293 	This allow some nice simplifications in our Texinfo recipes, and offer
 9294 	an automatic fix for bug#12320.
 9295 
 9296 	Another *very* good aspect of such a change is that it should be 100%
 9297 	transparent to the Automake users.
 9298 
 9299 	* lib/am/texinfos.am: Simplify moderately.
 9300 	* lib/am/texibuild.am: Simplify greatly the recipe for the creation
 9301 	of info files.
 9302 	* t/txinfo-makeinfo-error-no-clobber.sh: Adjust.
 9303 	* t/txinfo-no-split.sh: New test.
 9304 	* t/list-of-tests.mk: Add it.
 9305 
 9306 2013-01-05  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9307 
 9308 	texi: remove extra verbosity in creation of dirstamp directory
 9309 
 9310 	* lib/am/texi-vers.am (%STAMPVTI%): Here.
 9311 
 9312 2013-01-05  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9313 
 9314 	coverage: user can avoid distributing '.info' pages
 9315 
 9316 	Can be done like this:
 9317 
 9318 	        AUTOMAKE_OPTIONS = info-in-builddir
 9319 	        dist-info:
 9320 	            @:
 9321 
 9322 	Note that this usage is not yet documented: we might decide to go
 9323 	for a fully-fledged 'no-dist-info' flag, or something like that, in
 9324 	future automake version (this is not yet decided); in which case,
 9325 	it's better not to have people start to rely on the hack above.
 9326 	Still, there's no good reason to break it gratuitously, hence this
 9327 	test coverage.
 9328 
 9329 	* t/txinfo-nodist-info.sh: New test.
 9330 	* t/list-of-tests.mk: Add it.
 9331 
 9332 2013-01-04  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9333 
 9334 	texi: remove leftover support for DGJPP '.iNN' info files
 9335 
 9336 	A follow-up to commit 'v1.13-6-gad08bbf'.
 9337 
 9338 	* lib/am/texinfos.am (dist-info): Stop catering to info files
 9339 	with '.iNN' extension (as in '.i23' or '.i02').
 9340 
 9341 2013-01-04  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9342 
 9343 	Merge branch 'plans' into maint
 9344 
 9345 	* plans:
 9346 	  plans: add some on-going plans (already registered on the bug tracker)
 9347 
 9348 2013-01-04  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9349 
 9350 	plans: add some on-going plans (already registered on the bug tracker)
 9351 
 9352 2013-01-04  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9353 
 9354 	Merge branch 'plans' into maint
 9355 
 9356 	* plans:
 9357 	  plans: add the "PLANS" directory
 9358 
 9359 
 9360 	Conflicts:
 9361 		Makefile.am
 9362 
 9363 2013-01-03  Karl Berry  <karl@gnu.org>
 9364 
 9365 	docs: mention dist-hook help for EXTRA_DIST
 9366 
 9367 	* automake.texi (Basics of Distribution): mention dist-hook
 9368 	as working around the problems of whole directories in EXTRA_DIST.
 9369 
 9370 2013-01-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9371 
 9372 	texi: remove workaround for older Texinfo (4.1)
 9373 
 9374 	* lib/am/texibuild.am: Here, in the rules generating HTML output.
 9375 	We can do so because, since Automake 1.13, we require Texinfo >= 4.9
 9376 	anyway.
 9377 
 9378 	Basically a backport of Automake-NG commit '1.12.2-879-ge6caf5e'.
 9379 
 9380 2013-01-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9381 
 9382 	NEWS: improve wordings in entry deprecating suffix-less info files
 9383 
 9384 2013-01-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9385 
 9386 	Merge branch 'maint'
 9387 
 9388 	* maint:
 9389 	  build: don't enable 'color-tests' automake option explicitly
 9390 	  build: enable all warnings as fatal in our own build system
 9391 	  texi: Texinfo sources and CLEANFILES definition should co-exist peacefully
 9392 	  tests: make two new test executable
 9393 	  runtest: better command line API
 9394 	  tests: move runtest.in away from the top-lever directory
 9395 	  maint: move more maintainer files in the 'maint/' subdir
 9396 	  tests: more significant names for some tests
 9397 	  maint: add some of my maintainer-specific scripts
 9398 	  texi: deprecate hack about info files in CLEANFILES variables
 9399 	  texi: info files can be generated in the builddir
 9400 
 9401 2013-01-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9402 
 9403 	build: don't enable 'color-tests' automake option explicitly
 9404 
 9405 	It's enabled by default since commit 'v1.12.2-136-g2d5571e' (this
 9406 	change appeared in Automake 1.13).
 9407 
 9408 	* configure.ac (AM_INIT_AUTOMAKE): Drop 'color-tests'.
 9409 
 9410 2013-01-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9411 
 9412 	build: enable all warnings as fatal in our own build system
 9413 
 9414 	Automake should of course be able to bootstrap itself in a
 9415 	warning-free manner w.r.t. the Autotools.  So make any failure
 9416 	to do so fatal.  Not doing so caused the regression fixed by
 9417 	previous commit 'v1.13.1-22-ga790fae' to go unnoticed.
 9418 
 9419 	* configure.ac (AM_INIT_AUTOMAKE): Add '-Werror' and '-Wall'.
 9420 	* bootstrap.sh: Pass the '-Wall -Werror' options to aclocal,
 9421 	automake and autoconf invocations.
 9422 
 9423 2013-01-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9424 
 9425 	texi: Texinfo sources and CLEANFILES definition should co-exist peacefully
 9426 
 9427 	But they don't now, due to a regression introduced in commit
 9428 	'v1.13.1-4-gc1a8f56'.  Fix it.  The regression was hitting our
 9429 	own build system!
 9430 
 9431 	* automake.in (handle_texinfo_helper): Only complain if the
 9432 	'info-in-builddir' is not active and a '.info' file (not any
 9433 	random file!) is listed in CLEANFILES or DISTCLEANFILES.
 9434 
 9435 2013-01-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9436 
 9437 	tests: make two new test executable
 9438 
 9439 	* t/backslash-issues.sh: This.
 9440 	* t/extra-data.sh: And this.
 9441 
 9442 	Issue revealed by the 'sc_tests_executable' maintainer check.
 9443 
 9444 2013-01-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9445 
 9446 	Merge branch 'info-in-builddir' into maint
 9447 
 9448 	* info-in-builddir:
 9449 	  texi: deprecate hack about info files in CLEANFILES variables
 9450 	  texi: info files can be generated in the builddir
 9451 
 9452 2013-01-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9453 
 9454 	runtest: better command line API
 9455 
 9456 	* t/ax/runtest.in: Accept options '-k' and '--keep-testdirs' (same
 9457 	as exporting '$keep_testdirs' to "yes").  To improve compatibility
 9458 	with the "make check" interface, allow environment variables to be
 9459 	passes on the command line.  Minor adjustments while at it.
 9460 
 9461 2013-01-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9462 
 9463 	tests: move runtest.in away from the top-lever directory
 9464 
 9465 	Not only this leaves the top-lever directory less cluttered, but
 9466 	helps in keeping the testsuite-related files more "centralized".
 9467 
 9468 	* runtest.in: Move ...
 9469 	* t/ax/runtest.in: ... here.  While at it, add customary
 9470 	'@configure_input@' comment line.
 9471 	* Makefile.am (runtest, EXTRA_DIST): Adjust.
 9472 
 9473 2013-01-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9474 
 9475 	maint: move more maintainer files in the 'maint/' subdir
 9476 
 9477 	* maint.mk: Move ...
 9478 	* maint/maint.mk: ... here.
 9479 	* syntax-checks.mk: Move ...
 9480 	* maint/syntax-checks.mk: ... here.
 9481 	* Makefile.am: Adjust.
 9482 	* GNUmakefile: Likewise.
 9483 
 9484 2013-01-03  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9485 
 9486 	plans: add the "PLANS" directory
 9487 
 9488 	Individual files or sub-directories about future and on-going
 9489 	development plans in Automake will be added in follow-up commits.
 9490 
 9491 	This new set of documents is meant to help ensure a more controlled
 9492 	and smooth development and evolution for Automake, in several ways.
 9493 
 9494 	 - Having the plans clearly spelled out should will avoid messy
 9495 	   roadmaps with no clear way forward or with muddy or ill-defined
 9496 	   aims or purposes; a trap this is too easy to fall into.
 9497 
 9498 	 - Keeping planned changes cooking and re-hashed for a while should
 9499 	   ensure rough edges are smoothed up, transitions are planned in a
 9500 	   proper way (hopefully avoiding debacles like the AM_MKDIR_PROG_P
 9501 	   deprecation and the AM_CONFIG_HEADER too-abrupt removal), and
 9502 	   "power users" have more chances of getting informed in due time,
 9503 	   thus having all the time to prepare for the changes or raise
 9504 	   objections against them.
 9505 
 9506 	 - Having the plans clearly stated and registered in a "centralized"
 9507 	   location should make it more difficult to them to slip through
 9508 	   the cracks, getting forgotten or (worse) only half-implemented.
 9509 
 9510 	 - Even for discussions and plans registered on the Bug Tracker
 9511 	   as well, a corresponding entry in the PLANS directory can help
 9512 	   in keeping main ideas summarized, and consensus and/or objections
 9513 	   registered and easily compared.
 9514 
 9515 	Motivation:
 9516 	<http://blog.flameeyes.eu/2013/01/autotools-mythbuster-automake-pains>
 9517 	Not a flatting picture for us (and maybe a little too harsh), but
 9518 	basically true and even spot-on in some regards.
 9519 
 9520 	* PLANS/README: New.
 9521 	* Makefile.am (EXTRA_DIST): Distribute the whole PLANS directory.
 9522 
 9523 2013-01-02  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9524 
 9525 	tests: more significant names for some tests
 9526 
 9527 	* t/spy.sh: Rename...
 9528 	* t/spy-double-colon.sh: ... like this.
 9529 	* t/yacc4.sh: Rename...
 9530 	* t/yacc-misc.sh: ... like this.
 9531 	* t/yaccdry.sh: Rename...
 9532 	* t/yacc-dry.sh: ... like this.
 9533 	* t/yaccpp.sh: Rename...
 9534 	* t/yacc-cxx-grepping.sh: ... like this.
 9535 	* t/yaccvpath.sh: Rename...
 9536 	* t/yacc-vpath.sh: ... like this.
 9537 
 9538 2013-01-02  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9539 
 9540 	maint: add some of my maintainer-specific scripts
 9541 
 9542 	They are likely not general enough for widespread use, but they
 9543 	are useful nonetheless.
 9544 
 9545 	In the best-case scenario, they will start to be used by other
 9546 	people, and thus accordingly improved and made more general and
 9547 	flexible.
 9548 
 9549 	In the worst case scenario, well, I still get to keep them in a
 9550 	centralized, blessed place, simplifying the deployment and use
 9551 	of them; so still a win for me :-)
 9552 
 9553 	* maint/am-ft: New script.
 9554 	* maint/am-xft: Likewise.
 9555 	* maint/rename-tests: Likewise.
 9556 	* Makefile.am (EXTRA_DIST): Add them.
 9557 
 9558 2013-01-02  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9559 
 9560 	Merge branch 'maint'
 9561 
 9562 	* maint:
 9563 	  tests: reorganize tests on backslash issues
 9564 	  style: add trailing ':' to some test cases
 9565 	  tests: tweak tests on obsolete EXTRA_DATA variable
 9566 	  tests: more significant names for some tests
 9567 	  cosmetics: remove few occurrences of trailing whitespace
 9568 	  docs: re-introduce mention of two-args AM_INIT_AUTOMAKE invocation
 9569 	  texi: warn against '.txi' and '.texinfo' input suffixes
 9570 	  cleanup: remove two lines of dead code in automake
 9571 	  texi: warn against suffix-less info files
 9572 	  build: respect silent rules in generation of "amhello" example tarball
 9573 
 9574 2013-01-02  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9575 
 9576 	tests: reorganize tests on backslash issues
 9577 
 9578 	* t/backsl.sh, t/backsl2.sh, t/backsl3.sh: Merge ...
 9579 	* t/backslash-issues.sh: ... into this test.
 9580 	* t/backsl4.sh: Rename ...
 9581 	* t/backslash-before-trailing-whitespace.sh: ... like this.
 9582 	* t/list-of-tests.mk: Adjust.
 9583 
 9584 2013-01-02  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9585 
 9586 	style: add trailing ':' to some test cases
 9587 
 9588 2013-01-02  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9589 
 9590 	tests: tweak tests on obsolete EXTRA_DATA variable
 9591 
 9592 	* t/extra3.sh, t/extra4.sh: Merge ....
 9593 	* t/extra-data.sh: ... into this, with updated comments.
 9594 	* t/list-of-tests.mk: Adjust.  Also Tweak the order in
 9595 	which some other tests are listed.
 9596 
 9597 2013-01-02  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9598 
 9599 	tests: more significant names for some tests
 9600 
 9601 	* t/extra5.sh: Rename ....
 9602 	* t/extra-dist-vpath-dir.sh: ... like this.
 9603 	* t/extra6.sh: Rename ....
 9604 	* t/extra-dist-dirs-and-subdirs.sh: ... like this.
 9605 	* t/extra7.sh: Rename ....
 9606 	* t/extra-dist-vpath-dir-merge.sh: ... like this.
 9607 	* t/extra8.sh: Rename ....
 9608 	* t/extra-programs-misc.sh: ... like this.
 9609 	* t/extra9.sh: Rename ....
 9610 	* t/extra-programs-and-libs.sh: ... like this.
 9611 	* t/extra10.sh: Rename ....
 9612 	* t/extra-dist-wildcards.sh: ... like this.
 9613 	* t/extra11.sh: Rename ....
 9614 	* t/extra-dist-wildcards-gnu.sh: ... like this.
 9615 	* t/extra12.sh: Rename ....
 9616 	* t/extra-dist-wildcards-vpath.sh: ... like this.
 9617 	* t/extradep.sh: Rename ....
 9618 	* t/extra-deps.sh: ... like this, and adjust heading comments.
 9619 	* t/extradep2.sh: Rename ....
 9620 	* t/extra-deps-lt.sh: ... like this, and adjust heading comments.
 9621 	* t/list-of-tests.mk: Adjust.
 9622 
 9623 2013-01-02  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9624 
 9625 	cosmetics: remove few occurrences of trailing whitespace
 9626 
 9627 2013-01-02  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9628 
 9629 	docs: re-introduce mention of two-args AM_INIT_AUTOMAKE invocation
 9630 
 9631 	Albeit obsolescent and raising warnings in the 'obsolete' category,
 9632 	that usage is still supported, and will need to be until Autoconf
 9633 	improves its handling of configure-time-generated package version
 9634 	numbers.  So it's better to explicitly document it again, stating
 9635 	that it is obsoleted but still working (and why), rather then leaving
 9636 	it as Yet Another Undocumented Feature (that will mysteriously and
 9637 	suddenly break some random day in the future).
 9638 
 9639 	It's worth giving some background about how we ended up in the
 9640 	situation that this patch fixes.
 9641 
 9642 	We had originally removed support for the long-deprecated two-args
 9643 	AM_INIT_AUTOMAKE invocation (see commit v1.12-67-ge186355).  Before
 9644 	that removal could land in a released Automake version, Bob Friesenhahn
 9645 	made a quite compelling point that the two-args AM_INIT_AUTOMAKE
 9646 	invocation could still be useful for modern, maintained packages like
 9647 	GraphicsMagick, at least until Autoconf is fixed to offer better support
 9648 	for "dynamic" package versions (see commit v1.12.2-245-g2abe183 for more
 9649 	in-depth rationales and references).  However, in that commit we didn't
 9650 	revert the removal of the *documentation* for this two-arguments
 9651 	AM_INIT_AUTOMAKE invocation (and no rationale for not doing so was given
 9652 	in the commit message).  Time to remedy that.
 9653 
 9654 	Indirectly suggested by Diego Elio Pattenò:
 9655 	<http://blog.flameeyes.eu/2013/01/autotools-mythbuster-automake-pains>
 9656 
 9657 	* doc/automake.texi: Adjust.
 9658 	* NEWS: Update.
 9659 	* THANKS: Likewise.
 9660 
 9661 2013-01-02  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9662 
 9663 	texi: warn against '.txi' and '.texinfo' input suffixes
 9664 
 9665 	The warning being in the 'obsolete' category.  This is mostly to
 9666 	ease transition to Automake-NG (see commit v1.12.1-416-gd5459b9),
 9667 	and to discourage use of seldom-tested setups.
 9668 
 9669 	* automake.in (handle_texinfo_helper): Warn against Texinfo input
 9670 	files with '.txi' or '.texinfo' suffixes.
 9671 	* NEWS: Update.
 9672 	* t/txinfo-other-suffixes.sh: Adjust and enhance.
 9673 	* t/txinfo-no-repeated-targets.sh: No longer use '.txi' and
 9674 	'.texinfo' extensions.
 9675 
 9676 2013-01-02  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9677 
 9678 	cleanup: remove two lines of dead code in automake
 9679 
 9680 	* automake.in (scan_texinfo_file): Here, the definition and
 9681 	munging of '$infobase'.
 9682 
 9683 2013-01-02  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9684 
 9685 	texi: warn against suffix-less info files
 9686 
 9687 	The warning being in the 'obsolete' category.  This is mostly to
 9688 	ease transition to Automake-NG (see commit v1.12.1-392-ga0c7b6a),
 9689 	and to discourage use of seldom-tested setups.
 9690 
 9691 	* automake.in (scan_texinfo_file): Warn against '@setfilename'
 9692 	directives that specify suffix-less output info files.
 9693 	* t/txinfo-without-info-suffix.sh: Adjust and enhance.
 9694 	* t/txinfo-makeinfo-error-no-clobber.sh: No longer use suffix-less
 9695 	info files in '@setfilename' directives.
 9696 	* t/primary-prefix-valid-couples.sh: Likewise.
 9697 	* t/txinfo-setfilename-repeated.sh: Likewise.
 9698 	* t/txinfo-vtexi2.sh : Likewise.
 9699 	* t/mdate2.sh: Likewise.
 9700 	* NEWS: Update.
 9701 
 9702 2013-01-02  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9703 
 9704 	texi: deprecate hack about info files in CLEANFILES variables
 9705 
 9706 	For quite a long time, Automake has been implementing an undocumented
 9707 	hack which ensured that '.info' files which appeared to be cleaned (by
 9708 	e.g. being listed in the CLEANFILES or DISTCLEANFILES variables) were
 9709 	built in the builddir rather than in the srcdir; this hack was introduced
 9710 	to ensure better backward-compatibility with packages such as Texinfo,
 9711 	which did things like:
 9712 
 9713 	    info_TEXINFOS = texinfo.txi info-stnd.texi info.texi
 9714 	    DISTCLEANFILES = texinfo texinfo-* info*.info*
 9715 	    # Do not create info files for distribution.
 9716 	    dist-info:
 9717 	        @:
 9718 
 9719 	in order not to distribute .info files.
 9720 
 9721 	Now that we have the 'info-in-builddir' option that explicitly causes
 9722 	generated '.info' files to be placed in the builddir, this hack should
 9723 	be longer necessary, so we deprecate it with runtime warnings.  It is
 9724 	scheduled to be removed altogether in Automake 1.14.
 9725 
 9726 	* automake.in (handle_texinfo_helper): Raise proper runtime warnings
 9727 	if the hack is triggered.
 9728 	* NEWS: Update.
 9729 	* t/txinfo28.sh: Adjust.
 9730 	* t/txinfo23.sh: Likewise.
 9731 	* t/txinfo25.sh: Adjust and extend.
 9732 	* t/txinfo24.sh: Likewise.
 9733 
 9734 2013-01-01  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9735 
 9736 	texi: info files can be generated in the builddir
 9737 
 9738 	User can now ask info files to be built in the $(builddir), rather than
 9739 	the $(srcdir), by specifying the Automake option 'info-in-builddir'.
 9740 	This feature was requested by the developers of GCC, GDB, GNU binutils
 9741 	and the GNU bfd library.  See the extensive discussion about automake
 9742 	bug#11034 for more details.
 9743 
 9744 	OK, to be honest, having '.info' files built in the builddir was
 9745 	*already* possible, but only using ugly and undocumented hacks involving
 9746 	definition of the CLEANFILES and/or DISTCLEANFILES.  For example, the
 9747 	binutils project did something like this in the relevant 'Makefile.am':
 9748 
 9749 	    # Automake 1.9 will only build info files in the objdir if they are
 9750 	    # mentioned in DISTCLEANFILES.  It doesn't have to be unconditional,
 9751 	    # though, so we use a bogus condition.
 9752 	    if GENINSRC_NEVER
 9753 	    DISTCLEANFILES = binutils.info
 9754 	    endif
 9755 
 9756 	See also the extensive discussion about automake bug#11034; in
 9757 	particular, the following messages:
 9758 	<http://debbugs.gnu.org/cgi/bugreport.cgi?bug=11034#65>
 9759 	<http://debbugs.gnu.org/cgi/bugreport.cgi?bug=11034#80>
 9760 	<http://debbugs.gnu.org/cgi/bugreport.cgi?bug=11034#86>
 9761 	<http://debbugs.gnu.org/cgi/bugreport.cgi?bug=11034#101>
 9762 
 9763 	* lib/Automake/Options.pm (_is_valid_easy_option): Recognize the
 9764 	new 'info-in-builddir' option.
 9765 	* automake.in (handle_texinfo_helper): If that option is set,
 9766 	initialize '$insrc' to '0', so that info files will be generated
 9767 	in the builddir.  Adjust comments to match.
 9768 	* t/txinfo-builddir.sh: New test.
 9769 	* t/list-of-tests.mk: Add it.
 9770 	* NEWS: Update.
 9771 	* doc/automake.texi: Document the new options.
 9772 
 9773 2013-01-01  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9774 
 9775 	build: respect silent rules in generation of "amhello" example tarball
 9776 
 9777 	* Makefile.am ($(srcdir)/doc/amhello-1.0.tar.gz): Here.
 9778 
 9779 2013-01-01  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9780 
 9781 	Merge branch 'maint'
 9782 
 9783 	* maint:
 9784 	  NEWS: fixlets and updates
 9785 	  post-release: minor version bump (1.13.1a)
 9786 	  release: stable minor release 1.13.1
 9787 
 9788 2013-01-01  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9789 
 9790 	NEWS: fixlets and updates
 9791 
 9792 2013-01-01  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9793 
 9794 	post-release: minor version bump (1.13.1a)
 9795 
 9796 	* configure.ac, m4/amversion.m4: Bump version to 1.13.1a.
 9797 
 9798 2013-01-01  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9799 
 9800 	release: stable minor release 1.13.1
 9801 
 9802 	* configure.ac (AC_INIT): Bump version number to 1.13.1.
 9803 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
 9804 
 9805 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9806 
 9807 	Merge branch 'maint'
 9808 
 9809 	* maint:
 9810 	  tests: don't always look for a C++ compiler named 'RCC'
 9811 	  tests: fix bug in pkg-config-macros.sh, could cause spurious SKIPs
 9812 	  docs: autom4te cache can break CONFIGURE_DEPENDENCIES
 9813 	  coverage: test CONFIGURE_DEPENDENCIES
 9814 	  coverage: test CONFIG_STATUS_DEPENDENCIES
 9815 
 9816 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9817 
 9818 	tests: don't always look for a C++ compiler named 'RCC'
 9819 
 9820 	On MacOS X (10.8), since the file system is case-insensitive, RCC
 9821 	can point to the "Resource Compiler" of the Qt4 Toolkit:
 9822 
 9823 	    <http://doc.qt.digia.com/4.2/rcc.html>
 9824 
 9825 	That mismatch causes our configure script to erroneously think that
 9826 	no working C++ compiler is present, and that is thus necessary to
 9827 	skip all the test cases requiring such a compiler.
 9828 
 9829 	So only look for a compiler named 'RCC' if the file system is
 9830 	case-sensible.
 9831 
 9832 	Issue spotted analyzing the testsuite logs reported in bug#13317.
 9833 
 9834 	* configure.ac: Adjust.
 9835 
 9836 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9837 
 9838 	tests: fix bug in pkg-config-macros.sh, could cause spurious SKIPs
 9839 
 9840 	Issue spotted perusing the testsuite logs reported in bug#13317.
 9841 
 9842 	* t/pkg-config-macros.sh: Don't use (uninitialized) '$dir' where '$d'
 9843 	should have been used instead.  Set IFS to ':' before looping on the
 9844 	$PATH expansion.  Fix typo: 'alocal' instead of 'aclocal'.  These
 9845 	issues were causing the location in PATH of the 'pkg-config' program
 9846 	not to be found even when the program was present.
 9847 	* THANKS: Update.
 9848 
 9849 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9850 
 9851 	docs: autom4te cache can break CONFIGURE_DEPENDENCIES
 9852 
 9853 	In fact, the new test 'remake-configure-dependencies.sh' only
 9854 	works because it disables the autom4te cache :-(  That is more
 9855 	of an AUtotools design issue than an Automake bug, so better
 9856 	just document it ...
 9857 
 9858 	* doc/automake.texi (Rebuilding): ... here.
 9859 
 9860 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9861 
 9862 	coverage: test CONFIGURE_DEPENDENCIES
 9863 
 9864 	* t/remake-configure-dependencies.sh: New test.
 9865 	* t/list-of-tests.mk: Add it.
 9866 	* doc/automake.texi: Reference it in comments.
 9867 
 9868 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9869 
 9870 	coverage: test CONFIG_STATUS_DEPENDENCIES
 9871 
 9872 	* t/remake-config-status-dependencies.sh: New test.
 9873 	* t/list-of-tests.mk: Add it.
 9874 	* doc/automake.texi: Reference it in comments.
 9875 
 9876 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9877 
 9878 	Merge branch 'maint'
 9879 
 9880 	* maint:
 9881 	  tests: fix reference to related tests in some comments
 9882 	  tests: some enhancements to texinfo tests (from Automake-NG)
 9883 	  tests: fix reference to sister tests in two heading comments
 9884 
 9885 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9886 
 9887 	tests: fix reference to related tests in some comments
 9888 
 9889 	* t/acloca10.sh: Here.
 9890 	* t/check2.sh: And here.
 9891 	* t/remake-deleted-m4-file.sh: And here.
 9892 
 9893 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9894 
 9895 	tests: some enhancements to texinfo tests (from Automake-NG)
 9896 
 9897 	* t/txinfo-many-output-formats.sh: Backport improvements, almost verbatim,
 9898 	from Automake-NG commit 'v1.13-768-gb434acc' (plus minor tweaks for the
 9899 	sake of non-GNU make implementation).
 9900 	* t/txinfo-many-output-formats-vpath.sh: New test, backported almost
 9901 	verbatim from that same Automake-NG commit (again, with minor tweaks for
 9902 	the sake of non-GNU make implementations).
 9903 	* t/lsit-of-tests.mk: Adjust.
 9904 
 9905 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9906 
 9907 	tests: fix reference to sister tests in two heading comments
 9908 
 9909 	* t/color-tests.sh: Here.
 9910 	* t/color-tests2.sh: And here.
 9911 
 9912 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9913 
 9914 	maint: update copyright year for 2013 (in branch master)
 9915 
 9916 	* t/includes-no-more.sh: Only here (all other files already touched in maint)
 9917 
 9918 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9919 
 9920 	Merge branch 'maint'
 9921 
 9922 	* maint:
 9923 	  maint: update copyright year for 2013 (in branch maint)
 9924 
 9925 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9926 
 9927 	maint: update copyright year for 2013 (in branch maint)
 9928 
 9929 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9930 
 9931 	Merge branch 'maint'
 9932 
 9933 	* maint:
 9934 	  docs: adjust comment w.r.t. recent tests renaming
 9935 	  tests: some more renames for texinfo tests
 9936 
 9937 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9938 
 9939 	docs: adjust comment w.r.t. recent tests renaming
 9940 
 9941 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9942 
 9943 	tests: some more renames for texinfo tests
 9944 
 9945 	* t/txinfo.sh: Rename ...
 9946 	* t/txinfo-tex-dist.sh: ... like this.
 9947 	* t/txinfo10.sh: Rename ...
 9948 	* t/txinfo-dvi-recurs.sh: ... like this.
 9949 	* t/txinfo20.sh: Rename ...
 9950 	* t/txinfo-makeinfo-error-no-clobber.sh: ... like this.
 9951 	* t/txinfo-setfilename-suffix-match.sh: Rename ...
 9952 	* t/txinfo-setfilename-suffix-strip.sh: ... like this.
 9953 	* t/txinfo-suffix-less-info.sh: Rename ...
 9954 	* t/txinfo-without-info-suffix.sh: ... like this.
 9955 	* t/vtexi.sh: Rename ...
 9956 	* t/txinfo-vtexi.sh: ... like this.
 9957 	* t/vtexi2.sh: Rename ...
 9958 	* t/txinfo-vtexi2.sh: ... like this.
 9959 	* t/vtexi3.sh: Rename ...
 9960 	* t/txinfo-vtexi3.sh: ... like this, and adjust heading comments.
 9961 	* t/vtexi4.sh: Rename ...
 9962 	* t/txinfo-vtexi4.sh: ... like this, and adjust heading comments.
 9963 	* t/list-of-tests.mk: Adjust.
 9964 
 9965 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9966 
 9967 	Merge branch 'maint'
 9968 
 9969 	* maint:
 9970 	  tests: enhance and tweak some texinfo tests
 9971 	  tests: remove a redundant texinfo test
 9972 	  tests: don't list a removed tests in XFAIL_TESTS
 9973 	  tests: more explicative names for some tests
 9974 	  cleanup: drop a redundant %?FIRST% transform
 9975 	  maint: state when compatibility cruft should be removed or warned about
 9976 
 9977 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9978 
 9979 	tests: enhance and tweak some texinfo tests
 9980 
 9981 	* t/txinfo-subdir-pr343.sh: Move the tests checking that '.info.bak'
 9982 	files in subdirs are not unduly distributed into ...
 9983 	* t/txinfo-no-extra-dist.sh: ... in here.  Enhance a little while
 9984 	at it.
 9985 	* t/txinfo-other-suffixes.sh: Also check that the '.texinfo' suffix
 9986 	is accepted and works.
 9987 	* t/txinfo-setfilename-suffix-match.sh: Remove useless call to
 9988 	autoconf, enhance a little.
 9989 
 9990 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
 9991 
 9992 	tests: remove a redundant texinfo test
 9993 
 9994 	* t/txinfo-add-missing.sh: This one, its checks being already covered
 9995 	by the more extensive 't/add-missing.tap' test.
 9996 	* t/txinfo-add-missing2.sh: Rename ...
 9997 	* t/txinfo-add-missing-and-dist.sh: Like this.
 9998 	* t/list-of-tests.mk: Adjust.
 9999 
10000 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
10001 
10002 	tests: don't list a removed tests in XFAIL_TESTS
10003 
10004 	* t/list-of-tests.mk (XFAIL_TESTS): No longer list 't/txinfo5.sh' here,
10005 	it has been removed together with the Cygnus support.
10006 
10007 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
10008 
10009 	tests: more explicative names for some tests
10010 
10011 	Do so for several Texinfo-related tests.  Note that some of such tests
10012 	with suboptimal names are *not* renamed; that's because they are going
10013 	to be removed in the master branch anyway (since they check from some
10014 	soon-to-be-removed features), so renaming them would be wasted work,
10015 	and could even create useless merge conflicts.
10016 
10017 	* t/txinfo2.sh: Rename ...
10018 	* t/txinfo-no-extra-dist.sh: ... like this.
10019 	* t/txinfo3.sh: Rename ...
10020 	* t/txinfo-suffix-less-info.sh: ... like this.
10021 	* t/txinfo4.sh: Rename ...
10022 	* t/txinfo-unrecognized-info-suffix.sh: ... like this.
10023 	* t/txinfo6.sh: Rename ...
10024 	* t/txinfo-other-suffixes.sh: ... like this.
10025 	* t/txinfo7.sh: Rename ...
10026 	* t/txinfo-add-missing.sh: ... like this.
10027 	* t/txinfo8.sh: Rename ...
10028 	* t/txinfo-add-missing2.sh: ... like this.
10029 	* t/txinfo9.sh: Rename ...
10030 	* t/txinfo-no-repeated-targets.sh: ... like this.
10031 	* t/txinfo17.sh: Rename ...
10032 	* t/txinfo-setfilename-repeated.sh: ... like this.
10033 	* t/txinfo13.sh: Rename ...
10034 	* t/txinfo-subdir-pr343.sh: ... like this.
10035 	* t/txinfo16.sh: Rename ...
10036 	* t/txinfo-info-in-srcdir.sh: ... like this.
10037 	* t/txinfo21.sh: Rename ...
10038 	* t/txinfo-many-output-formats.sh: ... like this.
10039 	* t/txinfo22.sh: Rename ...
10040 	* t/txinfo-override-texinfo-tex.sh: ... like this.
10041 	* t/txinfo26.sh: Rename ...
10042 	* t/txinfo-absolute-srcdir-pr408.sh: ... like this.
10043 	* t/txinfo27.sh: Rename ...
10044 	* t/txinfo-no-installinfo.sh: ... like this.
10045 	* t/txinfo29.sh: Rename ...
10046 	* t/txinfo-override-infodeps.sh: ... like this.
10047 	* t/txinfo31.sh: Rename ...
10048 	* t/txinfo-setfilename-suffix-match.sh: ... like this.
10049 	* t/txinfo32.sh: Rename ...
10050 	* t/txinfo-bsd-make-recurs.sh: ... like this.
10051 	* t/txinfo33.sh: Rename ...
10052 	* t/txinfo-clean.sh: ... like this.
10053 	* t/list-of-tests.mk: Adjust.
10054 
10055 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
10056 
10057 	cleanup: drop a redundant %?FIRST% transform
10058 
10059 	* lib/am/texinfos.am: Here: this file is only included one time
10060 	per Makefile by Automake.
10061 
10062 	Cherry picked from Automake-NG commit 'v1.12.2-864-g5c580b0'.
10063 
10064 2012-12-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
10065 
10066 	maint: state when compatibility cruft should be removed or warned about
10067 
10068 	* automake.in: Here.
10069 	* lib/Automake/Options.pm: And here.
10070 	* m4/obsolete-err.m4: And here.
10071 
10072 2012-12-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
10073 
10074 	Merge branch 'maint'
10075 
10076 	* maint:
10077 	  fixup: copyright yars in two recent tests
10078 	  obsolete: fixup: don't use AU_DEFUN where AC_DEFUN is required
10079 
10080 2012-12-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
10081 
10082 	fixup: copyright yars in two recent tests
10083 
10084 2012-12-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
10085 
10086 	obsolete: fixup: don't use AU_DEFUN where AC_DEFUN is required
10087 
10088 	* m4/obsolete-err.m4: Here, for the definition of 'AM_PROG_CC_STDC'
10089 
10090 2012-12-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
10091 
10092 	obsolete: better error message if AM_PROG_MKDIR_P is used
10093 
10094 	In Automake 1.14, the long-deprecated macro AM_PROG_MKDIR_P will be
10095 	removed.  Instead of simply dropping the macro definition, which would
10096 	cause unclear error messages in aclocal and autoconf:
10097 
10098 	    configure.ac:4: warning: macro 'AM_PROG_MKDIR_P' not found in library
10099 	    configure.ac:4: error: possibly undefined macro: AM_PROG_MKDIR_P
10100 	      If this token and others are legitimate, please use m4_pattern_allow.
10101 	      See the Autoconf documentation.
10102 
10103 	we re-add an AM_PROG_MKDIR_P definition that simply raises a *clear* error
10104 	message when the macro is used.
10105 
10106 	See also recent similar commits 'v1.13-5-g9c07d89' and 'v1.13-4-g8e921bf'
10107 
10108 	* t/am-prog-mkdir-p-no-more.sh: New test.
10109 	* t/list-of-tests.mk: Add it.
10110 	* m4/obsolete-err.m4: Add the new "error-raising" definition for
10111 	AM_PROG_MKDIR_P.
10112 
10113 2012-12-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
10114 
10115 	NEWS: document removal of AM_PROG_MKDIR_P macro
10116 
10117 2012-12-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
10118 
10119 	Merge branch 'maint'
10120 
10121 	* maint:
10122 	  NEWS: minor fixup
10123 	  obsolete: better error message if AM_PROG_CC_STDC is used
10124 	  obsolete: better error message if AM_CONFIG_HEADER is used
10125 	  sync: update files from upstream with "make fetch"
10126 	  tests: better names for temporary configure.ac files
10127 
10128 2012-12-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
10129 
10130 	NEWS: minor fixup
10131 
10132 2012-12-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
10133 
10134 	obsolete: better error message if AM_PROG_CC_STDC is used
10135 
10136 	In Automake 1.13, the long-deprecated (since 2002) macro AM_PROG_CC_STDC
10137 	has been removed.  Such a removal, albeit sensible and justified, was
10138 	probably done in a too-abrupt way, since it didn't turn the pre-existing
10139 	warning messages into fatal error messages, but simply dropped the macro
10140 	definition, so that remaining usages of it would cause unclear error
10141 	messages, e.g.:
10142 
10143 	    configure.ac:4: warning: macro 'AM_PROG_CC_STDC' not found in library
10144 
10145 	from aclocal, and:
10146 
10147 	    configure.ac:4: error: possibly undefined macro: AM_PROG_CC_STDC
10148 	      If this token and others are legitimate, please use m4_pattern_allow.
10149 	      See the Autoconf documentation.
10150 
10151 	from autoconf.
10152 
10153 	In an attempt to mitigate this issue, we re-add an AM_PROG_CC_STDC
10154 	definition that simply raises a *clear* error message when the macro
10155 	is used.
10156 
10157 	* t/am-prog-cc-stdc-no-more.sh: New test.
10158 	* t/list-of-tests.mk: Add it.
10159 	* m4/obsolete-err.m4: Add the new "error-raising" definition for
10160 	AM_PROG_CC_STDC; the error message is a variation of the one already
10161 	present in the older version of this macro, before it got removed in
10162 	commit 'v1.12-15-gd2ca168'.
10163 	* NEWS: Update.
10164 
10165 2012-12-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
10166 
10167 	obsolete: better error message if AM_CONFIG_HEADER is used
10168 
10169 	In Automake 1.13, the long-deprecated macro AM_CONFIG_HEADER (deprecated
10170 	since 2002) has been removed in favour of AC_CONFIG_HEADERS.  However,
10171 	the removal was done without a proper deprecation period, and that
10172 	caused packages upgrading to Automake 1.13 to fail with very unclear
10173 	error messages, e.g.:
10174 
10175 	    configure.ac:4: warning: macro 'AM_CONFIG_HEADER' not found in library
10176 
10177 	from aclocal, and:
10178 
10179 	    configure.ac:4: error: possibly undefined macro: AM_CONFIG_HEADER
10180 	      If this token and others are legitimate, please use m4_pattern_allow.
10181 	      See the Autoconf documentation.
10182 
10183 	from autoconf.
10184 
10185 	In an attempt to mitigate this issue, we re-add an AM_CONFIG_HEADER
10186 	definition that simply raises a *clear* error message when the macro
10187 	is used.
10188 
10189 	Report by Paolo Bonzini:
10190 	<http://lists.gnu.org/archive/html/automake/2012-12/msg00039.html>
10191 
10192 	* t/am-config-header-no-more.sh: New test.
10193 	* t/list-of-tests.mk: Add it.
10194 	* m4/obsolete-err.m4: New file, contain the new AM_CONFIG_HEADER
10195 	"error-raising" definition, as well as the definition of the
10196 	similarly obsolete macros 'AM_C_PROTOTYPES' and 'fp_C_PROTOTYPES',
10197 	moved in from ...
10198 	* m4/protos.m4: ... this file, which has thus been removed.
10199 	* Makefile.am (dist_automake_ac_DATA): Adjust.
10200 	* t/ansi2knr-no-more.sh: Likewise.
10201 	* NEWS: Update.
10202 
10203 2012-12-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
10204 
10205 	sync: update files from upstream with "make fetch"
10206 
10207 	* lib/config.sub: Update.
10208 	* lib/config.guess: Likewise.
10209 
10210 2012-12-29  Stefano Lattarini  <stefano.lattarini@gmail.com>
10211 
10212 	tests: better names for temporary configure.ac files
10213 
10214 	When editing a pre-existent 'configure.ac' file in a test script, name
10215 	the temporary file 'configure.tmp' rather than 'configure.int'; the
10216 	latter was a relict of the times when we used 'configure.in' as the
10217 	default name for the Autoconf input files throughout the testsuite.
10218 
10219 	* t/cond43.sh: Adjust.
10220 	* t/depcomp8a.sh: Likewise.
10221 	* t/depcomp8b.sh: Likewise.
10222 	* t/gettext.sh: Likewise.
10223 	* t/mmode.sh: Likewise.
10224 	* t/pr401.sh: Likewise.
10225 	* t/pr401b.sh: Likewise.
10226 	* t/pr401c.sh: Likewise.
10227 	* t/python11.sh: Likewise.
10228 	* t/cond42.sh: Likewise.  Also, while at it, rename ...
10229 	(edit_configure_in): ... this function ...
10230 	(edit_configure_ac): ... like this.
10231 
10232 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10233 
10234 	NEWS: better distinction between obsolete and obsolescent features
10235 
10236 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10237 
10238 	cleanup: get rid of 'Automake::Configure_ac' module
10239 
10240 	Now that we unconditionally assume the Autoconf input is named
10241 	"configure.ac", it's easier and clearer to inline the checks on
10242 	the existence of such a file directly in the automake and
10243 	aclocal scripts.
10244 
10245 	This change is basically a backport of the 'v1.12-327-gee6d72a'
10246 	commit from Automake-NG.
10247 
10248 	* lib/Automake/Variable.pm (Automake::Configure_ac): Drop this import.
10249 	($configure_ac): Define to "configure.ac", unconditionally.
10250 	* automake.in: Likewise.  Also, in main code, check the existence
10251 	of 'configure.ac' directly, instead of calling the now-removed
10252 	'require_configure_ac' function.
10253 	* aclocal.in: Likewise.
10254 
10255 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10256 
10257 	maint: remove last vestiges of 'configure.in' support from the testsuite
10258 
10259 	* syntax-checks.mk: Don't whitelist any file in the maintainer
10260 	check looking against use of 'configure.in' in the testsuite.
10261 	* t/autodist-configure-no-subdir.sh: Remove last references to
10262 	'configure.in'.
10263 
10264 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10265 
10266 	tests: remove a botched comment from a test case
10267 
10268 	* t/backcompat6.sh: Here.
10269 
10270 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10271 
10272 	maint: reflect removal of 'configure.in' support in comments and messages
10273 
10274 	* t/help.sh, syntax-checks.mk, aclocal.in, old/TODO: In these files.
10275 
10276 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10277 
10278 	Drop support for 'configure.in' as the Autoconf input file
10279 
10280 	The autoconf input should be named 'configure.ac' instead.  The use
10281 	of 'configure.in' has been deprecated in Autoconf since at least
10282 	the 2.13 -> 2.50 transition, and future Autoconf versions (starting
10283 	with 2.70 probably) will start to warn about it at runtime.  Automake
10284 	has been warning about it since the 1.13 release.
10285 
10286 	This change is basically a backport of commit 'v1.12-325-g927d4cb',
10287 	from Automake-NG.
10288 
10289 	* lib/Automake/Configure_ac.pm (find_configure_ac): Don't look for
10290 	'configure.in' if 'configure.ac' is not found.
10291 	(require_configure_ac): On failure, don't tell that "'configure.ac'
10292 	or 'configure.in' is required", but just that "'configure.ac' is
10293 	required".
10294 	* automake.in (@common_sometimes): Don't list 'configure.in'.
10295 	(scan_autoconf_files): Adjust heading comments.
10296 	* NEWS: Update.
10297 	* t/help.sh: Adjust.
10298 	* t/configure.sh: Remove as obsolete.
10299 	* t/list-of-tests.mk: Adjust.
10300 
10301 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10302 
10303 	Drop support for $(ACLOCAL_M4_SOURCES), it's obsolete
10304 
10305 	The $(ACLOCAL_M4_SOURCES) variables has been obsolete by a long time,
10306 	and basically unused (since at least commit ab57cf28 of 2003-04-24), so
10307 	drop any mention of it.
10308 
10309 	* automake.in (scan_aclocal_m4): Flatly reject any ACLOCAL_M4_SOURCES
10310 	definition with a fatal error, rather than just warning about it.
10311 	if any deinifn
10312 	$(ACLOCAL_M4_SOURCES).  Don't append it to @ac_deps.  This makes such
10313 	variable always empty, so there's no need to return it.  Adjust the
10314 	signature of this function accordingly, and ...
10315 	(handle_configure): ... update this function accordingly.
10316 	* NEWS: Update.
10317 
10318 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10319 
10320 	tests: remove stray uses of 'Exit', in favour of simple 'exit'
10321 
10322 	* t/warnopts.sh: Here.  Suggested by the 'sc_tests_exit_not_Exit'
10323 	maintainer check.
10324 
10325 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10326 
10327 	Merge branch 'remove-djgpp-support'
10328 
10329 	* remove-djgpp-support:
10330 	  general: assume dot-starting file names are supported
10331 	  Drop support for DJGPP, MS-DOS, Windows 95/98/ME
10332 
10333 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10334 
10335 	general: assume dot-starting file names are supported
10336 
10337 	Now that we've dropped support for DJGPP, Windows 95/98/ME
10338 	and MS-DOS, that should always the case.
10339 
10340 	This change is basically a backport of Automake-NG commit
10341 	'v1.12-342-ge6af355'.
10342 
10343 	* lib/am/texibuild.am: Just assume we can have directories whose
10344 	name starts with a dot, i.e., that '${am__leading_dot}' expands
10345 	to '.'.
10346 	* automake.in (require_build_directory): Likewise.
10347 	* m4/depend.m4 (AM_SET_DEPDIR): Likewise.  Now $(DEPDIR) will
10348 	be set unconditionally to '.deps'.
10349 	* m4/init.m4 (AM_INIT_AUTOMAKE): Don't AC_REQUIRE the m4 macro
10350 	'AM_SET_LEADING_DOT' anymore.
10351 	* m4/lead-dot.m4: Delete.
10352 	* Makefile.am (dist_automake_ac_DATA): Remove it.
10353 	* t/subobj11a.sh: Just assume $(DEPDIR) expands to '.deps'.
10354 
10355 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10356 
10357 	Drop support for DJGPP, MS-DOS, Windows 95/98/ME
10358 
10359 	Basically a backport and squash-in of three Automake-NG commits:
10360 	v1.12.1-447-g041fe15, v1.12-343-g4526363, v1.12-341-gc853c20,
10361 	plus proper simplifications of the ylwrap script.
10362 
10363 	See also:
10364 	<http://lists.gnu.org/archive/html/automake-ng/2012-05/msg00104.html>
10365 
10366 	* automake.in (BEGIN): Don't override $ENV{SHELL} for DJGPP.
10367 	(handle_libtool): Do not bother cleaning '_libs' directories, they
10368 	are only used by Libtool on MS-DOS (and we no longer support that
10369 	setup).
10370 	* syntax-checks.mk (automake_diff_no): Adjust, as now there are only
10371 	seven (not eight) different lines between 'automake.in' and 'automake'.
10372 	* bootstrap.sh: Don't special case the default value for BOOTSTRAP_SHELL
10373 	on DJGPP.  While we are at, allow that variable to be overridden from
10374 	the environment. Also, adjust comments: we don't use "ln -s" not because
10375 	it's not properly supported by DJGPP, but because it might not be
10376 	properly supported by MinGW/MSYS.
10377 	* configure.ac (MODIFICATION_DELAY): Define to '2' unconditionally, as
10378 	we no longer have to cater to quirks of Windows 95/98/ME.
10379 	* lib/am/texibuild.am: Remove support for '*.iNN' files used on DJGPP.
10380 	* lib/am/texinfos.am: Likewise.
10381 	* lib/ylwrap: Remove support for DOS-specific filenames for bison
10382 	output.
10383 	* t/txinfo19.sh: Remove as obsolete.
10384 	* t/list-of-tests.mk: Adjust.
10385 	* NEWS: Update.
10386 
10387 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10388 
10389 	depcomp: remove support for SGI C/C++ compilers
10390 
10391 	These compilers are only meant to run on IRIX, and that system has seen
10392 	its last release in 2006, and is expected to lose support from SGI in
10393 	December 2013:
10394 	<http://www.sgi.com/services/support/irix_mips_support.html>
10395 
10396 	This change fixes automake bug#12967 and bug#12978.
10397 
10398 	* lib/depcomp (sgi): Remove this depmode.
10399 	* NEWS: Update.
10400 
10401 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10402 
10403 	tests: a test about warnings in category 'obsolete' was failing spuriously
10404 
10405 	* t/warnings-obsolete-default.sh : After our removal of $(INCLUDES)
10406 	support, this test has been spuriously failing because it was assuming
10407 	that a definition of the INCLUDES make variable only caused a warning
10408 	rather than a fatal error.  Fix it to use a definition of the newly
10409 	obsolescent ACLOCAL_AMFLAGS make variable.
10410 
10411 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10412 
10413 	ACLOCAL_AMFLAGS: fully deprecate, with warnings in the 'obsolete' category
10414 
10415 	Users are expected to use the AC_CONFIG_MACRO_DIR{,S} autoconf macros
10416 	instead.  And the use of the '--install' aclocal option in ACLOCAL_AMFLAGS
10417 	has proved to be a bad idea anyway -- see automake bug#9037.
10418 
10419 	* NEWS: Update.
10420 	* t/aclocal-amflags.sh: Enhance and adjust.
10421 	* t/acloca14b.sh: Adjust.
10422 	* t/acloca22b.sh: Likewise.
10423 	* t/ax/distcheck-hook-m4.am: Likewise.
10424 	* t/dist-missing-m4.sh: Likewise.
10425 	* t/remake-renamed-m4-macro.sh: Likewise.
10426 	* t/remake-renamed-m4-macro-and-file.sh: Likewise.
10427 	* t/remake-renamed-m4-file.sh: Likewise.
10428 	* t/remake-moved-m4-file.sh: Likewise.
10429 	* t/remake-gnulib-remove-header.sh: Likewise.
10430 	* t/remake-gnulib-add-acsubst.sh: Likewise.
10431 	* t/remake-deleted-m4-file.sh: Likewise.
10432 	* t/remake-deleted-m4-file.sh: Likewise.
10433 	* t/libtool-macros.sh: Likewise.
10434 	* t/gettext-macros.sh: Likewise.
10435 
10436 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10437 
10438 	compile: remove support for $(INCLUDES)
10439 
10440 	It has already been deprecated in the manual and by warnings in the
10441 	'obsolete' category for ages (at least since 2003), in favour of
10442 	AM_CPPFLAGS.  Automake-NG has removed support for it already.  And
10443 	the warnings in the obsolete category are on by default since the
10444 	1.12.4 release.
10445 
10446 	So, by removing it in Automake 1.14, we will simplify the transition
10447 	path for people that want to switch to Automake-NG.
10448 
10449 	See also Automake-NG commit v1.12-267-g6b74968 of 2012-05-21,
10450 	"[ng] compile: don't support $(INCLUDES) anymore, it's obsolete".
10451 
10452 	* NEWS, doc/automake.texi: Update.
10453 	* automake.in (generate_makefile): Reject INCLUDES unconditionally.
10454 	In all languages: don't add $(INCLUDES) to the compiler command line.
10455 	* t/includes-deprecation.sh: New test.
10456 	* t/list-of-tests.mk: Add it.
10457 	* t/werror2.sh : Adjust to avoid spurious failures.
10458 	* t/pluseq5.sh: Likewise.
10459 	* t/warnopts.sh: Likewise.
10460 
10461 	Suggested-by: Paolo Bonzini <bonzini@gnu.org>
10462 
10463 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10464 
10465 	Merge branch 'aclocal-lookup-order'
10466 
10467 	* aclocal-lookup-order:
10468 	  aclocal: allow third-party macros take precedence over automake ones
10469 
10470 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10471 
10472 	aclocal: allow third-party macros take precedence over automake ones
10473 
10474 	It makes sense to allow "local system-wide" m4 macros take precedence
10475 	over built-in macros (defined in the aclocal versioned m4 directory
10476 	"${prefix}/share/aclocal-${APIVERSION}"), the same way we allow
10477 	"environment-level" m4 macros (through the ACLOCAL_PATH environment
10478 	variable ) take precedence over system-wide third-party macros.
10479 
10480 	An example: after this change, a definition of AM_PROG_VALAC placed
10481 	in file (say) '/usr/local/share/aclocal/my-vala.m4' should take
10482 	precedence over the same-named automake-provided macro defined in
10483 	file '/usr/local/share/aclocal-1.14/vala.m4'.
10484 
10485 	Even more importantly, after this change, m4 macros accessible
10486 	through $ACLOCAL_PATH entries will take precedence over m4 macros
10487 	defined in the aclocal versioned m4 directory; it was quite counter
10488 	intuitive that it was not the case before.
10489 
10490 	* aclocal.in (scan_m4_files): Search files in @system_includes before
10491 	files in @automake_includes.
10492 	* t/aclocal-acdir.sh: Adjust.
10493 	* t/aclocal-path-precedence.sh: Likewise.
10494 	* t/dirlist.sh: Likewise.
10495 	* doc/automake.texi: Adjust.
10496 	* NEWS: Update.
10497 
10498 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10499 
10500 	Merge branch 'am-prog-mkdir-p-remove'
10501 
10502 	Finally backporting the AM_PROG_MKDIR_P removal that has already taken
10503 	place in Automake-NG.  We can safely do this since the latest Gettext
10504 	release (1.18.2) has removed all uses of AM_PROG_MKDIR_P, replacing
10505 	them with uses of AC_PROG_MKDIR_P instead.
10506 
10507 	References:
10508 	<http://lists.gnu.org/archive/html/bug-gettext/2012-12/msg00064.html>
10509 	<http://lists.gnu.org/archive/html/automake/2012-12/msg00024.html>
10510 	Commit v0.18.1.1-147-g2cccb4c in the Gettext git repository.
10511 
10512 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10513 
10514 	Merge branch 'maint'
10515 
10516 	Actually a dummy merge, since we retain our own "1.13a" version
10517 	number.
10518 
10519 	* maint:
10520 	  post-release: minor version bump
10521 
10522 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10523 
10524 	post-release: major version bump
10525 
10526 	This is for the master branch.
10527 
10528 	* configure.ac, m4/amversion.m4: Bump version to 1.13a.
10529 
10530 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10531 
10532 	post-release: minor version bump
10533 
10534 	This is for the maint branch.
10535 
10536 	* configure.ac, m4/amversion.m4: Bump version to 1.13.0a.
10537 
10538 2012-12-28  Stefano Lattarini  <stefano.lattarini@gmail.com>
10539 
10540 	release: stable major release 1.13
10541 
10542 	* configure.ac (AC_INIT): Bump version number to 1.13.
10543 	* m4/amversion.m4 (AM_AUTOMAKE_VERSION): Likewise (regenerated
10544 	by "make bootstrap").
10545 
10546 2012-12-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
10547 
10548 	news: fix botched URL
10549 
10550 2012-12-21  Akim Demaille  <akim@lrde.epita.fr>
10551 
10552 	ylwrap: various fixes
10553 
10554 	Rename properly header guards in generated header files, instead of
10555 	leaving Y_TAB_H.
10556 
10557 	Convert header guards in implementation files.  Because ylwrap failed
10558 	to rename properly #include in the implementation files, current
10559 	versions of Bison (e.g., 2.7) duplicate the generated header file in
10560 	the implementation file.  The header guard then protects the
10561 	implementation file from duplicate definitions from the header file.
10562 
10563 	Generate header guards with a single '_' for series of non alphabetic
10564 	characters, instead of several.  This is what Bison does.
10565 
10566 	Makes the test t/yacc-d-basic.sh pass again.
10567 
10568 	* lib/ylwrap (guard): Properly honor $1 to rename properly the
10569 	header guards.
10570 	Keep a single _ instead of several.
10571 	(rename_sed): Rename as...
10572 	(sed_fix_filenames): this.
10573 	Suggested by Stefano Lattarini.
10574 	(sed_fix_header_guards): New.
10575 	Use it.
10576 
10577 2012-12-21  Akim Demaille  <akim@lrde.epita.fr>
10578 
10579 	tests: strengthen the ylwrap tests
10580 
10581 	* t/yacc-d-basic.sh: Comment changes.
10582 	(generated): New.
10583 	Use it to factor various tests.
10584 	Check that Y_TAB_H is not issued.
10585 
10586 2012-12-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
10587 
10588 	NEWS: we plan to ditch MS-DOS/DJGPP support in future Automake versions
10589 
10590 2012-12-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
10591 
10592 	maint: typofixes in makefile comments
10593 
10594 	* maint.mk (git-upload-release): Here.
10595 
10596 2012-12-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
10597 
10598 	maint: remove extra verbosity from a maintainer rule
10599 
10600 	* maint.mk (git-upload-release): Here, by adding a forgotten
10601 	'@' recipe modifier.
10602 
10603 2012-12-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
10604 
10605 	NEWS: fix grammaro
10606 
10607 2012-12-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
10608 
10609 	maint: version bump after beta release
10610 
10611 	* configure.ac (AC_INIT): Bump version number to 1.12c, as per
10612 	HACKING suggestion.
10613 	* m4/amversion.m4: Likewise (autoupdated by "make bootstrap").
10614 
10615 2012-12-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
10616 
10617 	release: beta release 1.12b (will become 1.13)
10618 
10619 	* configure.ac (AC_INIT): Bump version number to 1.12b.
10620 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
10621 
10622 2012-12-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
10623 
10624 	cosmetics: comments fixlets in our Makefile
10625 
10626 	* Makefile.am: Prefer '#' comments over '##' ones where it makes sense.
10627 	Other minor style tweaks to comments about "updated later" variables.
10628 
10629 2012-12-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
10630 
10631 	cosmetics: consistency changes in some make recipes
10632 
10633 	* Makefile.am (install-exec-hook, uninstall-hook): Don't use extra
10634 	quoting in variable assignments.  More consistent use of whitespace.
10635 
10636 2012-12-18  Stefano Lattarini  <stefano.lattarini@gmail.com>
10637 
10638 	tests: avoid a spurious failure on NetBSD
10639 
10640 	* t/tags-pr12372.sh (configure.ac): AC_SUBST the LINK variable to a
10641 	dummy invocation, to avoid possible errors from make or the linker;
10642 	errors we do not care about in the least in this test.
10643 	(Makefile.am, sub/Makefile.am): Remove LINK definitions; simply
10644 	inherit that in configure.ac.
10645 
10646 2012-12-18  Stefano Lattarini  <stefano.lattarini@gmail.com>
10647 
10648 	tests: re-enable some checks disabled by mistake ...
10649 
10650 	* t/subpkg-macrodir.sh: ... in this test.  And fix some typos that would
10651 	have caused the newly re-enabled check to spuriously fail.  Issue revealed
10652 	by the lack of '.PHONY' support in Solaris CCS make (yes, so we've been
10653 	actually *helped* by that make implementation once, at last.  Scary ;-)
10654 
10655 2012-12-18  Stefano Lattarini  <stefano.lattarini@gmail.com>
10656 
10657 	tests: avoid a "grep -E" portability issue
10658 
10659 	As stated in the Autoconf manual, the empty alternative is not portable
10660 	with "grep -E".  For example, while with GNU grep we have:
10661 
10662 	    $ echo b | grep -E '(|a)b'; echo status = $?
10663 	    b
10664 	    status = 0
10665 
10666 	with Solaris 10 /usr/xpg4/bin/grep we have:
10667 
10668 	    $ echo b | grep -E '(|a)b'; echo status = $?
10669 	    1
10670 	    status = 1
10671 
10672 	* t/silent-texi.sh: That was causing a spurious failure here.  Fix it
10673 	by simply using '?' instead (as suggested by the Autoconf manual).
10674 
10675 2012-12-18  Stefano Lattarini  <stefano.lattarini@gmail.com>
10676 
10677 	tests: avoid a spurious failure due to a Clang bug
10678 
10679 	This version of clang:
10680 
10681 	    clang version 3.2 (trunk 163574)
10682 	    Target: powerpc64-unknown-linux-gnu
10683 	    Thread model: posix
10684 
10685 	caused the test 't/ltcond2.sh' to spuriously fail due to what
10686 	appeared like a clang bug.  Here is a part of the diagnostic (trimmed
10687 	down for better clarity):
10688 
10689 	    clang: .../cfarm/llvm/lib/MC/MCAsmStreamer.cpp:338: \
10690 	      virtual void {anonymous}::MCAsmStreamer::EmitLabel(llvm::MCSymbol*): \
10691 	      Assertion `Symbol->isUndefined() && "Cannot define a symbol twice!"' \
10692 	      failed.
10693 	    ...
10694 	    7  clang  0x0000000012a459c4 llvm::AsmPrinter::EmitGlobalVariable\
10695 	      (llvm::GlobalVariable const*) + 18446744073680468044
10696 	    8  clang  0x0000000012a490a8 llvm::AsmPrinter::doFinalization\
10697 	      (llvm::Module&) + 18446744073680481840
10698 	    ...
10699 	    Stack dump:
10700 	    0.  Program arguments: .../opt/cfarm/clang-2012.09.10/bin/clang \
10701 	          -cc1 -triple powerpc64-unknown-linux-gnu -S -disable-free \
10702 	          ...
10703 	    1.  <eof> parser at end of file
10704 	    2.  Code generation
10705 	    3.  Running pass 'Function Pass Manager' on module 'hello-generic.c'.
10706 	    clang: error: unable to execute command: Aborted
10707 	    clang: error: clang frontend command failed due to signal (use -v to \
10708 	           see invocation)
10709 	    clang version 3.2 (trunk 163574)
10710 	    Target: powerpc64-unknown-linux-gnu
10711 	    Thread model: posix
10712 
10713 	So tweak the affected test case to avoid triggering this bug.  This is
10714 	the easiest way for us to keep the testsuite result clean and meaningful
10715 	on our main Clang test bed.
10716 
10717 	* t/ltcond2.sh: Prefer using "extern const char *" variables rather
10718 	than functions returning a statically allocated "const char *" variable.
10719 
10720 2012-12-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
10721 
10722 	NEWS: start preparing for the 1.13 release
10723 
10724 	* NEWS: Here.  In particular, list (some of) the planned future
10725 	backward-incompatibilities that will take place in Automake 1.14.
10726 
10727 2012-12-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
10728 
10729 	Merge branch 'maint'
10730 
10731 	* maint:
10732 	  tests: typofix in the shebang line of 't/py-compile-destdir.sh'
10733 	  HACKING: release process: fixlets about the announcement
10734 	  maint: post-release minor version bump
10735 	  release: stable release 1.12.5
10736 	  tests: avoid errors due to underquoted third-party macros
10737 	  maintcheck: also consider contrib tests for syntax checks
10738 	  cosmetics: whitespace fixlets in syntax-checks.mk
10739 	  maint: prefer $(MKDIR_P) over $(mkinstalldirs) in maintainer rules
10740 
10741 2012-12-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
10742 
10743 	tests: typofix in the shebang line of 't/py-compile-destdir.sh'
10744 
10745 2012-12-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
10746 
10747 	HACKING: release process: fixlets about the announcement
10748 
10749 	* HACKING: The announcement should be generate *before* bumping
10750 	the version number to the next alpha number.  Also, only an
10751 	abridged version of the announcement should be copied on the
10752 	Savannah news feed (a link to the complete announcement should
10753 	be linked from there, though).
10754 
10755 2012-12-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
10756 
10757 	maint: post-release minor version bump
10758 
10759 	* configure.ac (AC_INIT): Bump version number to 1.12.6a.
10760 	* m4/amversion.m4: Likewise (automatically regenerated by
10761 	"make bootstrap").
10762 
10763 2012-12-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
10764 
10765 	release: stable release 1.12.5
10766 
10767 	* configure.ac (AC_INIT): Bump version number to 1.12.5.
10768 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
10769 
10770 2012-12-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
10771 
10772 	tests: avoid errors due to underquoted third-party macros
10773 
10774 	* t/pkg-config-macros.sh: Call aclocal with the warnings in the
10775 	"syntax" category disabled.
10776 
10777 2012-12-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
10778 
10779 	maintcheck: also consider contrib tests for syntax checks
10780 
10781 	* syntax-checks.mk (xtests): By listing them here.
10782 
10783 2012-12-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
10784 
10785 	cosmetics: whitespace fixlets in syntax-checks.mk
10786 
10787 2012-12-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
10788 
10789 	maint: prefer $(MKDIR_P) over $(mkinstalldirs) in maintainer rules
10790 
10791 	* maint.mk (check-coverage-run, recheck-coverage-run): Here.  This
10792 	is especially useful since $(mkinstalldirs) is no longer available
10793 	in Automake-NG.
10794 
10795 2012-12-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
10796 
10797 	Merge branch 'maint'
10798 
10799 	* maint:
10800 	  sync: update files from upstream with "make fetch"
10801 	  gitignore: align with recent changes
10802 	  fixup: add dummy rule for ChangeLog generation in Makefile.am
10803 	  maint: no longer use AM_MAKEFLAGS in maintainer rules
10804 	  maint: use more GNU make features in maintainer rules
10805 	  maint: move maintainer make rules in maint.mk
10806 
10807 2012-12-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
10808 
10809 	sync: update files from upstream with "make fetch"
10810 
10811 	* lib/gnupload: Update.  See:
10812 	<http://lists.gnu.org/archive/html/bug-gnulib/2012-12/msg00050.html>
10813 
10814 2012-12-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
10815 
10816 	gitignore: align with recent changes
10817 
10818 2012-12-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
10819 
10820 	fixup: add dummy rule for ChangeLog generation in Makefile.am
10821 
10822 	So that automake won't complain about the missing ChangeLog.
10823 	The real rule for ChangeLog generation is now in maint.mk (as
10824 	it is maintainer-specific).
10825 
10826 	* Makefile.am (ChangeLog): Dummy.
10827 
10828 2012-12-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
10829 
10830 	maint: no longer use AM_MAKEFLAGS in maintainer rules
10831 
10832 	They are not required with GNU make.
10833 
10834 	* maint.mk: Adjust accordingly.
10835 
10836 2012-12-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
10837 
10838 	maint: use more GNU make features in maintainer rules
10839 
10840 	* maint.mk: Here, throughout the file.  Other minor cleanups
10841 	and refactorings while at it.
10842 
10843 2012-12-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
10844 
10845 	maint: move maintainer make rules in maint.mk
10846 
10847 	This is done for two reasons:
10848 
10849 	  1. It will enable us to take advantage of more GNU make
10850 	     features in our maintainer-specific recipes, if we
10851 	     want to (and we probably will).
10852 
10853 	  2. The Makefile.am won't have to be modified each time a
10854 	     maintainer rule needs to be changed or tweaked; this
10855 	     will prevent useless remaking of (in particular) the
10856 	     't/testsuite-part.am' file.
10857 
10858 	* maint.mk: New, several recipes, rules and variables for
10859 	maintainer-specific tasks, extracted (with some tweaks)
10860 	from ...
10861 	* Makefile.am: ... this file.
10862 	(EXTRA_DIST): Add 'maint.mk'.
10863 	* GNUmakefile: Include 'maint.mk' after 'Makefile'.
10864 
10865 2012-12-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
10866 
10867 	fixup: source 'test-init.sh', not './defs', in a test
10868 
10869 	* t/pkg-config-macros.sh: Here.
10870 
10871 2012-12-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
10872 
10873 	Merge branch 'maint'
10874 
10875 	* maint:
10876 	  docs: fixlet about xz compression levels and command line options
10877 	  maint: more files excluded in 'update-copyright'
10878 	  maint: adapt 'update-copyright' recipe to the new $(FETCHFILES) format
10879 	  HACKING: update instructions to update the manuals at www.gnu.org
10880 	  maint: add an explicative comment in Makefile
10881 	  maint: factor out the name of the dir where web manuals are generated
10882 	  maint: rename rules to build/upload web manuals
10883 	  cosmetic: minor variable and rules reordering in Makefile.am
10884 	  docs: build the web manuals for CVS in the builddir, not in the srcdir
10885 	  release: automate uploading of web manuals in CVS
10886 	  docs: AM_CFLAGS: remove reference to non-existing "more detailed" desc
10887 	  tests: avoid two spurious failures on FreeBSD
10888 	  tests: improve 'pkg-config' requirement
10889 
10890 2012-12-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
10891 
10892 	Merge branch 'improve-release-process' into maint
10893 
10894 	* improve-release-process:
10895 	  HACKING: update instructions to update the manuals at www.gnu.org
10896 	  maint: add an explicative comment in Makefile
10897 	  maint: factor out the name of the dir where web manuals are generated
10898 	  maint: rename rules to build/upload web manuals
10899 	  cosmetic: minor variable and rules reordering in Makefile.am
10900 	  docs: build the web manuals for CVS in the builddir, not in the srcdir
10901 	  release: automate uploading of web manuals in CVS
10902 	  docs: AM_CFLAGS: remove reference to non-existing "more detailed" desc
10903 	  tests: avoid two spurious failures on FreeBSD
10904 
10905 2012-12-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
10906 
10907 	Merge branch 'pkgconfing-tests-requirement' into maint
10908 
10909 	* pkgconfing-tests-requirement:
10910 	  tests: improve 'pkg-config' requirement
10911 
10912 2012-12-14  Patrick Welche  <prlw1@cam.ac.uk>  (tiny change)
10913 
10914 	docs: fixlet about xz compression levels and command line options
10915 
10916 	* doc/automake.texi (The types of distributions): To display a
10917 	progress bar, xz takes "-v" for verbose.
10918 
10919 2012-12-13  Stefano Lattarini  <stefano.lattarini@gmail.com>
10920 
10921 	maint: more files excluded in 'update-copyright'
10922 
10923 	This will avoid the following spurious warnings when
10924 	"make update-copyright" is run:
10925 
10926 	    .autom4te.cfg: warning: copyright statement not found
10927 	    .git-log-fix: warning: copyright statement not found
10928 	    .gitattributes: warning: copyright statement not found
10929 	    .gitignore: warning: copyright statement not found
10930 	    AUTHORS: warning: copyright statement not found
10931 	    THANKS: warning: copyright statement not found
10932 	    contrib/README: warning: copyright statement not found
10933 	    contrib/multilib/README: warning: copyright statement not found
10934 	    doc/amhello/README: warning: copyright statement not found
10935 	    lib/install-sh: warning: copyright statement not found
10936 	    lib/mkinstalldirs: warning: copyright statement not found
10937 	    m4/acdir/README: warning: copyright statement not found
10938 	    t/README: warning: copyright statement not found
10939 
10940 	* Makefile.am (files_without_copyright): New.
10941 	(update-copyright): Use it when computing '$exclude_list'.
10942 	Also, skip any README file.
10943 
10944 2012-12-13  Stefano Lattarini  <stefano.lattarini@gmail.com>
10945 
10946 	maint: adapt 'update-copyright' recipe to the new $(FETCHFILES) format
10947 
10948 	This is a fixup for commit v1.12.5-27-g71ce1b4 of 2012-12-09, "fetch:
10949 	improve, and reduce code duplication".
10950 
10951 	* Makefile.am (update-copyright): Adjust.
10952 
10953 2012-12-13  Stefano Lattarini  <stefano.lattarini@gmail.com>
10954 
10955 	HACKING: update instructions to update the manuals at www.gnu.org
10956 
10957 2012-12-13  Stefano Lattarini  <stefano.lattarini@gmail.com>
10958 
10959 	maint: add an explicative comment in Makefile
10960 
10961 	* Makefile.am (web-manual-update): Here, about a slightly tricky
10962 	feature of rsync(1) used in the recipe.
10963 
10964 2012-12-13  Stefano Lattarini  <stefano.lattarini@gmail.com>
10965 
10966 	maint: factor out the name of the dir where web manuals are generated
10967 
10968 	Just a minor refactoring to reduce duplication; no semantic change is
10969 	intended.
10970 
10971 	* Makefile.am (web_manual_dir): New.
10972 	(web-manual, web-manual-update, clean-web-manual): Use it.
10973 
10974 2012-12-13  Stefano Lattarini  <stefano.lattarini@gmail.com>
10975 
10976 	maint: rename rules to build/upload web manuals
10977 
10978 	The new names are the same used by the similar-purpose rules provided by
10979 	the 'top/maint.mk' makefile in gnulib (as of commit v0.0-7741-g4a8c422).
10980 	No need to have gratuitous incompatibilities.
10981 
10982 	* Makefile.am (web-manuals): Rename ...
10983 	(web-manual): ... like this.
10984 	(upload-web-manuals): Rename ...
10985 	(web-manual-update): ... like this.
10986 	(clean-web-manuals): Rename ...
10987 	(clean-web-manual): ... like this.
10988 
10989 2012-12-13  Stefano Lattarini  <stefano.lattarini@gmail.com>
10990 
10991 	cosmetic: minor variable and rules reordering in Makefile.am
10992 
10993 2012-12-13  Stefano Lattarini  <stefano.lattarini@gmail.com>
10994 
10995 	docs: build the web manuals for CVS in the builddir, not in the srcdir
10996 
10997 	We can do so using the '-I' option of the gendocs.sh script.
10998 
10999 	Inspired by the 'web-manual' rule in the 'top/maint.mk' file provided
11000 	by gnulib (as of commit v0.0-7741-g4a8c422) as customized by Bison in
11001 	its 'cfg.mk' file (as of commit v2.6.5-1007-gf5fceda).
11002 
11003 	* Makefile.am (web-manuals): Modify and simplify to take advantage
11004 	of gendocs.sh's  '-I' option.
11005 	(upload-we-manuals): Accordingly adjust sanity checks and the rsync
11006 	invocation in the recipe.
11007 	(clean_texinfo_clutter_cmd): Delete this variable, it's no longer
11008 	required.
11009 	(clean-texinfo-clutter): Delete this phony target, it's no longer
11010 	required.
11011 	(clean-local): Don't depend on it.
11012 	* configure.ac: Invoke AC_PROG_LN_S, since we want to use $(LN_S)
11013 	in the new 'web-manuals' recipe.
11014 
11015 2012-12-13  Stefano Lattarini  <stefano.lattarini@gmail.com>
11016 
11017 	release: automate uploading of web manuals in CVS
11018 
11019 	This is partly inspired to the 'gnu-web-doc-update' script from
11020 	gnulib.  Unfortunately, we can't use that script directly, because
11021 	as of now it assumes a build system integrated with gnulib.
11022 
11023 	* Makefile.am (upload-web-manuals): New .PHONY rule.
11024 	(RSYNC, CVS, CVSU, CVS_USER, WEBCVS_ROOT): New variables, used by
11025 	the new rule.
11026 
11027 2012-12-13  Stefano Lattarini  <stefano.lattarini@gmail.com>
11028 
11029 	docs: AM_CFLAGS: remove reference to non-existing "more detailed" desc
11030 
11031 	* doc/automake.texi (Program variables): Here.  Reported by NightStrike:
11032 	<http://lists.gnu.org/archive/html/automake/2012-12/msg00012.html>
11033 
11034 2012-12-13  Stefano Lattarini  <stefano.lattarini@gmail.com>
11035 
11036 	tests: avoid two spurious failures on FreeBSD
11037 
11038 	Due to the FreeBSD make bug that causes of the well-known automake
11039 	bug#7884, two of our bison-requiring tests were failing on FreeBSD.
11040 	To avoid this, simply skip the part of these tests affected by that
11041 	bug.
11042 
11043 	* t/yacc-bison-skeleton.sh: Here, using the 'yl_distcheck' function.
11044 	* t/yacc-bison-skeleton-cxx.sh: Likewise.
11045 	* t/yacc-nodist.sh: While at it, use the 'yl_distcheck' here
11046 	as well, instead of running "$MAKE distcheck" only when $MAKE
11047 	is GNU make.
11048 
11049 2012-12-13  Stefano Lattarini  <stefano.lattarini@gmail.com>
11050 
11051 	tests: improve 'pkg-config' requirement
11052 
11053 	This improves (and unifies among different tests) the detection of
11054 	both the 'pkg-config' program and m4 macros provided by it; they are
11055 	needed in particular by tests on Vala support.
11056 
11057 	Before this, some Vala-related tests were failing spuriously when the
11058 	pkg-config program was available but the macros from pkg.m4 were not
11059 	made explicitly available in ACLOCAL_PATH.  These tests were:
11060 
11061 	    vala-mix vala-mix2 vala-parallel vala-vapi vala-vpath
11062 
11063 	See also similar commit 'v1.11-581-gb7d67d5' of 2011-12-22,
11064 	"tests: better handling of gettext and libtool requirements".
11065 
11066 	* t/ax/am-test-lib.sh (process_requirements): If a requirement
11067 	matches the 'pkg-config*' wildcard, try to make available to it
11068 	the pkg-config m4 macros, and skip it if it's not possible.
11069 	This is done by sourcing the 'get.sh' fragment generated by ...
11070 	* t/pkg-config-macros.sh: ... this new test, which looks in
11071 	advance for the 'pkg-config' program and the m4 macros that
11072 	comes with it.
11073 	* t/list-of-tests.mk: Add the new test.
11074 	* gen-testsuite-part (%deps_extractor): Tests requiring pkg-config
11075 	will need the m4 macros coming with it; so these tests should be
11076 	run after the 'pkg-config-macros.sh' test.  Automatically generate
11077 	proper declarations for such dependencies.
11078 	* t/vala-header.sh: No need to explicitly check that the macro
11079 	PKG_CHECK_MODULES has been expanded by autoconf.
11080 	* t/vala-libs.sh: Likewise.
11081 	* t/vala-non-recursive-setup.sh: Likewise.
11082 	* t/vala-per-target-flags.sh: Likewise.
11083 	* t/vala-recursive-setup.sh: Likewise.
11084 
11085 2012-12-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
11086 
11087 	Merge branch 'maint'
11088 
11089 	* maint:
11090 	  docs: AM_CFLAGS: remove reference to non-existing "more detailed" desc
11091 	  maintcheck: prefer '$(am__cd)' over plain 'cd'
11092 	  fetch: improve, and reduce code duplication
11093 	  docs: add rule to generate manuals for www.gnu.org
11094 	  docs: copy the 'gendocs.sh' script from Texinfo CVS repository
11095 	  sync: update files from upstream with "make fetch"
11096 	  release: generate a stub for the release announcement
11097 	  release: distinguish major and minor releases
11098 	  tests: avoid two spurious failures on FreeBSD
11099 
11100 2012-12-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
11101 
11102 	Merge branch 'improve-release-process' into maint
11103 
11104 	* improve-release-process:
11105 	  maintcheck: prefer '$(am__cd)' over plain 'cd'
11106 	  fetch: improve, and reduce code duplication
11107 	  docs: add rule to generate manuals for www.gnu.org
11108 	  docs: copy the 'gendocs.sh' script from Texinfo CVS repository
11109 	  sync: update files from upstream with "make fetch"
11110 	  release: generate a stub for the release announcement
11111 	  release: distinguish major and minor releases
11112 
11113 2012-12-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
11114 
11115 	docs: AM_CFLAGS: remove reference to non-existing "more detailed" desc
11116 
11117 	* doc/automake.texi (Program variables): Here.  Reported by NightStrike:
11118 	<http://lists.gnu.org/archive/html/automake/2012-12/msg00012.html>
11119 
11120 2012-12-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
11121 
11122 	maintcheck: prefer '$(am__cd)' over plain 'cd'
11123 
11124 	In a couple of rules in out own build system, as suggested by the
11125 	maintainer check 'sc_cd_relative_dir'.
11126 
11127 	* Makefile.am (web-manuals, clean_texinfo_clutter_cmd): Adjust.
11128 
11129 2012-12-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
11130 
11131 	fetch: improve, and reduce code duplication
11132 
11133 	* Makefile.am (WGET_SV_CVS, WGET_SV_GIT_CF, WGET_SV_GIT_AC,
11134 	WGET_SV_GIT_GL): Delete.
11135 	(SV_CVS, SV_GIT_CF, SV_GIT_AC, SV_GIT_GL): New.
11136 	(fetch): Simplify and enhance.
11137 	(update-copyright): Adjust.
11138 
11139 2012-12-09  Stefano Lattarini  <stefano.lattarini@gmail.com>
11140 
11141 	docs: add rule to generate manuals for www.gnu.org
11142 
11143 	Less clerical work to do by hand.
11144 
11145 	* Makefile.am (web-manuals): New rule.
11146 	(clean_texinfo_clutter): New macro.
11147 	* clean-web-manuals, clean-texinfo-clutter: New cleaning
11148 	rules.
11149 	(clean-local): Depend on them.
11150 	* .gitignore: Update.
11151 	* HACKING: Adjust instructions for generation of manuals.
11152 
11153 2012-12-08  Stefano Lattarini  <stefano.lattarini@gmail.com>
11154 
11155 	docs: copy the 'gendocs.sh' script from Texinfo CVS repository
11156 
11157 	As well as the 'gendocs_template' file.  And enhance the "make fetch"
11158 	target to keep them synced.  It's better than having to fetch these
11159 	files by hand each time a release is to be made.
11160 
11161 	* Makefile.am (FETCHFILES): List 'gendocs.sh' and 'gendocs_template'.
11162 	(fetch): Fetch them, into 'lib/'.
11163 	(EXRA_DIST): Distribute them.
11164 	* HACKING: Adjust instructions for generation of manuals.
11165 
11166 2012-12-08  Stefano Lattarini  <stefano.lattarini@gmail.com>
11167 
11168 	sync: update files from upstream with "make fetch"
11169 
11170 	* lib/config.sub: Update.
11171 
11172 2012-12-08  Stefano Lattarini  <stefano.lattarini@gmail.com>
11173 
11174 	release: generate a stub for the release announcement
11175 
11176 	It's much better than having to write it my hand each time; after all,
11177 	most of it is either boilerplate or a cope of NEWS entries.
11178 
11179 	* Makefile.am (determine_release_type): Also set the shell variable
11180 	'$announcement_type' appropriately.
11181 	(print-release-type): Print the value of this new variable as well.
11182 	(announcement): New phony target, generate a files with the same name.
11183 	The recipe uses the shell variable '$announcement_type'.
11184 	(CLEANFILES): Clean it.
11185 	(PACKAGE_MAILINGLIST): New make macro, used when generating the
11186 	'announcement' file.
11187 	* HACKING: Explain how to take advantage of the new convenience
11188 	target.
11189 	* .gitignore: Ignore the 'announcement' file.
11190 
11191 2012-12-08  Stefano Lattarini  <stefano.lattarini@gmail.com>
11192 
11193 	release: distinguish major and minor releases
11194 
11195 	This is mostly a preparatory change, in view of future
11196 	improvements in our release process.
11197 
11198 	* Makefile.am (determine_release_type): Also distinguish
11199 	between major and minor releases, with the help of ...
11200 	(stable_major_version_rx, stable_minor_version_rx): ... these
11201 	new macros.
11202 	(stable_version_rx): Remove, it's obsolete now.
11203 	(version_rx): Rename ...
11204 	(base_version_rx): ... like this, and adjust.
11205 	(print-release-type): New target; helps in debugging the code
11206 	that determines the release type.
11207 
11208 2012-11-30  Stefano Lattarini  <stefano.lattarini@gmail.com>
11209 
11210 	tests: avoid two spurious failures on FreeBSD
11211 
11212 	Due to the FreeBSD make bug that causes of the well-known automake
11213 	bug#7884, two of our bison-requiring tests were failing on FreeBSD.
11214 	To avoid this, simply skip the part of these tests affected by that
11215 	bug.
11216 
11217 	* t/yacc-bison-skeleton.sh: Here, using the 'yl_distcheck' function.
11218 	* t/yacc-bison-skeleton-cxx.sh: Likewise.
11219 	* t/yacc-nodist.sh: While at it, use the 'yl_distcheck' here
11220 	as well, instead of running "$MAKE distcheck" only when $MAKE
11221 	is GNU make.
11222 
11223 2012-11-27  Stefano Lattarini  <stefano.lattarini@gmail.com>
11224 
11225 	Merge branch 'depcomp-work'
11226 
11227 	* depcomp-work:
11228 	  depcomp: remove 'icc' depmode, add 'tcc' one
11229 	  news: update w.r.t. recent depcomp changes
11230 	  tests: fix spurious failures with tcc
11231 	  depcomp: improve comments about tcc support
11232 	  depcomp: remove last relics of libtool 1.4 support
11233 	  depcomp: remove an unused variable
11234 	  depcomp: rewrite a sed invocation for less quoting
11235 	  depcomp: whitespace fixes
11236 	  depcomp: shell code style changes
11237 	  depcomp: make some code more self-documenting
11238 	  depcomp: less duplication between AIX and Tru64 modes
11239 	  depcomp: avoid an extra fork when possible
11240 	  depcomp: reduce code duplication
11241 	  depcomp: don't rely on character ranges working as in C locale
11242 	  depcomp: cosmetic fixlets to a comment
11243 	  depcomp: safer quoting in variable expansion
11244 	  depcomp: correctly propagate exit status in exit trap
11245 	  depcomp: use "rmdir", not "rm -rf", to remove the empty lock directory
11246 	  depcomp: remove useless quoting in variable definitions
11247 	  depcomp: be more idiomatic in checking for zero exit statuses
11248 
11249 2012-11-27  Stefano Lattarini  <stefano.lattarini@gmail.com>
11250 
11251 	Merge branch 'maint'
11252 
11253 	* maint:
11254 	  news: we'll remove AM_PROG_MKDIR in 1.14, not 1.13
11255 	  tests: give few vala tests more significant names
11256 	  vala tests: don't use the 'posix' profile, it's no longer supported
11257 	  news: fixup: bug#8847 is fixed in 1.12.6, not in 1.12.5
11258 	  news: document the fix for automake bug#10227
11259 	  tests: fix a spurious testsuite failure on Solaris
11260 	  python: make installed modules find by default on Debian and Ubuntu
11261 
11262 2012-11-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
11263 
11264 	news: we'll remove AM_PROG_MKDIR in 1.14, not 1.13
11265 
11266 	See also commit v1.12.4-158-gdf23daf.
11267 
11268 	* NEWS (Future backward-incompatibilities): Adjust accordingly.
11269 
11270 2012-11-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
11271 
11272 	tests: give few vala tests more significant names
11273 
11274 	* t/vala.sh: Rename ...
11275 	* t/vala-grepping.sh: ... like this.
11276 	* t/vala2.sh: Rename ...
11277 	* t/vala-recursive-setup.sh: ... like this.
11278 	* t/vala3.sh: Rename ...
11279 	* t/vala-non-recursive-setup.sh: ... like this.
11280 	* t/vala4.sh: Rename ...
11281 	* t/vala-configure.sh: ... like this.
11282 	* t/vala5.sh: Rename ...
11283 	* t/vala-per-target-flags.sh: ... like this, and slightly
11284 	improve heading comments.
11285 	* t/list-of-tests.mk: Adjust.
11286 
11287 2012-11-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
11288 
11289 	vala tests: don't use the 'posix' profile, it's no longer supported
11290 
11291 	Fixes automake bug#12934 a.k.a. bug#12522.
11292 
11293 	Some of automake's Vala tests used to rely on "valac --profile posix"
11294 	(which makes it not use glib).  However, the posix profile was removed
11295 	in August 2012, and is no longer present in the GNOME 3.6.2 version
11296 	of valac (see commit ca020bf0 in the vala Git repository).
11297 
11298 
11299 	* t/vala-libs.sh: Adjust to avoid using the posix profile.
11300 	* t/vala-mix.sh: Likewise.
11301 	* t/vala-mix2.sh: Likewise.
11302 	* t/vala-parallel.sh: Likewise.
11303 	* t/vala-vapi.sh: Likewise.
11304 	* t/vala-vpath.sh: Likewise.
11305 	* NEWS: Update and adjust.
11306 
11307 2012-11-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
11308 
11309 	news: fixup: bug#8847 is fixed in 1.12.6, not in 1.12.5
11310 
11311 	* NEWS (Bugs fixed in 1.12.5): So move it from here ...
11312 	(Bugs fixed in 1.12.6): ... to here.
11313 
11314 2012-11-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
11315 
11316 	news: document the fix for automake bug#10227
11317 
11318 	This is a follow-up commit for v1.12.5-14-g1f113f6.
11319 
11320 	* NEWS: Update.
11321 
11322 2012-11-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
11323 
11324 	Merge branch 'python-fix-pr10227' into maint
11325 
11326 	* python-fix-pr10227:
11327 	  python: make installed modules find by default on Debian and Ubuntu
11328 
11329 2012-11-26  Stefano Lattarini  <stefano.lattarini@gmail.com>
11330 
11331 	tests: fix a spurious testsuite failure on Solaris
11332 
11333 	Reported in automake bug#11524.
11334 
11335 	* t/amhello-binpkg.sh: When non-GNU tar is in use, relax grepping of
11336 	"tar cvf ..." output a little more.
11337 
11338 2012-11-25  Stefano Lattarini  <stefano.lattarini@gmail.com>
11339 
11340 	tests: give few aclocal tests more significant names
11341 
11342 	* t/dirlist-abspath.sh: Rename ...
11343 	* t/aclocal-dirlist-abspath.sh: ... like this.
11344 	* t/dirlist.sh: Rename ...
11345 	* t/aclocal-dirlist.sh: ... like this.
11346 	* t/dirlist2.sh: Rename ...
11347 	* t/aclocal-dirlist-globbing.sh: ... like this.
11348 	* t/list-of-tests.mk: Adjust.
11349 
11350 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11351 
11352 	tests: remove a redundant workaround for a testsuite weakness
11353 
11354 	* t/warnings-unknown.sh: Here, since we have improved our wrapper
11355 	scripts for aclocal and automake enough to make that workaround
11356 	unneeded.
11357 
11358 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11359 
11360 	depcomp: remove 'icc' depmode, add 'tcc' one
11361 
11362 	According to the comments in the now-removed test 'depcomp5.test' (removed
11363 	by commit 'v1.11-1829-g5066c1b' of 2012-02-08, "tests: improve and rework
11364 	tests on dependency tracking") "icc 8.0 and greater understand gcc options,
11365 	so depmode is set to gcc".  According to Wikipedia, the 8.0 version of the
11366 	Intel C compiler has been released in December 2003, that is, almost nine
11367 	yeas ago.  So we can remove explicit support for icc 7.x from the depcomp
11368 	script.
11369 
11370 	Since our support for tcc has so far been relied on the icc depmode, we
11371 	must write a new explicit 'tcc' depmode.  That is easy to do, since we can
11372 	rely on the simpler and better-specified format of tcc-specified makefiles;
11373 	for more info, see:
11374 	<http://lists.gnu.org/archive/html/automake/2012-10/msg00043.html>
11375 
11376 	* lib/depcomp (icc): Remove.
11377 	(tcc): New.
11378 	* NEWS: Update.
11379 
11380 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11381 
11382 	news: update w.r.t. recent depcomp changes
11383 
11384 	* NEWS: Here.
11385 
11386 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11387 
11388 	tests: fix spurious failures with tcc
11389 
11390 	* t/depcomp8a.sh, t/depcomp8b.sh: Here, by avoiding initialization of
11391 	global 'extern' variables, preferring functions definitions instead.
11392 
11393 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11394 
11395 	depcomp: improve comments about tcc support
11396 
11397 	* lib/depcomp (icc): Here.  See also tcc commit 0394caf7 "Emit spaces
11398 	for -MD" and automake commit v1.11-768-gb96263 "depcomp: support tcc
11399 	(Tiny C Compiler)".
11400 
11401 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11402 
11403 	depcomp: remove last relics of libtool 1.4 support
11404 
11405 	* lib/depcomp (tru64): Here.
11406 
11407 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11408 
11409 	depcomp: remove an unused variable
11410 
11411 	* lib/depcomp ($alnum): This.
11412 
11413 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11414 
11415 	depcomp: rewrite a sed invocation for less quoting
11416 
11417 	* lib/depcomp (dashmstdout): Here.
11418 
11419 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11420 
11421 	depcomp: whitespace fixes
11422 
11423 	* lib/depcomp: We indent by two spaces, and without using tabs.
11424 
11425 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11426 
11427 	depcomp: shell code style changes
11428 
11429 	* lib/depcomp: Here, in comments and line continuations.
11430 
11431 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11432 
11433 	depcomp: make some code more self-documenting
11434 
11435 	* lib/depcomp (make_dummy_makefile): With the help of this function.
11436 	Use it throughout.
11437 
11438 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11439 
11440 	depcomp: less duplication between AIX and Tru64 modes
11441 
11442 	* lib/depcomp (tru64, aix): In these entries of the big 'case'
11443 	switch, with the help of ...
11444 	(aix_post_process_depfile): ... this new function.
11445 
11446 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11447 
11448 	depcomp: avoid an extra fork when possible
11449 
11450 	* lib/depcomp (set_dir_from): Here.
11451 
11452 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11453 
11454 	depcomp: reduce code duplication
11455 
11456 	* lib/depcomp: Here, when we have to get the directory and basename
11457 	components of objects or source files; do so with the help of ...
11458 	(set_dir_from, set_base_from): ... these new shell functions.
11459 
11460 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11461 
11462 	depcomp: don't rely on character ranges working as in C locale
11463 
11464 	* lib/depcomp (pgcc): Here.
11465 	Related reorganization, with the unconditional definition of ...
11466 	($upper, $lower, $digits, $alnum, $alpha): ... these shell variables.
11467 
11468 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11469 
11470 	depcomp: cosmetic fixlets to a comment
11471 
11472 	* lib/depcomp (pgcc): Here, as that's unportable according to the
11473 	autoconf manual.  In fact, there's no need to re-trap the signals
11474 	just before exiting, since at that point a signal would cause at
11475 	most a useless but harmless attempt to remove the
11476 
11477 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11478 
11479 	depcomp: safer quoting in variable expansion
11480 
11481 	* lib/depcomp (pgcc): Here.  Probably not needed, but since the
11482 	rest of the script seems to employ proper quoting, better to be
11483 	consistent.
11484 
11485 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11486 
11487 	depcomp: correctly propagate exit status in exit trap
11488 
11489 	* lib/depcomp (pgcc): Here.
11490 
11491 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11492 
11493 	depcomp: use "rmdir", not "rm -rf", to remove the empty lock directory
11494 
11495 	* lib/depcomp (pgcc): Here.
11496 
11497 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11498 
11499 	depcomp: remove useless quoting in variable definitions
11500 
11501 	* lib/depcomp (pgcc): Here.
11502 
11503 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11504 
11505 	depcomp: be more idiomatic in checking for zero exit statuses
11506 
11507 	* lib/depcomp: Prefer "if test $stat -ne 0; then ..." over the longer
11508 	"if test $stat -eq 0; then :; else ...".
11509 
11510 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11511 
11512 	tests: give some tests more significant names
11513 
11514 	* t/req.sh: Rename ...
11515 	* t/aclocal-req.sh: ... like this.
11516 	* t/reqd.sh: Rename ...
11517 	* t/add-missing-multiple.sh: ... like this.
11518 	* t/block.sh: Rename ...
11519 	* t/comment-block.sh: ... like this.
11520 	* t/list-of-tests.mk: Adjust.
11521 
11522 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11523 
11524 	Merge branch 'maint'
11525 
11526 	* maint:
11527 	  tests: fix a spurious failure with older flex versions
11528 
11529 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11530 
11531 	python: make installed modules find by default on Debian and Ubuntu
11532 
11533 	This change fixes automake bug#10227.
11534 
11535 	The code used to get the python package directory was wrong for Python 3,
11536 	at least on Debian and Ubuntu distributions.  In the case the installation
11537 	was using the default prefix "/usr/local", python modules were incorrectly
11538 	installed in the directory
11539 
11540 	    /usr/local/lib/python3/dist-packages
11541 
11542 	(which is *not* searched by default), rather than in a directory like
11543 
11544 	    /usr/local/lib/python3.x/dist-packages
11545 
11546 	which is searched by default.
11547 
11548 	* m4/python.m4 (AM_PATH_PYTHON): Try to use the 'sysconfig' module if
11549 	possible, for better interactions with python 3.x.
11550 
11551 	Helped-by: Reuben Thomas <rrt@sc3d.org>
11552 	Helped-by: Roumen Petrov <bugtrack@roumenpetrov.info>
11553 
11554 2012-11-24  Stefano Lattarini  <stefano.lattarini@gmail.com>
11555 
11556 	tests: fix a spurious failure with older flex versions
11557 
11558 	See automake bug#11524 and bug#12836.
11559 
11560 	* t/lex-header.sh: Here, we use flex option '--header-file',
11561 	but that option is not supported by some older flex versions.
11562 	Simply skip this test if such an old flex version is detected.
11563 
11564 2012-11-23  Stefano Lattarini  <stefano.lattarini@gmail.com>
11565 
11566 	tests: avoid a failure due to m4 builtin 'sinclude' in package name
11567 
11568 	Even with Autoconf 2.69, AC_INIT is not very robust when to package
11569 	names that contain some selected m4 builtins; for example:
11570 
11571 	    $ echo 'AC_INIT([sinclude], [1.0])' | autoconf -o/dev/null -
11572 	    stdin:1: warning: file `' included several times
11573 
11574 	    $ echo 'AC_INIT([dnl], [1.0])' | autoconf -o/dev/null -
11575 	    /usr/bin/m4:stdin:1: excess arguments to builtin `m4_define' ignored
11576 	    autom4te: /usr/bin/m4 failed with exit status: 1
11577 
11578 	Reference:
11579 	<http://lists.gnu.org/archive/html/bug-autoconf/2012-11/msg00005.html>
11580 
11581 	* t/aclocal-m4-sinclude.sh (configure.ac): Use 'm4sinclude' as package
11582 	name in the AC_INIT invocation, to avoid tickling the just-described
11583 	bug.
11584 
11585 2012-11-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
11586 
11587 	Merge branch 'maint'
11588 
11589 	* maint:
11590 	  tests: fix a spurious failure when $PYTHON is in the environment
11591 	  python tests: support PEP-3147 installation layout
11592 	  python: uninstall cater to PEP-3147
11593 	  tests: improve a comment
11594 	  tests: honour $PYTHON override
11595 	  tests: typofix in message
11596 	  news: document fix for bug#8847 (PEP-3147, __pycache__)
11597 	  python: improve support for modern python (CPython 3.2 and PyPy)
11598 
11599 2012-11-21  Stefano Lattarini  <stefano.lattarini@gmail.com>
11600 
11601 	Merge branch 'python-pep-3147' into maint
11602 
11603 	* python-pep-3147:
11604 	  tests: fix a spurious failure when $PYTHON is in the environment
11605 	  python tests: support PEP-3147 installation layout
11606 	  python: uninstall cater to PEP-3147
11607 	  tests: improve a comment
11608 	  tests: honour $PYTHON override
11609 	  tests: typofix in message
11610 	  news: document fix for bug#8847 (PEP-3147, __pycache__)
11611 	  python: improve support for modern python (CPython 3.2 and PyPy)
11612 
11613 2012-11-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
11614 
11615 	tests: fix a spurious failure when $PYTHON is in the environment
11616 
11617 	* t/python11.sh: This test doesn't interact well with user-overrides of
11618 	$PYTHON; and, given its particular nature, neither should it be expected
11619 	to honour that override.  Just unset that $PYTHON variable and live
11620 	happy.
11621 
11622 2012-11-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
11623 
11624 	python tests: support PEP-3147 installation layout
11625 
11626 	This fixes several spurious testsuite failures with python >= 3.2,
11627 	introduced by recent commit v1.12.4-43-ge0e99ed, "python: improve
11628 	support for modern python (CPython 3.2 and PyPy)".
11629 
11630 	* t/ax/am-test-lib.sh (python_has_pep3147, pyc_location, py_installed):
11631 	New functions.
11632 	($am_pep3147_tag): New variable.
11633 	* t/py-compile-basic2.sh: Deleted, too difficult to adapt for the
11634 	gain it would offer; move the still relevant parts ...
11635 	* t/py-compile-basic.sh: ... here.  Adapt and adjust the rest of
11636 	the test as well.
11637 	* t/nobase-python.sh: Adapt and adjust.
11638 	* t/py-compile-basedir.sh: Likewise.
11639 	* t/py-compile-destdir.sh: Likewise.
11640 	* t/py-compile-option-terminate.sh: Likewise.
11641 	* t/python-pr10995.sh: Likewise.
11642 	* t/python-virtualenv.sh: Likewise.
11643 	* t/python10.sh: Likewise.
11644 	* t/python12.sh: Likewise.
11645 	* t/python3.sh: Likewise.
11646 	* t/list-of-tests.mk: Adjust list of tests.
11647 
11648 2012-11-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
11649 
11650 	python: uninstall cater to PEP-3147
11651 
11652 	After recent commit commit v1.12.4-43-ge0e99ed, "python: improve
11653 	support for modern python (CPython 3.2 and PyPy)", the python install
11654 	rules have been made smart enough to install "*.pyc" byte-compiled
11655 	files according to PEP-3147 with modern (post-3.2) pythons.  However,
11656 	the uninstall rules hadn't been updated accordingly, causing leftover
11657 	files to remain around after "make uninstall", as well as failures in
11658 	"make distcheck".
11659 
11660 	* lib/am/python.am (am__pep3147_tweak): New internal macro, used
11661 	in ...
11662 	(uninstall-%DIR%PYTHON): ... the recipe of this target, which has
11663 	been adjusted to cater to PEP-3147
11664 
11665 2012-11-19  Stefano Lattarini  <stefano.lattarini@gmail.com>
11666 
11667 	tests: more explicative names for few aclocal tests
11668 
11669 	* t/aclocal3.sh: Rename ...
11670 	* t/aclocal-missing-macros.sh: ... like this.
11671 	* t/aclocal4.sh: Rename ...
11672 	* t/aclocal-deps.sh: ... like this.
11673 	* t/aclocal5.sh: Rename ...
11674 	* t/aclocal-deps-subdir.sh: ... like this.
11675 	* t/aclocal6.sh: Rename ...
11676 	* t/aclocal-remake-misc.sh: ... like this.
11677 	* t/aclocal7.sh: Rename ...
11678 	* t/aclocal-no-force.sh: ... like this.
11679 	* t/acloca8.sh: Rename ...
11680 	* t/aclocal-no-unused-required.sh: ... like this.
11681 	* t/acloca9.sh: Rename ...
11682 	* t/aclocal-I-order-1.sh: ... like this.
11683 	* t/acloca11.sh: Rename ...
11684 	* t/aclocal-I-order-2.sh: ... like this, and update heading comments.
11685 	* t/acloca12.sh: Rename ...
11686 	* t/aclocal-I-order-3.sh: ... like this, and update heading comments.
11687 	* t/acloca13.sh: Rename ...
11688 	* t/aclocal-m4-sinclude.sh: ... like this, and update heading comments.
11689 	* t/acloca14.sh: Rename ...
11690 	* t/aclocal-m4-include-are-scanned.sh: ... like this.
11691 	* t/acloca14.sh: Rename ...
11692 	* t/aclocal-m4-include-are-scanned-aclocal-amflags.sh: ... like this,
11693 	and update heading comments.
11694 	* t/acloca15.sh: Rename ...
11695 	* t/aclocal-scan-configure-ac-pr319.sh: ... like this.
11696 	* t/acloca16.sh: Rename ...
11697 	* t/aclocal-no-extra-scan.sh: ... like this, and enhance slightly.
11698 	* t/acloca18.sh: Rename ...
11699 	* t/aclocal-serial.sh: ... like this.
11700 	* t/acloca19.sh: Rename ...
11701 	* t/aclocal-pr450.sh: ... like this.
11702 	* t/acloca20.sh: Rename ...
11703 	* t/aclocal-no-symlinked-overwrite.sh: ... like this.
11704 	* t/acloca21.sh: Rename ...
11705 	* t/aclocal-comments-respected.sh: ... like this.
11706 	* t/acloca22.sh: Rename ...
11707 	* t/aclocal-deleted-header.sh: ... like this.
11708 	* t/acloca22b.sh: Rename ...
11709 	* t/aclocal-deleted-header-aclocal-amflags.sh: ... like this.
11710 	* t/acloca23.sh: Rename ...
11711 	* t/aclocal-underquoted-defun.sh: ... like this.
11712 	* t/list-of-tests.mk: Update.
11713 
11714 2012-11-17  Stefano Lattarini  <stefano.lattarini@gmail.com>
11715 
11716 	tests: improve coverage on Texinfo + silent rules
11717 
11718 	* t/silent-texi.sh: Here, by basic importing the version of the
11719 	test that is the 'ng/master' branch (that of Automake-NG), plus
11720 	a minor tweak to avoid a spurious failure with NetBSD make.
11721 
11722 2012-11-17  Stefano Lattarini  <stefano.lattarini@gmail.com>
11723 
11724 	Merge branch 'maint'
11725 
11726 	* maint:
11727 	  build: fix rebuild rules for Makefile.in and aclocal.m4
11728 	  hacking: release procedure: fix order of some steps
11729 	  maint: post-release minor version bump
11730 	  release: stable release 1.12.5
11731 
11732 2012-11-17  Stefano Lattarini  <stefano.lattarini@gmail.com>
11733 
11734 	build: fix rebuild rules for Makefile.in and aclocal.m4
11735 
11736 	They have been broken after our removal of the 'perllibdir' hack
11737 	in yesterday's commit v1.12.4-45-g4872dfe.
11738 
11739 	* configure.ac: Rather then redefining AUTOMAKE and ACLOCAL to
11740 	explicitly override the perllibdir, the acdir and the libdir, use
11741 	the ready-made t/wrap/{aclocal,automake}-$APIVERSION scripts.
11742 	* NEWS: Document the fix.
11743 
11744 2012-11-17  Stefano Lattarini  <stefano.lattarini@gmail.com>
11745 
11746 	hacking: release procedure: fix order of some steps
11747 
11748 	* HACKING (Release procedure): The manuals should be rebuilt and
11749 	uploaded at at www.gnu.org *before* bumping the version number to
11750 	the next alpha release.
11751 
11752 2012-11-17  Stefano Lattarini  <stefano.lattarini@gmail.com>
11753 
11754 	maint: post-release minor version bump
11755 
11756 	* configure.ac (AC_INIT): Bump version number to 1.12.5a.
11757 	* m4/amversion.m4: Likewise (automatically regenerated by
11758 	"make bootstrap").
11759 
11760 2012-11-17  Stefano Lattarini  <stefano.lattarini@gmail.com>
11761 
11762 	release: stable release 1.12.5
11763 
11764 	* configure.ac (AC_INIT): Bump version number to 1.12.5.
11765 	* m4/amversion.m4: Likewise (auto-updated by "make bootstrap").
11766 
11767 2012-11-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
11768 
11769 	Merge branch 'maint'
11770 
11771 	* maint:
11772 	  maintcheck: avoid failures, sync with 'ng/master' more
11773 	  maintcheck: remove 'sc_perl_syntax'
11774 	  sync: update files from upstream with "make fetch"
11775 	  maint: get rid of $perllibdir environment variable hack
11776 	  configure: respect the '-q' option better
11777 	  cosmetics: de-tabify configure.ac
11778 
11779 2012-11-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
11780 
11781 	maintcheck: avoid failures, sync with 'ng/master' more
11782 
11783 	* syntax-checks.mk (sc_diff_aclocal_in_aclocal,
11784 	sc_diff_automake_in_automake): Rename ...
11785 	(sc_diff_aclocal, sc_diff_automake): ... respectively to these, rewrite
11786 	to use static pattern rules and 'diff -u' instead of bare 'diff'.  This
11787 	makes the rules more robust in the face of edits to 'automake.in' and
11788 	'aclocal.in', and get them in sync with the corresponding checks in the
11789 	'ng/master' branch.
11790 	(syntax_check_rules): Adjust.
11791 
11792 2012-11-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
11793 
11794 	maintcheck: remove 'sc_perl_syntax'
11795 
11796 	It is quite useless, since a "make bootstrap" or "make check" would
11797 	immediately catch a breakage in the 'aclocal' or 'automake' scripts
11798 	anyway.  In addition, the recent removal of the 'perllibdir' environment
11799 	variable hack has broken this check.
11800 
11801 	* syntax-checks.mk (sc_perl_syntax): Remove.
11802 	(syntax_check_rules): No longer list it.
11803 
11804 2012-11-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
11805 
11806 	sync: update files from upstream with "make fetch"
11807 
11808 	* lib/texinfo.tex, lib/config.sub: Update.
11809 
11810 2012-11-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
11811 
11812 	maint: get rid of $perllibdir environment variable hack
11813 
11814 	That was needed when the automake-$APIVERSION and aclocal-$APIVERSION
11815 	wrapper scripts used by the testsuite were shell script, rather than
11816 	perl scripts acting as a thin layers.
11817 
11818 	* aclocal.in (BEGIN): Update @INC based on the contents of the array
11819 	'@Aclocal::perl_libdirs', rather than of the environment variable
11820 	'perllibdir'.
11821 	* t/wrap/aclocal.in (BEGIN): Initialize '@Aclocal::perl_libdirs'
11822 	rather than $ENV{'perllibdir'}.
11823 	* automake.in (BEGIN): Update @INC based on the contents of the array
11824 	'@Automake::perl_libdirs', rather than of the environment variable
11825 	'perllibdir'.
11826 	* t/wrap/automake.in (BEGIN): Initialize '@Automake::perl_libdirs'
11827 	rather than $ENV{'perllibdir'}.
11828 	* NEWS: Update.
11829 
11830 2012-11-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
11831 
11832 	configure: respect the '-q' option better
11833 
11834 	* configure.ac: Here, by avoiding to print the warnings about
11835 	using a non-stable Automake version if the '$silent' variable
11836 	is set to "yes".
11837 
11838 2012-11-16  Stefano Lattarini  <stefano.lattarini@gmail.com>
11839 
11840 	cosmetics: de-tabify configure.ac
11841 
11842 2012-11-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
11843 
11844 	fixup: Automake installation works again
11845 
11846 	By renaming 'dist_automake_ac_DATA' to 'nobase_dist_automake_ac_DATA' in
11847 	a recent patch, we have broken the Automake installation, since m4 files
11848 	that were expected to be installed in '${prefix}/aclocal-${APIVERSION}'
11849 	are now being installed in '${prefix}/aclocal-${APIVERSION}/m4' instead.
11850 	Fix this mess.
11851 
11852 	* Makefile.am (nobase_dist_automake_ac_DATA): Rename ...
11853 	(dist_automake_ac_DATA): ... like this.  And move
11854 	'm4/internal/ac-config-macro-dirs.m4' ...
11855 	(dist_automake_internal_ac_DATA): ... here.
11856 	(automake_internal_acdir): New, simply defined to
11857 	'$(automake_acdir)/internal'.
11858 
11859 2012-11-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
11860 
11861 	Merge branch 'ac-config-macro-dirs'
11862 
11863 	* ac-config-macro-dirs:
11864 	  tests: AC_CONFIG_MACRO_DIRS: ignore inevitable failures with old autoconf
11865 	  aclocal: AC_CONFIG_MACRO_DIRS: work around autom4te option parsing bugs
11866 	  aclocal: tracing AC_CONFIG_MACRO_DIRS can work with older autoconf as well
11867 
11868 2012-11-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
11869 
11870 	tests: AC_CONFIG_MACRO_DIRS: ignore inevitable failures with old autoconf
11871 
11872 	When "older" version of autoconf are used (that is, those before commit
11873 	v2.69-44-g1ed0548), we have no sane way to prevent the autom4te invocation
11874 	issued from aclocal to possibly display warnings "MACRO m4_require'd but
11875 	not m4_defun'd".  That's not a big deal, because that just means that
11876 	people using pre-2.70 autoconf with cutting-edge automake will see few
11877 	spurious warnings, but the actual semantics will remain correct.
11878 
11879 	However, this blemish was causing a couple of annoying testsuite failures.
11880 	Solve this by simply skipping the affected tests when older (pre-2.70)
11881 	autoconf versions are used.
11882 
11883 	* t/aclocl-macrodir.tap ("AC_CONFIG_MACRO_DIR interaction with
11884 	AC_REQUIRE"): Skip when older autoconf is in use.
11885 	* t/aclocl-macrodirs.tap ("AC_CONFIG_MACRO_DIR interaction with
11886 	AC_REQUIRE"): Likewise.
11887 
11888 2012-11-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
11889 
11890 	recursion: remove _AM_EXTRA_RECURSIVE_TARGETS indirection
11891 
11892 	Since automake automatically smashes extra whitespace and newlines
11893 	in the arguments of the macros it traces, we can remove the extra
11894 	indirection with the _AM_EXTRA_RECURSIVE_TARGETS.  This doesn't
11895 	truly simplify the existing code, but make it slightly easier to
11896 	understand anyway, since leave it with one less indirection.
11897 
11898 	* automake.in: Trace AM_EXTRA_RECURSIVE_TARGETS directly, rather
11899 	than tracing _AM_EXTRA_RECURSIVE_TARGETS.  Remove trailing and
11900 	leading empty fields in the extracted arguments.
11901 	* m4/extra-recurs.m4 (_AM_EXTRA_RECURSIVE_TARGETS): Remove, no longer
11902 	needed.
11903 	(AM_EXTRA_RECURSIVE_TARGETS): Adjust by making it a simple no-op.
11904 
11905 2012-11-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
11906 
11907 	aclocal: AC_CONFIG_MACRO_DIRS: work around autom4te option parsing bugs
11908 
11909 	The autom4te program coming with autoconf 2.68 and earlier had a bug
11910 	which caused the "-" command line argument (with which we tell it to
11911 	read some input from from standard input) to aways be pushed at the
11912 	*end* of the command line, regardless of where the user specified it
11913 	(that bug was fixed by autoconf commit 'v2.68-120-gf4be358', "getopt:
11914 	new Autom4te::Getopt module").
11915 
11916 	This broken semantics conflict with our usage in aclocal, where we
11917 	need to pass some input to the invoked autom4te program early, and
11918 	have so far been using the stdin to do so.  Now we start using an
11919 	external file instead.
11920 
11921 	* m4/internal/ac-config-macro-dirs.m4: New file, contain a fallback
11922 	definition of the AC_CONFIG_MACRO_DIRS macro for older autoconf
11923 	releases.
11924 	* aclocal.in (trace_used_macros): When invoking autom4te, use that
11925 	file instead of "abusing" standard input.
11926 	* Makefile.am (dist_automake_ac_DATA): Rename ...
11927 	(nobase_dist_automake_ac_DATA): ... like this.
11928 	Add 'm4/internal/ac-config-macro-dirs.m4' to it.
11929 	* t/aclocal-acdir.sh: Adjust to avoid spurious failures.
11930 
11931 	Helped-by: Eric Blake <eblake@redhat.com>
11932 
11933 2012-11-15  Stefano Lattarini  <stefano.lattarini@gmail.com>
11934 
11935 	aclocal: tracing AC_CONFIG_MACRO_DIRS can work with older autoconf as well
11936 
11937 	This will allow our users to interact also with pre-2.70 autoconf without
11938 	need for the user to add ACLOCAL_AMFLAGS in Makefile.am.  For example,
11939 	before this change, in order to have aclocal look for macros in 'm4/dir1'
11940 	and 'm4/dir2' also when (say) autoconf 2.69 was used, our users would
11941 	have had to add something like:
11942 
11943 	    ACLOCAL_AMFLAGS = -I m4/dir1 -I m4/dir2
11944 
11945 	in Makefile.am, in addition to the
11946 
11947 	    AC_CONFIG_MACRO_DIRS([m4/dir1 m4/dir2])
11948 
11949 	in configure.ac.  Now, the AC_CONFIG_MACRO_DIRS call is enough.
11950 
11951 	See the long-winded discussion on automake bug#12845 for more details:
11952 	<http://debbugs.gnu.org/cgi/bugreport.cgi?bug=12845>
11953 
11954 	* aclocal.in ($ac_config_macro_dirs_fallback): New global variable,
11955 	contains m4 code to issue a fallback definition of AC_CONFIG_MACRO_DIRS
11956 	as an alias for the private macro _AM_CONFIG_MACRO_DIRS.
11957 	(trace_used_macros): Handle and trace that macro.  Do some code
11958 	reorganization and fix related botched indentation while at it.
11959 	(write_aclocal): Output '$ac_config_macro_dirs_fallback' early in
11960 	the generated aclocal.m4.
11961 	* t/aclocal-macrodirs.tap: Run unconditionally, even with older
11962 	autoconf.
11963 	* t/subpkg-macrodir.sh: Likewise.
11964 	* doc/automake.texi: Document only AC_CONFIG_MACRO_DIRS, rather
11965 	than AC_CONFIG_MACRO_DIR.
11966 
11967 2012-11-14  Stefano Lattarini  <stefano.lattarini@gmail.com>
11968 
11969 	Merge branch 'ac-config-macro-dirs'
11970 
11971 	* ac-config-macro-dirs:
11972 	  news: we trace AC_CONFIG_MACRO_DIRS
11973 	  aclocal: trace AC_CONFIG_MACRO_DIR_TRACE
11974 	  aclocal: avoid spurious warnings from autom4te with AC_CONFIG_MACRO_DIRS
11975 	  coverage: expose a bug in aclocal (spurious warnings)
11976 	  aclocal: smash newlines in arguments of traced macros
11977 	  tests: better coverage for AC_CONFIG_MACRO_DIRS
11978 	  aclocal: diagnose non-existing directories in AC_CONFIG_MACRO_DIRS better
11979 	  aclocal: multiple local m4 macro dirs with AC_CONFIG_MACRO_DIRS
11980 
11981 2012-11-11  Stefano Lattarini  <stefano.lattarini@gmail.com>
11982 
11983 	tests: improve a comment
11984 
11985 	* t/ax/am-test-lib.sh (require_tool): Here, about python support
11986 	for '--version' option.
11987 
11988 2012-11-11  Stefano Lattarini  <stefano.lattarini@gmail.com>
11989 
11990 	tests: honour $PYTHON override
11991 
11992 	* t/ax/am-test-lib.sh (require_tool): Here.
11993 	* t/python-too-old.sh: And here.
11994 	* t/python-vars.sh: And here.
11995 	* t/python-virtualenv.sh: And here.  Also add some sanity
11996 	checks while at it.
11997 
11998 2012-11-11  Stefano Lattarini  <stefano.lattarini@gmail.com>
11999 
12000 	tests: typofix in message
12001 
12002 	* t/python-virtualenv.sh: Here.
12003 
12004 2012-11-11  Stefano Lattarini  <stefano.lattarini@gmail.com>
12005 
12006 	news: document fix for bug#8847 (PEP-3147, __pycache__)
12007 
12008 	* NEWS (Bugs fixed in 1.12.5): Here.
12009 
12010 2012-11-10  Yaakov Selkowitz  <yselkowitz@users.sourceforge.net>  (tiny change)
12011 
12012 	python: improve support for modern python (CPython 3.2 and PyPy)
12013 
12014 	This fixes automake bug#8847.
12015 
12016 	* m4/python.m4 (AM_PATH_PYTHON): Add python3.3 to
12017 	_AM_PYTHON_INTERPRETER_LIST.
12018 	* lib/py-compile: Fix compiled filenames for PEP-3147, currently
12019 	implemented in CPython 3.2 and newer.  Do not create '.pyo' files
12020 	for PyPy.
12021 
12022 2012-11-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
12023 
12024 	news: we trace AC_CONFIG_MACRO_DIRS
12025 
12026 	* NEWS (New in 1.13): Update accordingly.
12027 
12028 2012-11-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
12029 
12030 	news: improve wording
12031 
12032 	* NEWS (New in 1.13): Prefer "no longer" over "not anymore",
12033 	wherever possible.  Related re-wrapping of text.
12034 
12035 2012-11-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
12036 
12037 	aclocal: trace AC_CONFIG_MACRO_DIR_TRACE
12038 
12039 	This is now the documented way, and indeed the only correct way, to
12040 	support AC_CONFIG_MACRO_DIR and AC_CONFIG_MACRO_DIRS in third-party
12041 	tools.  See autoconf commit v2.69-45-g5269030 of 2012-11-09,
12042 	"AC_CONFIG_MACRO_DIRS: improve tracing and add sanity checks".
12043 
12044 	* aclocal.in (trace_used_macros): Trace 'AC_CONFIG_MACRO_DIR_TRACE'
12045 	rather than 'AC_CONFIG_MACRO_DIRS'.  We still have to trace
12046 	'AC_CONFIG_MACRO_DIR' explicitly though, for compatibility with
12047 	Autoconf versions before 2.70.
12048 	* t/aclocal-macrodirs.tap: Enhance w.r.t. precedence of arguments
12049 	in one and several AC_CONFIG_MACRO_DIRS calls.
12050 
12051 2012-11-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
12052 
12053 	aclocal: avoid spurious warnings from autom4te with AC_CONFIG_MACRO_DIRS
12054 
12055 	When some macro expanded in configure.ac calls AC_REQUIRE on another
12056 	macro that is defined in one of the local m4 macro dirs specified
12057 	with AC_CONFIG_MACRO_DIRS, aclocal prints spurious warnings like:
12058 
12059 	    configure.ac:4: warning: MY_BAR is m4_require'd but not m4_defun'd
12060 	    configure.ac:3: MY_FOO is expanded from...
12061 
12062 	Such warnings come from autom4te, and are due to the fact that the
12063 	*first* autom4te invocation issued by aclocal is not yet able to
12064 	"see" the m4 macro definitions in the local m4 dirs (because they
12065 	can be looked for only after the AC_CONFIG_MACRO_DIRS call has been
12066 	traced, and tracing it requires running autom4te).
12067 
12068 	To allow us to work around this issue, autom4te has introduced a new
12069 	"witness" macro 'm4_require_silent_probe', that, when defined, allows
12070 	us to silence that particular kind of warnings (and only it).
12071 
12072 	Reported by Nick Bowler; see point (4) of:
12073 	<http://lists.gnu.org/archive/html/autoconf-patches/2012-11/msg00000.html>
12074 
12075 	* aclocal.in (trace_used_macros): Pre-define the special macro
12076 	'm4_require_silent_probe' when invoking autom4te.
12077 	* t/aclocal-macrodirs.tap ("AC_CONFIG_MACRO_DIR interaction with
12078 	AC_REQUIRE"): This test passes now: remove the "TODO" directive.
12079 	* t/aclocal-macrodir.tap ("AC_CONFIG_MACRO_DIRS interaction with
12080 	AC_REQUIRE"): Likewise.
12081 	* t/acloca17.sh: Remove.
12082 	* t/list-of-tests.mk: Adjust.
12083 
12084 2012-11-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
12085 
12086 	coverage: expose a bug in aclocal (spurious warnings)
12087 
12088 	When some macro expanded in configure.ac calls AC_REQUIRE on another
12089 	macro that is defined in one of the local m4 macro dirs specified
12090 	with one of the macros AC_CONFIG_MACRO_DIRS or AC_CONFIG_MACRO_DIR,
12091 	aclocal prints spurious warnings like:
12092 
12093 	    configure.ac:4: warning: MY_BAR is m4_require'd but not m4_defun'd
12094 	    configure.ac:3: MY_FOO is expanded from...
12095 
12096 	Expose this weakness in our testsuite.
12097 
12098 	Reported by Nick Bowler; see point (4) of:
12099 	<http://lists.gnu.org/archive/html/autoconf-patches/2012-11/msg00000.html>
12100 
12101 	* t/aclocal-macrodir.tap ("AC_CONFIG_MACRO_DIR interaction with
12102 	AC_REQUIRE"): New test, still xfailing.
12103 	* t/aclocal-macrodirs.tap ("AC_CONFIG_MACRO_DIRS interaction with
12104 	AC_REQUIRE"): Likewise.
12105 
12106 2012-11-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
12107 
12108 	aclocal: smash newlines in arguments of traced macros
12109 
12110 	This change fixes the existing issues with AC_CONFIG_MACRO_DIRS
12111 	containing newlines:
12112 	<http://lists.gnu.org/archive/html/autoconf-patches/2012-11/msg00000.html>
12113 
12114 	Likely, it will also allow a less involved implementation of the
12115 	AM_EXTRA_RECURSIVE_TARGETS macro (but that is left for potential
12116 	follow-up patches).
12117 
12118 	* aclocal.in (trace_used_macros): When calling autom4te, pass its
12119 	'--trace' option an argument containing '${::}%' rather than '$1'.
12120 	According to the autoconf manual (as of version 2.69), that will expand
12121 	to the concatenation, with the '::' string, of all the arguments passed
12122 	to a macro, with all newline characters in such arguments smashed.
12123 	Related adjustments when handling the macro AC_CONFIG_MACRO_DIRS, to
12124 	ensure leading whitespace in its argument are handled correctly.
12125 	* t/aclocal-macrodirs.tap ("AC_CONFIG_MACRO_DIRS: extra whitespace"):
12126 	No longer declare it as an xfailing test.
12127 
12128 2012-11-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
12129 
12130 	tests: better coverage for AC_CONFIG_MACRO_DIRS
12131 
12132 	Suggested by a report from Nick Bowler:
12133 	<http://lists.gnu.org/archive/html/autoconf-patches/2012-11/msg00000.html>
12134 
12135 	* t/aclocal-macrodirs.tap: Enhance to check use of extra whitespaces
12136 	and newline characters in the arguments to AC_CONFIG_MACRO_DIRS.  This
12137 	test currently fails, so mark it as "TODO".  While at it, throw in other
12138 	minor edits and enhancements.
12139 
12140 2012-11-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
12141 
12142 	aclocal: diagnose non-existing directories in AC_CONFIG_MACRO_DIRS better
12143 
12144 	This new implementation ensures that any directory (possibly excluding
12145 	the first one, if the '--install' option is used) that is declared with
12146 	AC_CONFIG_MACRO_DIRS and that is non-existent will cause an error from
12147 	aclocal.
12148 
12149 	* aclocal.in (scan_m4_dirs): Add a new argument, telling whether it's OK
12150 	for the scanned directory to be non-existing.  Adjust the implementation
12151 	accordingly.
12152 	($first_user_m4dir): Remove, no more needed.
12153 	(scan_m4_files): Update 'scan_m4_dirs' invocations so that aclocal will
12154 	not complain if the first user macro directory is non-existing and the
12155 	'--install' option is given: such directory will be created later by
12156 	aclocal itself.
12157 	* t/aclocal-macrodir.tap: Do not mark the last test as TODO anymore;
12158 	it now passes.  Make stricter by ensuring a non-existing directory in
12159 	AC_CONFIG_MACRO_DIRS causes an hard error, not a warning.
12160 
12161 2012-11-10  Stefano Lattarini  <stefano.lattarini@gmail.com>
12162 
12163 	aclocal: multiple local m4 macro dirs with AC_CONFIG_MACRO_DIRS
12164 
12165 	A new macro 'AC_CONFIG_MACRO_DIRS' has been recently introduced in
12166 	autoconf (and is expected to appear in the autoconf 2.70 release),
12167 	allowing us to declare several local m4 macro directories for a
12168 	package.
12169 
12170 	It can be done either passing several arguments to a single invocation:
12171 
12172 	    AC_CONFIG_MACRO_DIRS([dir1 dir2])
12173 
12174 	or issuing more invocations:
12175 
12176 	    AC_CONFIG_MACRO_DIRS([dir1])
12177 	    AC_CONFIG_MACRO_DIRS([dir2])
12178 
12179 	or a combination of the two:
12180 
12181 	    AC_CONFIG_MACRO_DIRS([dir1 dir2])
12182 	    AC_CONFIG_MACRO_DIRS([dir3])
12183 
12184 	This will allow projects to use several m4 macro local dirs, without the
12185 	need to use ACLOCAL_AMFLAGS (which we want to make obsolete and finally
12186 	remove).  This is especially important for projects that are used as
12187 	nested subpackages of larger projects.
12188 
12189 	For more information and rationales, refer to these past discussions:
12190 
12191 	<http://lists.gnu.org/archive/html/autoconf/2011-12/msg00037.html>
12192 	<http://lists.gnu.org/archive/html/automake-patches/2012-07/msg00010.html>
12193 	<http://lists.gnu.org/archive/html/autoconf-patches/2012-07/msg00000.html>
12194 	<http://lists.gnu.org/archive/html/autoconf-patches/2012-07/msg00012.html>
12195 	<http://thread.gmane.org/gmane.comp.sysutils.autoconf.patches/8037/>
12196 	<http://thread.gmane.org/gmane.comp.sysutils.autoconf.patches/8087>
12197 	<http://thread.gmane.org/gmane.comp.sysutils.automake.patches/8956>
12198 
12199 	as well as to Automake commit v1.12.1-165-gcd1a9cc of 2012-07-03,
12200 	"aclocal: deprecate ACLOCAL_AMFLAGS, trace AC_CONFIG_MACRO_DIR instead",
12201 	autoconf commit v2.69-42-gd73770f of 2012-10-17, "AC_CONFIG_MACRO_DIRS:
12202 	new macro, mostly for aclocal".
12203 
12204 	* aclocal.in ($ac_config_macro_dir): Turn this global scalar it into ...
12205 	(@ac_config_macro_dirs): ... this global array.
12206 	(trace_used_macros): Update '@ac_config_macro_dirs' instead of
12207 	re-defining '$ac_config_macro_dir'.  Cater to calls the now-preferred
12208 	macro 'AC_CONFIG_MACRO_DIRS' in addition to the "obsolescent" one
12209 	AC_CONFIG_MACRO_DIR.
12210 	(main loop): Append '@ac_config_macro_dirs', not '$ac_config_macro_dir',
12211 	to '@user_includes'.
12212 	* t/subpkg-macrodir.sh: New test.
12213 	* t/aclocal-macrodirs.tap: Likewise.
12214 	* t/list-of-tests.mk: Add them.
12215 	* t/aclocal-macrodir.tap: Adjust and extend a little to keep it more in
12216 	sync with 'aclocal-macrodirs.tap'.
12217 
12218 2012-11-07  Stefano Lattarini  <stefano.lattarini@gmail.com>
12219 
12220 	tests: remove an obsolete comment
12221 
12222 	* t/silent-yacc.sh: Here.
12223 
12224 2012-11-07  Stefano Lattarini  <stefano.lattarini@gmail.com>
12225 
12226 	tests: remove mostly-redundant tests on silent rules
12227 
12228 	We used to have several couples of tests named like 'silent-foo-gcc.sh'
12229 	and 'silent-foo-generic.sh'.  Differently from what the names suggest,
12230 	the first test in such a couple (that is, "silent-foo-gcc.sh") was not
12231 	meant to check specific GCC-related features, but rather to check how
12232 	the silent rules behave in combination with automatic dependency tracking
12233 	when the 'gcc' depmode (that targets GCC versions before the 3.x and 4.x
12234 	release series) is forced.
12235 
12236 	Such depmode forcing was done exclusively to cover the code paths in
12237 	'lib/am/depend2.am' that actually invoke the 'depcomp' script, rather
12238 	than using the inlined, GCC-specific compiler invocation (the so-called
12239 	"fastdep" mode), which is the default with modern GCC or with other
12240 	modern compilers that can emulate the GCC command-line interface (e.g.,
12241 	clang 3.0).
12242 
12243 	But whenever we run the silent-*-generic.sh" tests with other supported
12244 	compilers which have an associated depmode different from 'gcc3', these
12245 	same "non-fastdep" code paths are covered, since in those tests we run
12246 	./configure with the '--enable-dependency-tracking' option, which causes
12247 	slower depmodes not to be rejected.  Examples of such compilers are the
12248 	Sun C and C++ compilers (at least since version 5.9, a.k.a. Sun Studio
12249 	12.1), and the Tiny C Compiler (from version 0.9.26); and I run the
12250 	Automake testsuite quite regularly with those compilers.
12251 
12252 	So, the "silent-*-gcc.sh" test cases don't offer any real coverage
12253 	enhancements, while still using testsuite runtime and causing some
12254 	(admittedly minor, but still annoying) synchronization headaches with
12255 	the sister tests "silent-foo-general.sh" tests.  So let's just remove
12256 	these "silent-*-gcc.sh" tests.
12257 
12258 	* t/silent-c-gcc.sh: Remove.
12259 	* t/silent-cxx-gcc.sh: Likewise.
12260 	* t/silent-lt-gcc.sh: Likewise.
12261 	* t/silent-many-gcc.sh: Likewise.
12262 	* t/silent-c-generic.sh: Rename ...
12263 	* t/silent-c.sh: ... like this, and adjust heading comments.
12264 	* t/silent-cxx-generic.sh: Rename ...
12265 	* t/silent-cxx.sh: ... like this, and adjust heading comments.
12266 	* t/silent-lt-generic.sh: Rename ...
12267 	* t/silent-lt.sh: ... like this, and adjust heading comments.
12268 	* t/silent-many-generic.sh: Rename ...
12269 	* t/silent-many-languages.sh: ... like this, and adjust heading
12270 	comments.
12271 	* t/list-of-tests.mk: Adjust.
12272 
12273 2012-11-07  Stefano Lattarini  <stefano.lattarini@gmail.com>
12274 
12275 	tests: enhance coverage on silent-rules and C compilation rules a little
12276 
12277 	* t/silent-c-generic.sh: Here, by trying them also with automatic
12278 	dependency tracking explicitly disabled.
12279 
12280 2012-11-07  Stefano Lattarini  <stefano.lattarini@gmail.com>
12281 
12282 	Merge branch 'maint'
12283 
12284 	* maint:
12285 	  tests: new variable $am_testaux_builddir
12286 	  tests: rename $am_testauxdir -> $am_testaux_srcdir
12287 
12288 2012-11-07  Stefano Lattarini  <stefano.lattarini@gmail.com>
12289 
12290 	tests: rename few tests (on silent-rules support)
12291 
12292 	Either to more informative names, or to names more consistent with the
12293 	ones of similar/related tests.
12294 
12295 	* t/silent.sh: Rename ...
12296 	* t/silent-c-generic.sh: ... like this, and adjust heading comments.
12297 	* t/silent2.sh: Rename ...
12298 	* t/silent-c-gcc.sh: ... like this, and adjust heading comments.
12299 	* t/silent3.sh: Rename ...
12300 	* t/silent-lt-generic.sh: ... like this, and adjust heading comments.
12301 	* t/silent4.sh: Rename ...
12302 	* t/silent-lt-gcc.sh: ... like this, and adjust heading comments.
12303 	* t/silentcxx.sh: Rename ...
12304 	* t/silent-cxx-generic.sh: ... like this, and adjust heading comments.
12305 	* t/silentcxx-gcc.sh: Rename ...
12306 	* t/silent-cxx-gcc.sh: ... like this, and adjust heading comments.
12307 	* t/silentf77.sh: Rename ...
12308 	* t/silent-f77.sh: ... like this, and adjust heading comments.
12309 	* t/silentf90.sh: Rename ...
12310 	* t/silent-f90.sh: ... like this, and adjust heading comments.
12311 	* t/silent7.sh: Rename ...
12312 	* t/silent-gen.sh: ... like this, and adjust heading comments.
12313 	* t/silent6.sh: Rename ...
12314 	* t/silent-custom.sh: ... like this.
12315 	* t/silent8.sh: Rename ...
12316 	* t/silent-texi.sh: ... like this.
12317 	* t/list-of-tests.mk: Adjust.
12318 
12319 2012-11-07  Stefano Lattarini  <stefano.lattarini@gmail.com>
12320 
12321 	tests: merge two tests on silent-rules with libtool
12322 
12323 	This helps to reduce code duplication, and also to (marginally)
12324 	speed up the involved tests.
12325 
12326 	Before this change, we had:
12327 
12328 	  $ diff -u t/silent3.sh t/silent9.sh
12329 
12330 	  --- t/silent3.sh      2012-11-07 19:38:05.000000000 +0100
12331 	  +++ t/silent9.sh      2012-11-07 19:27:54.000000000 +0100
12332 	  @@ -14,9 +14,9 @@
12333 	   # You should have received a copy of the GNU General Public License
12334 	   # along with this program.  If not, see <http://www.gnu.org/licenses/>.
12335 
12336 	  -# Check silent-rules mode, with libtool, standard depmode case.
12337 	  +# Check silent-rules mode, with libtool, nodep case.
12338 
12339 	  -# Please keep this file in sync with 'silent4.sh' and 'silent9.sh'.
12340 	  +# Please keep this file in sync with 'silent3.sh' and 'silent4.sh'.
12341 
12342 	   required='cc libtoolize'
12343 	   . test-init.sh
12344 	  @@ -26,8 +26,8 @@
12345 	   cat >>configure.ac <<'EOF'
12346 	   AC_CONFIG_FILES([sub/Makefile])
12347 	   AC_PROG_CC
12348 	  -AM_PROG_AR
12349 	   AM_PROG_CC_C_O
12350 	  +AM_PROG_AR
12351 	   AC_PROG_LIBTOOL
12352 	   AC_OUTPUT
12353 	   EOF
12354 	  @@ -61,7 +61,7 @@
12355 	   $AUTOMAKE --add-missing
12356 	   $AUTOCONF
12357 
12358 	  -./configure --enable-silent-rules
12359 	  +./configure --disable-dependency-tracking --enable-silent-rules
12360 	   $MAKE >stdout || { cat stdout; exit 1; }
12361 	   cat stdout
12362 	   $EGREP ' (-c|-o)' stdout && exit 1
12363 
12364 	* t/silent9.sh: Remove, merge ...
12365 	* t/silent3.sh: ... in here.
12366 	* t/list-of-tests.mk: Adjust.
12367 
12368 2012-11-07  Stefano Lattarini  <stefano.lattarini@gmail.com>
12369 
12370 	tests: new variable $am_testaux_builddir
12371 
12372 	And a related fix that solves a spurious testsuite failures in
12373 	VPATH builds.
12374 
12375 	* t/ax/test-defs.in ($am_testaux_builddir): New, counterpart of
12376 	$am_testaux_srcdir, but pointing inside the build directory.
12377 	* t/self-check-shell-no-trail-bslash.sh: Use $am_testaux_builddir,
12378 	not $am_testaux_srcdir, when fetching the 'shell-no-trail-bslash'
12379 	script.
12380 
12381 2012-11-07  Stefano Lattarini  <stefano.lattarini@gmail.com>
12382 
12383 	tests: rename $am_testauxdir -> $am_testaux_srcdir
12384 
12385 	This is just a preparatory change in view of a future commit.
12386 
12387 	* t/ax/test-defs.in: Here.
12388 	* t/ax/tap-summary-aux.sh: And here.
12389 	* t/ax/testsuite-summary-checks.sh: And here.
12390 	* t/distcheck-missing-m4.sh: And here.
12391 	* t/distcheck-outdated-m4.sh: And here.
12392 	* t/self-check-shell-no-trail-bslash.sh: And here.
12393 	* t/test-driver-acsubst.sh: And here.
12394 	* t/test-driver-cond.sh: And here.
12395 	* t/test-driver-custom-multitest.sh: And here.
12396 	* t/test-driver-custom-multitest-recheck.sh: And here.
12397 	* t/test-driver-custom-multitest-recheck2.sh: And here.
12398 	* t/testsuite-summary-count-many.sh: And here.
12399 
12400 2012-10-31  Stefano Lattarini  <stefano.lattarini@gmail.com>
12401 
12402 	Merge branch 'maint'
12403