"Fossies" - the Fresh Open Source Software Archive

Member "s-nail-14.9.19/NEWS" (26 Apr 2020, 92791 Bytes) of package /linux/misc/s-nail-14.9.19.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 "NEWS": 14.9.18_vs_14.9.19.

    1 S - n a i l / S - m a i l x  N e w s
    2 ====================================
    3 
    4   mdocmx(7) anchors are denoted by a number-sign #: typing
    5   "^A ANCHOR" while reading the man(1)ual in a capable less(1)
    6   will scroll to the manual's Point-Of-Interest, and pointing
    7   a web browser to the "#ANCHOR" of the online manual works.
    8 
    9 v14.9.19 ("Tufted titmouse (Indianermeise)"), 2020-04-26
   10 --------------------------------------------------------
   11 
   12 Fixing an unknown-8bit/iconv(1/3) misbehaviour when displaying
   13 mails, and making this MUA ready for OpenSSL 3.0.
   14 
   15 Credits, in order of commit appearance: Anirudh Oppiliappan and
   16 Claus Assmann.
   17 
   18 We welcome Anirudh Oppiliappan and Claus Assmann in THANKS.
   19 
   20 git(1) shortlog (edited)
   21 ^^^^^^^^^^^^^^^^^^^^^^^^
   22 
   23 Steffen Nurpmeso (14):
   24 623b64e6 Fix: ensure `environ'ment vars are sh(1) compatible! (Anirudh
   25         Oppiliappan)
   26 f5e202a5 THANKS: Anirudh Oppiliappan
   27 6fa32bdc Fix: a_crese_make_ref_and_cs(): *reply-in-same-charset*: "handle
   28         unknown-8bit"
   29 a9ec20d6 Fix: "revert" [ab0cd3b8] from 2017-10-20.. (Claus Assmann)..
   30 eb2db180 THANKS: Claus Assmann
   31 b0d0920a mime.types: add audio/{aac,opus}, adjust audio/mp4
   32 bc89c8ff Adjustments for OpenSSL 3.0 (and cpp namespace cleanup)
   33 
   34 v14.9.18 ("It is spring time, what a bliss"), 2020-04-18
   35 --------------------------------------------------------
   36 
   37 Some bugfixes and tweaks that accumulated over the months.
   38 
   39 Credits, in order of commit appearance: Jens Schleusener,
   40 Noctambule, Kyle Evans.
   41 
   42 It is courtesy to give a special credit to Coverity.com.
   43 (<https://scan.coverity.com/projects/s-nail>, project 444.)
   44 
   45 We welcome Noctambule and Kyle Evans in THANKS.
   46 
   47 NOTES, ChangeLog (packager-affine)
   48 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
   49 
   50 - Warning: we _will_ have v15-compat=yes as a default in v14.10!
   51 
   52 - Fix for `~f'#327, `~m'#333, `~Q'#335, `~U'#341 and `~u'#342:
   53   no longer include all MIME parts (bug since at least v14.9.16).
   54 
   55 - Fix for handling of "The Mailcap files"#36:
   56   a single copiousoutput/x-mailx-tmpfile-fill|nametemplate
   57   combination was falsely handled, mostly affecting binary file
   58   formats (which got broken by applied character set conversion).
   59 
   60 - Fix for automatic S/MIME encrypted key / certificate password
   61   lookup: used $LOGNAME@[no hostname].smime-cert-key as
   62   a fallback, instead of using the value of *from*#430.
   63   (Noctambule)
   64 
   65 - `~R'#337 and `~r'#338, as well as all prompts which expect
   66   file names and did not yet do so, expect shell-quoted names.
   67   (See "Shell-style argument quoting"#23.)
   68 
   69 ChangeLog
   70 ^^^^^^^^^
   71 
   72 - `~Q'#335 now acts as if *quote*#518 is set.
   73   Does not really make sense otherwise.
   74 
   75 - New environment variable $SOCKS5_PROXY#633 is tight together
   76   with *socks-proxy*#567: setting the one affects the other.
   77   This $SOCKS5_PROXY is introduced by FreeBSD to affect all
   78   programs which are capable to proxy via SOCKS5 (it seems).
   79   (Kyle Evans)
   80 
   81 git(1) shortlog: Steffen Nurpmeso (26)
   82 
   83 v14.9.17 ("To bind, or not to bind.."), 2020-02-02
   84 --------------------------------------------------
   85 
   86 A shadowed key bindings report on bash-bugs@ made me aware that
   87 our `bind' code has never seen a real review, and not only were we
   88 incapable too, but the code was a piece of shit, when i looked at
   89 it.  So please find here a bugfix release to smoothly end v14.9.
   90 
   91 Credits, in order of commit appearance: Jens Schleusener,
   92 Johannes Schöpfer, Russell Bell, and Koichi Murase.
   93 
   94 It is courtesy to give a special credit to Coverity.com.
   95 (<https://scan.coverity.com/projects/s-nail>, project 444.)
   96 
   97 We welcome Koichi Murase in THANKS.
   98 
   99 And now for something completely different.
  100 
  101 NOTES, ChangeLog (packager-affine)
  102 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  103 
  104 - "make test" now works when run by root or on read-only
  105   file-systems.  (Johannes Schöpfer)
  106 
  107 - Our `bind'#149 now supports overall key-sequence timeouts,
  108   as opposed to inter-byte timeouts.  For this i have obsoleted
  109   *bind-timeout* in favour of *bind-inter-byte-timeout*#379
  110   and *bind-inter-key-timeout*#380 (not set by default).
  111 
  112   The bind tree code had two bugs, for one the "shortcut"
  113   key-bindings would all have been created in the "base" context,
  114   not where they really belong ("default" and "compose").
  115   And due to false list relinking shadowed key bindings did not
  116   work.  (Koichi Murase)
  117 
  118   When used with 3x*verbose*#? (aka -v#85) the bind tree
  119   is now dumped when it has been build (once used first, and after
  120   modifications.  Putting all this together, a resource file
  121 
  122     cat >/tmp/t.sh <<'__EOT'
  123     MAILX=s-nail
  124     cat >/tmp/t.rc <<'_EOT'
  125     set line-editor-no-defaults
  126     bind base $'\n' mle-commit
  127     bind base $'\c?' mle-del-bwd
  128     bind base $'\cT' echo one
  129     bind base $'\cT',$'\cT' echo two
  130     bind base abc echo 0
  131     bind base ab,c echo 1
  132     bind base abc,d echo 2
  133     bind base ac,d echo 3
  134     bind base a,b,c echo 4
  135     bind base a,b,c,d echo 5
  136     bind base a,b,cc,d echo 6
  137     set quiet noheader
  138     set bind-inter-key-timeout=2500
  139     set bind-inter-byte-timeout=250
  140     bind base a,b,c,d # now works, too!
  141     _EOT
  142     MAILRC=/tmp/t.rc $MAILX -R:u -Y 'set verbose=3'
  143     rm -f /tmp/t.sh /tmp/t.rc
  144     __EOT
  145     sh /tmp/t.sh
  146 
  147   will now do the expected.
  148 
  149 - `Reply'#256: no longer honours *recipients-in-cc*#526
  150 
  151 ChangeLog
  152 ^^^^^^^^^
  153 
  154 - Manual: after review i think the following sections are of
  155   acceptable quality: "On terminal control and line editor"#17,
  156   "Coloured display"#18.
  157 
  158 - *verbose*#605 is no longer a boolean but can be assigned
  159   a numeric value.  But "set verbose verbose verbose"
  160   also still works.
  161 
  162 - New variable *reply-to-swap-in*#534 tries to work around
  163   the DKIM/DMARC "HUMAN via LIST <LIST@ADDRESS>" ugliness
  164   when `reply'#258ing (or `Reply'#256ing) to such a message.
  165   It tries to move the HUMAN into responsibility.
  166 
  167 git(1) shortlog: Steffen Nurpmeso (65)
  168 
  169 v14.9.16 ("Message of Winter, your hopes shall be crushed"), 2019-12-29
  170 -----------------------------------------------------------------------
  171 
  172 At the end of the v14.9 series we finally deliver the RFC 1524
  173 mailcap support, as well as other improvements, and new features.
  174 Many bugfixes and tests arrive.
  175 A whole bunch of things are backward-incompatible, but i would
  176 assume that most use cases are not at all affected.
  177 
  178 Credits, in order of commit appearance: Brian Evans,
  179 Paride Legovini, Yasuhiro KIMURA, John P. Linderman, Leo,
  180 Martin Neitzel, Alexander Harm, Ken Hornstein, Martynas Bendorius,
  181 Russell Bell, Goesta Smekal, Kevin McCarthy, Ralph Keller,
  182 Ralph Corderoy, Viktor Szépe, Jelle van der Waa, Arnout Engelen,
  183 Stuart Henderson, elo, and Benjamin A. Wong.
  184 
  185 We welcome Yasuhiro KIMURA, John P. Linderman, Leo,
  186 Martynas Bendorius, Goesta Smekal, Jelle van der Waa,
  187 Arnout Engelen, elo, and Benjamin A. Wong in THANKS.
  188 
  189 NOTES, ChangeLog (packager-affine)
  190 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  191 
  192 - Renamed *expandaddr*#415 namehostex to nametoaddr.  (Better.)
  193 
  194 - *tls-features*#593 string now starts and ends with comma.
  195   (Just like *features*#418 does.)
  196 
  197 - -L#71 longname --header-search renamed to --search.
  198 
  199 - Config option OPT_FORCED_STACKPROT=xy now
  200   OPT_AUTOCC_STACKPROT=y, and now auto-enabled for OPT_AUTOCC=y
  201   (even though i hate that, protectors in shipout code..).
  202 
  203 - `account'#140 return value now matters, and can thus be used to
  204   abort account switching.
  205 
  206 - -:#55 has new "x" mode, which executes the directives of the
  207   compiled-in resource file.  (The template content is now
  208   compiled in too, so the real file does not need to be loaded.)
  209   (Brian Evans)
  210 
  211 - Quote etc. injections (*quote-inject-head*#523) now always
  212   happen (if set), regardless of *quote*#518.
  213 
  214 - We now truly honour POSIX command abbreviations (order).
  215   As part of that i finally implemented a simple command lookup
  216   speedup, these are now almost alphabetical (unless abbreviations
  217   prevent it).
  218 
  219 - Added *mta-bcc-ok*#480.  Bummer.  It seems exim and courier
  220   do not remove Bcc: headers as required by standards unless
  221   invoked with a special command line argument.  So we now
  222   do not pass Bcc: headers to file-based MTAs unless this
  223   variable is set explicitly.  (Kevin McCarthy).
  224 
  225 - "COMMAND ESCAPES"#28 now have a $ command modifier, which
  226   causes a shell-style `eval'#191uation before the command
  227   escape is executed.
  228 
  229   And so `~<'#314, `~R'#337 and `~r'#338 no longer expand $VAR
  230   expressions by themselves.
  231 
  232   And so -a#57, `~@'#317, and "attachment insert" of `~^'#319
  233   and `digmsg'#174 only perform ~/-style expansions.
  234 
  235 - `~^'#319: use shell-style argument expansion.
  236   We handled `~^' and `digmsg' differently, but furthermore
  237   turned the shell-parsed data of the latter into whitespace
  238   separated data, which made it impossible to, for example,
  239   use attachments with whitespace in their names.
  240 
  241   This increases the interaction protocol version number of
  242   *on-compose-splice*#490 from "0 0 1" to "0 0 2"!
  243   Because, we do also quote the output, since using `read'
  244   (or read(1)) causes *ifs*#443 ($IFS) normalization.
  245   All that could be done would be (for ourselves):
  246 
  247       define x {
  248         set ifs=
  249         read vany
  250         unset ifs
  251         vpospar set $vany
  252         # And now assign the desired real fields
  253         set real-var1=$1 real-var2=$2  ...
  254       }
  255 
  256   Sick!  So instead introduce a `readsh'#251 command which works
  257   like `read'#250 but splits fields at shell token boundaries,
  258   for example from within *on-compose-splice*#490:
  259 
  260     echo '~^h s subject'; read stat name; readsh sub; read i
  261 
  262   There would be better examples.  (Ralph Keller)
  263 
  264 - `~F'#326, `~f'#327, `~M'#332, `~m'#333, `~U'#341 and `~u'#342 now
  265   honour *forward-inject-head*#428 and *forward-inject-tail*#429.
  266 
  267 - New option OPT_MAILCAP, by default enabled.
  268   Disable at runtime via *mailcap-disable*#458, all documented
  269   in "The Mailcap files"#36.
  270 
  271 - OpenBSD: really auto-find number of processors in test script.
  272   Work around fflush(3) not adapting POSIX behaviour.
  273 
  274 - Manual: after review i think the following sections are of
  275   acceptable quality: "Encrypted network communication"#13,
  276   "A starter"#6, "On URL syntax and credential lookup"#12,
  277   "The Mailcap files"#36, and
  278   "But, how about XOAUTH2 / OAUTHBEARER?"#45.
  279 
  280   The latter is actually (Stuart Henderson, Benjamin A. Wong),
  281   and now provides a copy+paste example of how to keep a
  282   OAUTHBEARER token up-to-date with S-nail (with some care).
  283   Yes, it is terrible.
  284 
  285 - Default .rc file: keeps Sender: by default (Ken Hornstein),
  286   sets *followup-to-honour*#425 and *reply-to-honour*#533,
  287   and gives more *history-gabby*#437ness.
  288 
  289 ChangeLog
  290 ^^^^^^^^^
  291 
  292 - Add `Lfollowup'#218.  (Russell Bell)
  293 
  294 - `?'#138: prefix \ (quoted!) to command to avoid
  295   `commandalias'#162  matching:
  296 
  297   ? ? S
  298   S -> spamspam: Teach the spam detector that <msglist> is spam
  299   ? ? \\S
  300   S (Save): Like `save', but derive filename from first sender
  301 
  302 - `history'#212 has new "delete NUMBER" subcommand.
  303 
  304 - Add *forward-add-cc*#426 and *quote-add-cc*#519 to Cc: the
  305   originator of a forwarded or quoted message, respectively.
  306   (Goesta Smekal)
  307 
  308 - *history-gabby*#437 now has a value to allow for more.
  309   This changes second argument of *on-history-addition*#492
  310   from boolean to context string.
  311 
  312 - New command `mtaaliases'#235.  We no longer automatically
  313   update the *mta-aliases*#475 cache.  (Maybe much later we will
  314   have a path_monitor or something, until then, not.)
  315 
  316   And `netrc'#236 "load" is now indeed "clear" + "load".
  317 
  318   `netrc' also gained a "lookup" subcommand.  (Ralph Corderoy)
  319 
  320 - *headline*#433 %L format will announce possibility that
  321   a message could be a list.
  322 
  323 - `chdir'#157, `rename'#255 and `remove'#254 use shell-style
  324   argument quoting.
  325 
  326 - `tls'#288 gained "certchain" and "certificate" subcommands.
  327 
  328 - `folder'#200 can open RFC 5322 messages via eml:// protocol,
  329   as in "folder eml:///tmp/msg.eml".  Yet primitive and only
  330   read-only.  (Viktor Szépe)
  331 
  332 - `vexpr'#302 gained "date-utc", "date-stamp-utc" and "epoch"
  333   subcommands.  (Benjamin A. Wong)
  334 
  335 - *on-main-loop-tick*#493 now also happens for commands passed
  336   via -Y#87.
  337 
  338 git(1) shortlog: Yasuhiro KIMURA (1), Steffen Nurpmeso (263)
  339 
  340 v14.9.15 ("Tit family in the trees"), 2019-08-18
  341 ------------------------------------------------
  342 
  343 Plugging a bug regarding copying data out of invalid MBOX mail
  344 databases which is present in all BSD Mails and in Unix V10 mail,
  345 and bringing in some tweaks, this update hopefully really marks
  346 the end of the v14.9.* series.
  347 
  348 After more than four and a half years i again have a VM testbed,
  349 with an increasing number of VM combinations.  (Yet still too few,
  350 but nonetheless, a dramatical improvement.)  This includes
  351 a GSS-API testbed, with an ArchLinux server and Linux and FreeBSD
  352 clients (do not ask why no additional FreeBSD server, i want to
  353 use binary packages).  This brought GSSAPI tweaks.
  354 
  355 Credits, in order of commit appearance: Ralph Corderoy,
  356 Chet Ramey, Robert Elz, Jilles Tjoelker, Steve Izma, Viktor Szépe,
  357 and Jean-Marc Pigeon.
  358 
  359 Very special thanks go to Tarqi Kazan and Ivan Vučica, who tested
  360 GSS-API in the past until it worked (again), testing against my
  361 blind flight patches!  Thank you very much, guys!
  362 
  363 We welcome Chet Ramey, Jilles Tjoelker and Steve Izma in THANKS.
  364 
  365 NOTES, ChangeLog (packager-affine)
  366 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  367 
  368 - *features*#418 and *tls-features*#593 are now prefixed with
  369   a comma ",", not with the number sign "#" (which could
  370   increasingly `eval'#191uate to a comment).
  371 
  372 - We now support parallelized tests.  It takes a numeric job
  373   number out of $MAKEFLAGS, or tries to fetch the number of
  374   processors otherwise (really!).  To go singleprocessor
  375   "$ make testnj" has to be called explicitly.
  376   With or without, we will terminate tests which take too long.
  377 
  378   This is truly tremendous, on the unstable9s machine of the
  379   OpenCSW.org cluster for example we now need 24 seconds instead
  380   of by far more than 300.  What a release!
  381 
  382   With help of (Chet Ramey, Robert Elz, Jilles Tjoelker and
  383   Steve Izma)
  384 
  385 - EXTERNAL authentication is truly a mess.  It has been fixed for
  386   POP3, where it was broken on our side.  But it seems the
  387   internet does not like that, or cannot (pass user credentials
  388   from a certificate gracefully to the authenticator).
  389   Anyway.  I have introduced EXTERNANON in addition for all of
  390   IMAP, POP3 and SMTP.  This could now result in a usable
  391   combination, regardless of what server(s) are contacted.
  392 
  393 - The hook *on-account-cleanup*#486 will now be called even
  394   upon program exit (i.e., implicitly leaving the account).
  395 
  396 ChangeLog
  397 ^^^^^^^^^
  398 
  399 - New *followup-to-add-cc*#424 will place the user in the Cc:
  400   list if it will place her in the Mail-Followup-To:.
  401 
  402 - New hook *on-program-exit*#494.
  403 
  404 - *pop3-auth*#508=gssapi is now supported.
  405   For IMAP, SASL-IR will be used for GSSAPI if possible (saving
  406   a packet round-trip).
  407 
  408 - *expandaddr*#415 has the new keyword "namehostex".
  409   If set, plain name addressees, like "To: steffen", will be
  410   expanded to NAME@HOSTNAME (where the latter could be
  411   *hostname*#441) if NAME is a valid user on the current host.
  412   (Viktor Szépe, Jean-Marc Pigeon)
  413 
  414 git(1) shortlog: Steffen Nurpmeso (45)
  415 
  416 v14.9.14 ("Great tit passed moult"), 2019-07-27
  417 -----------------------------------------------
  418 
  419 This is an unwanted and unplanned but unfortunately necessary
  420 bugfix release.  I hope it marks the end of the v14.9.* series.
  421 
  422 I presume you would be surprised if it would not also bring some
  423 features, this time mostly support of MTA-style aliases as
  424 inquired by Jean-Marc Pigeon, some authentication work (XOAUTH2/
  425 OAUTHBEARER support), and as usual development to the last minute.
  426 
  427 It fixes IMAP GSSAPI authentication, thanks to Ivan Vučica for
  428 reporting and testing this issue (Debian #930691; still have no
  429 testbed, but will soon!), and imap-delim, which i broke in July
  430 2017, thanks to Ralph Keller for repetitive reporting.
  431 
  432 For OpenBSD and SunOS 5.9 this release fixes long standing (must
  433 be) race conditions regarding child processes and their I/O setup.
  434 
  435   Never seen before, but my new box (i stepped a decade of
  436   hardware improvements, finally) rather regulary has shown them
  437   when running the test suite.  (On the OpenCSW cluster my speed
  438   varies, but i had a very good day and seen them there once.)
  439 
  440   This (finally) caused the complete rewrite of the child process
  441   (and termios) handling that i (had to) mention in communication
  442   with Gavin Troy already back in, i do not know -- 2013?  (Still
  443   not event loop based, but near getting good feelings there.)
  444 
  445   Funnily the problem (child descriptors were closed by the parent
  446   before the fork(2)ed childs had the opportunity to dup(2)licate
  447   their file descriptors) reminded me of a message of the german
  448   computer magazine c't, maybe around 2001/2002, when OpenBSD
  449   improved their fork(2) performance in a day or two after having
  450   appeared declassified in a comparison with other OSes.
  451   (Of course it was nothing but our own fault to not synchronize on
  452   the child, but blindly assuming that a fork(2) child gets the
  453   opportunity to run immediately.)
  454 
  455   Dear Predrag: would it now be possible for you to upgrade from
  456   v14.8.12?  I really would like to know!
  457 
  458 Credits, in order of commit appearance: Martin Lucina,
  459 Viktor Szépe, Alexander Harm, Anders Magnusson, Thomas Haigh,
  460 Martin T, Ivan Vučica, Nicholas Marriott, Alexander Harm,
  461 Steven Penny, Jean-Marc Pigeon, Martin Neitzel, Paul Vojta,
  462 Russell Bell, Paride Legovini and Ralph Keller.
  463 
  464 A special credit to Coverity.com once again, it found bugs!
  465 (<https://scan.coverity.com/projects/s-nail>, project number 444.)
  466 
  467 We welcome Martin Lucina, Anders Magnusson, Thomas Haigh,
  468 Martin T, Ivan Vučica, Nicholas Marriott, Steven Penny and
  469 Ralph Keller in THANKS.
  470 
  471 NOTES, ChangeLog (packager-affine)
  472 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  473 
  474 - The (very backward) Debian mawk is now supported directly.
  475   (Martin Lucina, Viktor Szépe)
  476 
  477 - GCC (8.3.0) -Os inlining bug (wmt) is worked around.
  478   And GNU awk 5 warnings have been fixed (before CRUX).
  479 
  480 - "|PIPE RECEIVER" errors seen on SunOS 5.9 and OpenBSD have first
  481   been fixed, and then caused a major rewrite of the child and
  482   termios handling for a rather "real" fix.
  483 
  484   The latter has the side effect that *pipe-TYPE/SUBTYPE*#500
  485   handlers will now have their standard output go to /dev/null.
  486 
  487 - IMAP GSSAPI authentication should work again.  (Ivan Vučica)
  488 
  489 - -C#60 testable can be used more than two times.
  490 
  491 - The "grappa" mode of mk/make-release.sh can now be used
  492   "everywhere".  See INSTALL on interest.  (It now gracefully
  493   fails if s-nail is not installed: we need that for hashing.)
  494 
  495 - New option OPT_MTA, by default enabled.
  496   Set *mta-aliases*#475 to a valid path in aliases(5) syntax,
  497   and we will expand them.  All Postfix directives but :include:
  498   are supported.  Only clear text files are supported, no DBs.
  499   (Jean-Marc Pigeon)
  500 
  501 - `~^'#319 will now verify *expandaddr*#415 right away in
  502   "~^ head ins to|cc|bcc", as is already done for `~t'#340,
  503   `~c'#323 and `~b'#322.
  504 
  505 - OPT_AGENT and OPT_SPAM_SPAMD are gone; they were obsoleted on
  506   2017-07-16, and it is not expected to become noticed.
  507 
  508 - *sendwait*#542 is now initially set, and it gained an optional
  509   value, a comma-separated list of case-insensitive strings naming
  510   specific subsystems for which synchronousness shall be ensured
  511   (only).  Possible values are "mta" for *mta*#474 delivery,
  512   and "pcc" for command-pipe receivers.
  513 
  514   P.S.: you can get a list of all initial values plus via
  515 
  516     $ s-nail -:/ -v -Xset -Xx
  517 
  518 - Colours may now happen even in quickrun mode (-e#65,
  519   -H#68, -L#71)!  I thought it is ok nonetheless, because
  520   we i think always documented to enwrap `colour'#160 setting
  521   in an according `if'#214, as in
  522 
  523     \if terminal && [ "$features" =% +colour ]
  524       \colo iso view-header fg=red
  525     \end
  526 
  527 - OPT_SOCKETS has been renamed to OPT_NET.
  528 
  529 - -Y#87 is now well defined under all conditions, with tests:
  530 
  531     The commands will be evaluated successively in the given
  532     order, and as if given on the program's standard input --
  533     before interactive prompting begins in interactive mode,
  534     after standard input has been consumed otherwise.
  535 
  536 - XOAUTH2 / OAUTHBEARER (OAuth 2.0 bearer token, RFC 6750)
  537   autentication is now supported for all protocols.
  538   New FAQ entry "But, how about XOAUTH2 / OAUTHBEARER?"#45
  539   For driving the necessary external update tool a new
  540   *on-main-loop-tick*#493 hook has been introduced.
  541   (I am thinking about adding support for an optional built-in
  542   token refresh.)
  543 
  544 - Support for the EXTERNAL authentication method has been
  545   introduced.  This is UNTESTED, though.  (I am still in the
  546   process of re-setting up my VM test environment.)
  547   We do not verify presence of a client certificate etc., but only
  548   that a TLS secured channel is active, when using this method.
  549   (We now *verbose*#605 log the used TLS version and cipher, too.)
  550 
  551 - *imap-delim*#665 works again as advertised!  This was broken in
  552   [1b9897a9] ((BWDIC!) Fix *imap-delim* behaviour.., 2017-07-01),
  553   and i think i was pretty much irritated by then.  Sorry!!
  554   While here, take *imap-delim* into account for `imapcodec'#661.
  555   (Ralph Keller).
  556 
  557 ChangeLog
  558 ^^^^^^^^^
  559 
  560 - Manual section "HISTORY"#51 improved a bit.  (Thomas Haigh)
  561 
  562 - New variable *line-editor-cpl-word-breaks*#452 (yet a bit
  563   restricted).
  564 
  565 - MLE: add mle-raise-{int,quit,tstp} functions.  Ie., raising
  566   those signals via ^C and ^Z is no longer hard-wired (in the
  567   MLE), but can be reassigned.  (Nicholas Marriott)
  568 
  569 - The makefiles no longer contain any awk code, that all has been
  570   separated into files under mk/.  (Alexander Harm)
  571 
  572 - We now have "test" *mta*#474, which dumps to standard output
  573   or optionally to a file, and honours *mbox-fcc-and-pcc*#461:
  574 
  575      $ echo text | s-nail -:/ -Smta=test -s ubject user@exam.ple
  576      $ </dev/null s-nail -:/ -Smta=test://./xy -s ub user@exam.ple
  577 
  578 - -##89 now acts as if variables were set via -S#79, rather than
  579   as via `set'#268, meaning that they are "frozen" for a while.
  580 
  581 - We have a new `colour'#160 mapping, "mle-error".
  582   It applies to error note in *prompt*#515, as well as to any
  583   other error logged to the terminal.  The latter is temporary
  584   until we gain a more fine grained logging facility, which then--
  585   likely--introduces an entire mapping family.
  586 
  587   Btw., for messages not explicitly generated (by, eg.,
  588   `echoerr'#182), we now also avoid writing adjacent duplicates,
  589   but rather do a syslog-like thing.  That error generator has
  590   been rewritten a bit, now more expensive, but looks right.
  591   (Martin Neitzel, Paul Vojta)
  592 
  593 - The value of $COLUMNS#613 seen in child processes is now the
  594   real one, rather than the one we use (not necessarily identical,
  595   dependent on termcap/terminfo/terminal support).
  596 
  597 - There are now three -v#85 / *verbose*#605 levels, not two.
  598 
  599 - The HTML filter knows more HTML 4.0 entities.  (Russell Bell)
  600 
  601 git(1) shortlog: Steffen Nurpmeso (161)
  602 
  603 v14.9.13 ("Blue tit's spiral marriage swoop"), 2019-03-08
  604 ---------------------------------------------------------
  605 
  606 Not so much happened as i was busy with other things in the second
  607 half of 2018, and could not find a grip: at the source level some
  608 efforts to turn this to mailx have taken place, including early
  609 work on a code abstraction that will be shared with my (g)roff
  610 clone: it was tiring to come back to old ideas that i have already
  611 implemented multiple times and in different languages.  To mention
  612 it for the curious.
  613 
  614 So: this is a bugfix release, in fact it fixes a tremendous amount
  615 thereof, [master] was ahead by 62 commits from [v14.9.11] (not all
  616 bugs though, hrmhrm), but also with, well, a few new features, and
  617 of course, development to the last minute. ^_^
  618 
  619   v14.9.13 replaces v14.9.12 from yesterday which would pick up an
  620   ISO C 2011 statement that is impracticable, thanks to Johannes
  621   Schöpfer and Jürgen Daubert for reporting this (i should test
  622   without OPT_AUTOCC more often.)  I have also fixed -T to match
  623   NEWS (manual and code did the opposite).
  624   v14.9.12 balls will be removed from the server.
  625 
  626 Credits, in order of commit appearance: Jörg Schilling,
  627 Paride Legovini, Olav Mørkrid, Ralph Corderoy, Rich Felker,
  628 Predrag Punosevac, Russell Bell, Dirk-Wilhelm Peters,
  629 Jean-Marc Pigeon, Warren Toomey, Cág, Martin Neitzel,
  630 Dr. Werner Fink, Dr. Matthias St. Pierre, Kurt Roeckx, Mike Sharov,
  631 Joan and Johannes Schöpfer.
  632 
  633 Very special thanks go to Jean-Marc Pigeon of OSUKISS Linux, who
  634 provided me access to a VZGOT container on the most "beefy"
  635 machine i have ever had access to!  It is ever so astonishing to
  636 have work done in 90 seconds on this supercomputer which requires
  637 an hour or more here.  Thank you, Jean-Marc.
  638 
  639 And thanks to Kimura-san not only my web server became accessible
  640 via TLSv1.3 (out-of-line), but we also have landed on FreeBSD!
  641 Kimura-san is the one who installed a sympathic cron job that
  642 reminds me of cherry petals wafting in a warm spring breeze --
  643 in winter!
  644 
  645 We welcome Russell Bell, Jean-Marc Pigeon, Warren Toomey,
  646 Dr. Matthias St. Pierre, Kurt Roeckx, Mike Sharov, Joan and
  647 Johannes Schöpfer in THANKS.
  648 
  649 NOTES, ChangeLog (packager-affine)
  650 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  651 
  652 - Packagers could follow stable/* via git(1) and the "grappa" mode
  653   of the release script (see README); perl(1) is required to be
  654   totally en par with an official release.  (Paride Legovini)
  655 
  656     $ git checkout stable/stable
  657     $ sh mk/make-release.sh grappa mybranch
  658     ...
  659     Program version is [.], packager release addition shall be: xy
  660     Is s-nail <v[.]-xy> correct? [y/n] y
  661     Switched to branch 'mybranch'
  662     $ git commit -S -n -m 'My release [.]-xy'
  663 
  664 - $MAKEJOBS vanished, just use -j or whatever your make(1) supports.
  665   Luckily the tested make(1)s can be persuaded to dig each others
  666   .WAIT / .WAIT: / .NOTPARALLEL: targets.. and do the right thing.
  667 
  668 - $OBJDIR support added, i use it for building / testing on tmpfs.
  669   It works in conjunction with make-emerge.sh, too, thus
  670   out-of-tree out-of-tree is possible (more or less; see INSTALL).
  671     $ make tangerine OBJDIR=/tmp/x/y/z
  672 
  673 - Option VAL_PRIVSEP_USER has been renamed to VAL_PS_DOTLOCK_USER
  674   (to reflect the new "deep tree" directory layout).
  675 
  676 - Option OPT_QUOTE_FOLD has been renamed to OPT_FILTER_QUOTE_FOLD
  677   (to reflect later code changes upwards compatibly).
  678 
  679 - New options OPT_CMD_VEXPR and OPT_CMD_CSOP, by default enabled.
  680   To include the commands `vexpr'#302 and the new `csop'#166,
  681   which now provides the byte string functions of the former.
  682   (Still available through `vexpr' until v15.)
  683   (I hope to be able to later provide a `usop' or `unisop' or so.)
  684 
  685 - *v15-compat*#604 can now have a value: if it is set, the
  686   `wysh'#130 command modifier which chooses shell quoting rules
  687   for some commands is implicit.
  688 
  689 - We have some (more) backward incompatible changes, though it is
  690   likely most users will not recognize the differences.
  691 
  692   o *headline*#433 format %T is obsolete, %L fits better.
  693 
  694   o `csop'#166 `hash' and `hash32' subcommands (formerly from
  695     `vexpr'#302) use a slightly changed hash algorithm.
  696     (Which results in an improved distribution for tested sets of
  697     words in power-of-two spaced dictionary.)
  698     These are affected by the change in the second next item, too.
  699 
  700   o Changed to use shell quoting rules for arguments:
  701 
  702     + `mimetype'#224 and `unmimetype'#225.
  703       This is affected by the change in the next item, too.
  704 
  705     + `shortcut'#272 and `unshortcut'#273.
  706 
  707     + `mlist'#227 and `unmlist'#228 as well as `mlsubscribe'#229
  708       and `unmlsubscribe'#230.
  709 
  710     + `alias'#143, too.
  711 
  712   o Changed (with legacy compat) the "@[i]" modifier prefix to
  713     a question-mark ?[case|..] suffix, as is known from URLs.
  714     We head towards direction URL syntax, now here too.
  715 
  716     + `if'#214 and `elif'#186.
  717       E.g., 'wysh if "abc" ==?case "ABC"' is true, as well as is
  718       'wysh if 0xFFFFFFFFFFFFFFFF -eq?saturated 36#1Y2P0IJ32E8E7'.
  719       "==?" and "-eq?" would have been sufficient, here.
  720       (No unsigned mode (yet) for `if'#214.)
  721 
  722       Yes, `if'#214 and `elif'#186 now support `wysh'#130, and
  723       see already expanded arguments, then.  No more "triggers".
  724       This finally makes it possible to write things like
  725         ? wysh if X;A;wysh elif Y;B;else;C;end
  726       Note 'else;C' not 'else C'.
  727 
  728       New operators: '-n "$VAR"' and '-z "$VAR"' work like in the
  729       shell, '-N varname' and '-Z varname' do not test the
  730       expansion but the existence of variables instead.
  731       Two argument forms require `wysh'#130.
  732 
  733     + `mimetype'#224 markers have changed likewise; this also
  734       affects *pipe-TYPE/SUBTYPE*#500 and *pipe-EXTENSION*#507 (with
  735       legacy compatibility and -v/-d obsoletion warnings)!
  736 
  737         ? mimetype ?t text/x-awk  awk
  738         ? wysh set pipe-application/pdf='?=&?\
  739             trap "rm -f \"${MAILX_FILENAME_TEMPORARY}\"" EXIT;\
  740             trap "trap \"\" INT QUIT TERM; exit 1" INT QUIT TERM;\
  741             mupdf "${MAILX_FILENAME_TEMPORARY}"'
  742 
  743     + `vexpr'#302 (and `csop'#166) modifiers changed likewise.
  744       The case-insensitive subcommands "ifind" and "iregex" have
  745       been obsoleted, just use the ?[case] modifier to the regular
  746       function instead.
  747 
  748       P.S.: Thanks to Rich Felker the `regex' subcommands now
  749       works as desired even with empty intermediate submatches.
  750 
  751   o Changed address parse mode for command line arguments plus.
  752     This modifies decade old tradition, but results in a more
  753     predictable behaviour i think.  Most people will possibly even
  754     be surprised to see the old behaviour:
  755 
  756       $ </dev/null s-nail-old -dsubject ' du , de <p@p> , pp , du '
  757       -> To: du, de <p@p>, pp
  758     new:
  759       $ </dev/null s-nail -dsubject ' du , de <p@p> , pp , du '
  760       -> To: "du , de , pp , du " <p@p>
  761 
  762     Of course anything but perfect, our address parser is very
  763     complicated yet far from being acceptable.  (Dr. Werner Fink)
  764     This affects:
  765 
  766     + -b#59, -c#61 and To: receivers, as above.
  767       (We also have a new -T#81 receiver multiplexer, which is
  768       configurable in this regard, please see below for more.)
  769 
  770     + The -r#78 address.
  771       This saw more changes: the content is no longer evaluated
  772       via shell expression parser (when *v15-compat* is set)..
  773       unless explicitly requested via the *expandaddr*#415
  774       flag "shquote".
  775 
  776     + *sender*#541 variable.
  777 
  778     + `addrcodec'#142 command, likewise; old:
  779         ? addrc e du , e <w@d> , d
  780         du , e <w@d> , d
  781         #?1!22/INVAL
  782       new:
  783         ? addrc e du , e <w@d> , d
  784         "du , e , d" <w@d>
  785         #?0!0/NONE
  786 
  787     + `digmsg'#174 and `~^'#319 now use this parse mode fix for
  788       headers which need a single receiver, which is backward
  789       compatible but now safer since it can be fooled less easily
  790       (to split into a list what should be a single address, as
  791       shown above for `addrcodec'#142).
  792 
  793       They now can also be forced to use that parse mode for To:,
  794       Cc:, Bcc: with a new question mark modifier "?single", here
  795       the word "single" is optional.
  796 
  797        ~^ header insert To?single: exa, <m@ple>
  798        ~^ header show to
  799 
  800 - By established rules and popular demand occurrances of '^From_'
  801   (see *mbox-rfc4155*#462) will be MBOXO quoted (prefixed with
  802   greater-than sign '>') instead of causing a non-destructive
  803   encoding like 'quoted-printable' to be chosen, unless context
  804   (e.g., message signing) requires otherwise.
  805   Only with *mime-encoding*#471=8bit.
  806 
  807 - We now support long "Options"#5 -- try --long-help.
  808 
  809 - Finally, it is possible to force sending out messages with the
  810   new *mime-force-sendout*#472 variable.  If this MUA has been
  811   compiled with iconv(3) support it can happen that sending
  812   otherwise valid text messages fails because of invalid bytes
  813   sequences according to the locale; setting this new variable
  814   will avoid this; use *mime-counter-evidence*#470 to view such
  815   messages nonetheless.  (Dr. Werner Fink)
  816 
  817 ChangeLog
  818 ^^^^^^^^^
  819 
  820 - `mimeview'#226 works again with binary formats.  (Russell Bell)
  821 
  822 - IMAP searches via IMAP without matches no longer report a single
  823   match.  (Dirk-Wilhelm Peters)
  824 
  825 - New -Y#87 aka --cmd= option to inject commands to be executed
  826   when startup is completed (as opposed to the earlier -X#86 aka
  827   --startup-cmd=).
  828   These commands appear as if the user had typed them in.
  829 
  830 - A new *on-history-addition*#492 can be used to filter what
  831   enters the `history'#212.
  832 
  833 - New "fcc" flag for *expandaddr*#415.  (Olav Mørkrid)
  834 
  835   And "domaincheck" will cause target domain comparison against
  836   entries in the new *expandaddr-domaincheck*#416.  (Olav Mørkrid)
  837 
  838 - New *mbox-fcc-and-pcc*#461 will write out file and pipe addresses
  839   as a plain RFC5322 message rather than an MBOX.  (Olav Mørkrid)
  840 
  841 - The `errors'#190 queue existance and size is announced via
  842   *^ERRQUEUE-EXISTS*#354 and *^ERRQUEUE-COUNT*#353.
  843   (Russell Bell, Martin Neitzel)
  844 
  845 - Our MBOX parser is now truly compliant to POSIX.
  846   (Dr. Werner Fink)
  847 
  848 - We follow symbolic links again when writing files.
  849   (Russell Bell)
  850 
  851 - *tls-rand-file*#596 is in fact now necessarily one of the
  852   optional *tls-features*#593.  (Mike Sharov)
  853 
  854 - New command line option -T#81 aka --target='FIELD: BODY'.
  855   FIELD can be To:, Cc:, Bcc: or Fcc:.
  856   The BODY is parsed as a list (just as if the given FIELD would
  857   be part of a template message fed in via -t#82), but the
  858   "?single" modifier suffix can be used to avoid this.
  859   (Dr. Werner Fink)
  860 
  861 git(1) shortlog: Steffen Nurpmeso (277 + 9)
  862 
  863 v14.9.11 ("Tit family enjoying a bath"), 2018-08-08
  864 ---------------------------------------------------
  865 
  866 A hot summer bugfix release, but it surely brings in some new
  867 features, like TLS fingerprinting and `digmsg' message access.
  868 
  869 An embarassing number of bugfixes have been seen, to fix IMAP UID
  870 handling on 32-bit hosts, UTF-8, `readall' with empty lines, rare
  871 endless iconv(3) loops, false qsort(3)ing of addressee lists,
  872 crashes due to false user shell quoting, acceptance of "0" port
  873 numbers, and more.  Most of these cases now have tests.
  874 
  875 Credits, in order of commit appearance: Paride Legovini,
  876 Andrew Gee, Olav Mørkrid, Kevin McCarthy, Michael Dressel,
  877 Jürgen Bruckner, Robert Elz, Rudolf Sykora, Doug McIlroy,
  878 Gavin Troy and Jörg Schilling.
  879 A special credit to Coverity.com again, it found a bug!
  880 (<https://scan.coverity.com/projects/s-nail>, project number 444.)
  881 
  882 We welcome Andrew Gee, Kevin McCarthy, Michael Dressel,
  883 Olav Mørkrid and Jürgen Bruckner in THANKS.
  884 
  885 NOTES, ChangeLog (packager-affine)
  886 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  887 
  888 - I have discovered that GnuPG can --export-secret-subkey so that
  889   the real/full private key is no longer needed to --sign, and the
  890   excerpts of the private one can have a different password, too.
  891   So i have created a new for-signing subkey: EEC8C2FF.
  892 
  893   Unfortunately it is not possible to verify new signatures with
  894   the old public key, an update is necessary.  For example via
  895     https://ftp.sdaoden.eu/steffen.asc
  896   or just update 1883A0DD via normal gpg(1).
  897 
  898 - Maildir support is now optional but default via OPT_MAILDIR.
  899   (Paride Legovini)
  900 
  901 - I hope for the last time there has been a change to SSL
  902   configuration: i have renamed all ssl* variables to tls*.
  903   The old ssl* ones still exist until v15, though, yet obsoletion
  904   warnings will be produced.
  905   This is so because i expect that in a not too distant future
  906   only the term TLS will be around.
  907 
  908   Also the term CipherList was falsely used, it should have been
  909   CipherString.
  910 
  911 - Most (if not all) commands which take a message list and a file
  912   target now use shell-style quoting.  (Before that say `copy'#165
  913   scanned backwards over "something possibly quoted", took that
  914   off, then treated the rest as a message list.  Now shell tokens
  915   are parsed starting at the front, the last is taken off, and
  916   anything before that is the message list.)  (Gavin Troy.  2013.)
  917 
  918 - `~^#'? "header show" now backward-incompatibly shows the address
  919   type in field 1, but since this _only_ applies to non-network
  920   addresses i made the change.
  921 
  922 - We will find ncurses on DragonFly BSD.
  923 
  924 - On Solaris tests no longer need GNU cksum(1): the Solaris cksum
  925   is different only for whitespace separators.  (Jörg Schilling)
  926 
  927 - All generated files reside in .obj/, and the tests run in there,
  928   too.  A "rm -f .obj" should suffice to clean anything up.
  929 
  930 ChangeLog
  931 ^^^^^^^^^
  932 
  933 - `~F'#326, `~f'#327, `~M'#332, `~m'#333, `~U'#341 and `~u'#342
  934   now default to the current message (the "dot").  (Andrew Gee)
  935 
  936 - *indentprefix*#448 handling has had its pitfalls when quoting
  937   messages.  (Andrew Gee)
  938 
  939 - -r#78 will again set *from*#430 even after -S#79 has been used
  940   to set *from*.  (Michael Dressel)
  941 
  942 - No longer process From: (*from*#430) content via `alternates'#145
  943   when Sender: (*sender*#541) is set.  (Michael Dressel)
  944 
  945 - Because the priority class of headers was not taken into
  946   account, it could happen that addressees in Cc: would remain but
  947   the same in To: were removed.  (Michael Dressel)
  948 
  949 - IMAP accounts for RFC 4551 (and 7162) and supports 64-bit
  950   UIDVALIDITY.
  951 
  952 - *spamfilter-rate-scanscore*#578 could crash if specification
  953   did not match program output.
  954 
  955 - Shims for TLSv1.3 support, e.g., for *tls-config-pairs*#590.
  956 
  957 - Obsoleted *dotlock-ignore-error*, added *dotlock-disable*#409.
  958   (Paride Legovini)
  959 
  960 - In compose-mode, removing the In-Reply-To: header breaks an old,
  961   and starts a new thread.  (Doug McIlroy)
  962 
  963 - Added new *forward-inject-tail*#429, *quote-inject-head*#523
  964   and *quote-inject-tail*#524 variables, and extended the meaning
  965   of *quote*#518.
  966 
  967   All of *{forward,quote}-inject-{head,tail}* now support
  968   a compose-mode specific set of formats (see
  969   *quote-inject-head*#523), for now a few only.
  970   (This adds meaning onto the content of *forward-inject-head*#428
  971   as introduced in v14.9.0.)
  972 
  973   The generated output honours *quote-fold*#522, which now takes
  974   an optional third argument in order to produce better output.
  975 
  976   While here, introduce the new command escape `~Q'#335 which
  977   performs full *quote*#518 cycles on the given message list.
  978 
  979 - Fcc: headers are now understood in -t#82 templates or when
  980   placed in compose mode (`~v'#343, *editalong*#410 etc.).
  981   Since each such header only takes one addressee, no quoting
  982   issues apply, the entire header body is the value.
  983 
  984 - `~|'#318 will pass the entire message including headers when
  985   used as "~||", e.g., prepend a file-carbon-copy message header:
  986 
  987     ~|| echo Fcc: /tmp/test; cat
  988 
  989 - New `tls'#288 multiplexer command.  Yet primitive and only
  990   supports a `fingerprint' subcommand.  Supports `vput'#129.
  991 
  992   The new *tls-fingerprint*#594 variable chain aids in adding
  993   support for connection verification without an installed CA
  994   certificate pool in conjunction with the new
  995   *tls-fingerprint-digest*#595 chain.
  996 
  997   Consequently *smime-sign-message-digest* has been renamed to
  998   *smime-sign-digest*#561 (old version will cease in v15).
  999   The latter now defaults to SHA512 if possible.
 1000 
 1001 - New MLE commands mle-go-screen-bwd and mle-go-screen-fwd to go
 1002   backward and forward one screenful.
 1003   And a new mle-clear-screen command.  (Todd C. Miller)
 1004 
 1005 - New *expandaddr*#415 setting "shquote" will evaluate addresses
 1006   as if specified within $'' shell-quotes for -b#59, -c#61,
 1007   and all direct command line receivers.  This allows for, e.g.,
 1008 
 1009     $ s-nail -Sexpandaddr=shquote '\$contact-mail'
 1010 
 1011 - *quote-as-attachment*#520 no longer needs to be set before
 1012   compose mode is entered in order to become honoured.
 1013 
 1014 - Even for -H#68 or -L#71 *folder-hook*#421s will now be called.
 1015   Possible sorting is also applied.
 1016 
 1017 - `='#137 now optionally supports message list arguments and the
 1018   `vput'#129 modifier in order to store the result list.
 1019 
 1020   The new `digmsg'#174 multiplexer adds some message access, just
 1021   like `~^'#319 does in compose mode.  In fact the set of commands
 1022   is shared, yet only in compose mode `digmsg' can change messages
 1023   or access attachments until v15, however.  For example,
 1024 
 1025   #?0[steffen@essex nail.git]$ cat > /tmp/z.rc <<'_EOT'
 1026   define one {
 1027      if [ "${#}" -gt 0 ]
 1028         digmsg create $1 - # no `read'/`readall' overlay but stdout
 1029         #digmsg $1 header list
 1030         digmsg $1 header show subject
 1031         digmsg remove $1
 1032         shift
 1033         \eval xcall one "$@"
 1034      \end
 1035   }
 1036   define all {
 1037      local set all # localize ("localopts yes" would do too)
 1038      vput = all *; echo all: $all; eval call one $all
 1039   }
 1040 _EOT
 1041   #?0[steffen@essex nail.git]$ MAILRC=/tmp/z.rc \
 1042     .obj/s-nail -:u -Snoheader -Squiet -Rf /tmp/z
 1043   ? call all
 1044   all: 1 2 3
 1045   212 Subject
 1046   Re: [S-mailx] FYI: after USB stick loss i have rotated keys, plus
 1047 
 1048   212 Subject
 1049   Re: Problem with page?
 1050 
 1051   212 Subject
 1052   Re: s-nail Source ...
 1053 
 1054   ? x
 1055 
 1056 git(1) shortlog: Steffen Nurpmeso (203)
 1057 
 1058 v14.9.10 ("(40th Mail anniversary) Blue tit"), 2018-03-25
 1059 ---------------------------------------------------------
 1060 
 1061 On this day in 1978 Kurt Shoens placed the following comment in
 1062 def.h (now it is in nail.h):
 1063 
 1064   /*
 1065    * Mail -- a mail program
 1066    *
 1067    * Author: Kurt Shoens (UCB) March 25, 1978
 1068    */
 1069 
 1070 v14.9.10 is mostly a stability and bugfix release.
 1071 It has seen a full test series including Coverity.com scans.
 1072 It fixes bugs i have introduced (also a double free in IMAP cache
 1073 that i introduced for v14.9.* series to address Coverity CID
 1074 1376978..).
 1075 
 1076 In the end i am saying thanks to Gunnar Ritter for the IMAP
 1077 module, and absolutely especially his really neat idea of an IMAP
 1078 cache including offline work queue.  (IMAP will nonetheless
 1079 temporarily go in v15, but these ideas will come back thereafter.)
 1080 I have gray hairs now.
 1081 
 1082 Credits, in order of commit appearance: William Yodlowsky,
 1083 Stuart Henderson, Jörg Schilling, Viktor SZÉPE, Rich Felker,
 1084 Ralph Corderoy and Philipp Gesang.
 1085 
 1086 A special credit to Coverity.com again.  Because:
 1087     tcc is 618496 bytes, pcc is 851968+24576 bytes,
 1088     but gcc is 73355264 bytes and clang is even
 1089     147406848 bytes, i wonder why the latter two never
 1090     said a word that would have addressed the pretty
 1091     obvious CID 1387053!
 1092     [Use of initialized value, the author.]
 1093 
 1094 We welcome Stuart Henderson and Philipp Gesang in THANKS.
 1095 
 1096 NOTES, ChangeLog (packager-affine)
 1097 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 1098 
 1099 - The balls are now build with umask 0022 not 0027.
 1100   (William Yodlowsky, Stuart Henderson)
 1101 
 1102 - One actual development of this version was the addition of
 1103   multiple choice VAL_ues, as documented in make.rc.
 1104   For now we have VAL_IDNA (for OPT_IDNA)
 1105     VAL_IDNA="idnkit idn2 idn"
 1106   and VAL_RANDOM (by itself)
 1107     VAL_RANDOM="arc4 ssl libgetrandom sysgetrandom urandom builtin"
 1108   (Stuart Henderson)
 1109 
 1110   In brief:
 1111     The value is interpreted as a whitespace separated list of
 1112     strings, like "idn2 idn idnkit", case is ignored, order is
 1113     important.
 1114     The special strings "all" and "any" as well as the empty value
 1115     are wildcard matches; if any entry in the list is a wildcard
 1116     match, the rest of the list is ignored.
 1117 
 1118     The special string "error" will abort configuration once its
 1119     list position is reached; this is only supported if
 1120     documented, and not with an accompanying OPT_ (which then
 1121     offers "require", as below).
 1122 
 1123   Since this VAL_RANDOM approach is so much better i have dropped
 1124   OPT_SSL_RANDOM and OPT_NOEXTRANDOM that were recently
 1125   introduced again.  They were c..p.
 1126 
 1127 - Support for idnkit 2.3 has been added.
 1128   Support for idnkit 1 (especially as idnkitlite) has been fixed.
 1129 
 1130 - For the first time this codebase should be able to handle
 1131   invalid MBOX mailboxes (produced by, e.g., dma(1)) gracefully.
 1132   I hope i have found all places (sic) where code has to be fixed.
 1133   E.g., "? copy * INVALID-MBOX" now works.
 1134   (Smalltalk already knew objects which know what they are doing
 1135   are for the better...  This is v15, then.)
 1136 
 1137 - P.S.: the two FreeBSD test failures are noted in INSTALL.
 1138 
 1139 ChangeLog
 1140 ^^^^^^^^^
 1141 
 1142 - *asksend*#368 will now really allow recomposing.
 1143 
 1144 - `help'#211 now supports recursive `commandalias'#162es, and
 1145   command self-recursion detection now works differently, it has
 1146   been false for something like
 1147 
 1148       commandalias x q; commandalias q echo au
 1149 
 1150   since q became expanded to `quit'#249 (alias expansion equals
 1151   new command word).  New behaviour: we allow equals once:
 1152 
 1153     commandalias q q; commandalias x q; x
 1154   ->
 1155     x -> q -> q -> quit
 1156 
 1157 - *editalong*#410 can have a value, say "set editalong=v" and it
 1158   will startup $VISUAL#638 not $EDITOR#615.
 1159 
 1160 - Path separators are now normalized, thus all places, including
 1161   MLE tab-expansion ("On terminal control and line editor"#17),
 1162   can expand something like "///t*////t*".
 1163 
 1164 - -E#64 flag will not be obsoleted.
 1165   -D#62 flag has been reintroduced (sets *disconnected*#662 right
 1166   away, was not reinstantiated with the rest of the IMAP support.)
 1167 
 1168 git(1) shortlog: Steffen Nurpmeso (71)
 1169 
 1170 v14.9.9 ("Marsh tit savours first spring sun, II") 2018-03-06
 1171 -------------------------------------------------------------
 1172 
 1173 A bugfix release.
 1174 I hope with this the fallout of the Christmas 2016 "address the
 1175 Dr. Problem workshop" has been fully resolved and thus MIME for
 1176 header address fields, even if iconv(3) is involved, been fully
 1177 restored!  We have even more tests for this now.
 1178 
 1179 The release v14.9.8 was broken on big endian machines.
 1180 I will remove the v14.9.8 balls from the server by the weekend.
 1181 Sorry for the inconvenience!
 1182 
 1183 Credits, in order of commit appearance: Slavko, Matej Mužila,
 1184 Rich Felker, Simon McVittie, Paride Legovini, Cág,
 1185 Predrag Punosevac.
 1186 
 1187 We welcome Slavko, Matej Mužila, Rich Felker and Simon McVittie in
 1188 THANKS.
 1189 
 1190 NOTES, ChangeLog (packager-affine)
 1191 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 1192 
 1193 - The v14.9.* series called *pipe-TYPE/SUBTYPE*#500 handlers in
 1194   display or quote mode with CR (carriage-return) bytes stripped
 1195   because of a missing output file comparison check, which broke
 1196   binary formats etc.  (Slavko)
 1197 
 1198 - We now have native support for Libidn2.  (Matej Mužila)
 1199 
 1200 - uname(1) is now hookable by setting the shell variable uname
 1201   when calling make ("uname=MY-UNAME make config" etc.).
 1202   (Simon McVittie)
 1203 
 1204   We no longer bake the kernel version into the binary, and
 1205   `version'#301 includes uname(2) output.
 1206   (Simon McVittie, Paride Legovini)
 1207 
 1208 - We now support a fallback P(seudo)R(andomNumber)G(enerator)
 1209   initialization even if getrandom(2)/getrandom(3) has been found
 1210   by the configuration, just like we do for "/dev/urandom" usage.
 1211   This does not affect systems with arc4random(3) or OpenSSL
 1212   random usage.  (David Čepelík, Simon McVittie)
 1213 
 1214   A new OPT_SSL_RANDOM make.rc variable, by default initialized to
 1215   the value of OPT_SSL.
 1216 
 1217 ChangeLog
 1218 ^^^^^^^^^
 1219 
 1220 - `~@'#317 list-edit behaviour in -##89 batch mode was broken.
 1221 
 1222 - Character set names will now undergo generic normalization,
 1223   including stripping of iconv(3) //SUFFIXes.
 1224 
 1225 git(1) shortlog: Steffen Nurpmeso (33)
 1226 
 1227 v14.9.7 ("Marsh tit patiently scraping bark") 2018-02-16
 1228 --------------------------------------------------------
 1229 
 1230 A maintenance release which fixes bugs and brings in features.
 1231 
 1232 Credits, in order of commit appearance: Alexander Harm,
 1233 Viktor SZÉPE, Paul Eggert, Joseph Bisch, Paride Legovini,
 1234 and Peter J. Holzer.
 1235 A special credit to the disappearing mutt(1) bug tracker.
 1236 And to Gmane.org for creating gmane.mail.s-mailx.general!
 1237 
 1238 Thanks Paride Legovini for becoming maintainer of the Debian port.
 1239 
 1240 We welcome Joseph Bisch, Paride Legovini, and Peter J. Holzer in
 1241 THANKS.
 1242 
 1243 NOTES, ChangeLog (packager-affine)
 1244 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 1245 
 1246 - The USB stick loss reported for v14.9.6 was fake news, so to
 1247   say, the stick exists and therefore the old key is not
 1248   compromised.
 1249 
 1250 - We are back at Gmane.org!
 1251     news.gmane.org/gmane.mail.s-mailx.general
 1252 
 1253 - `history'#212 has learned to be context-sensitive a bit, and
 1254   has two new subcommands, `load' and `save'.
 1255 
 1256     This is in parts backward incompatible because it needs a new
 1257     *history-file*#436 format; however, the old format can be loaded
 1258     yet compose-mode commands will not appear in compose mode no
 1259     more.  Iirc you can start with an old format then `save' to
 1260     the new, then replace the "d" in the first column with "c" for
 1261     compose-mode commands which should appear correctly.
 1262 
 1263 - Obsoletion warnings for variables now happen at `set'#268
 1264   time instead of when used.  Running once via -v#85 may
 1265   be beneficial.
 1266 
 1267 - The saturation modifier of `vexpr'#302 is henceforth a prefix,
 1268   the suffix version is obsolete (but still supported for a while).
 1269 
 1270 - A network address that contains no domain-, but only a valid local
 1271   user <name> in angle brackets will be automatically expanded to
 1272   a valid address when *hostname*#441 is set to a non-empty value;
 1273   setting it to the empty value instructs us that the used *mta*#474
 1274   (including builtin SMTP) will perform the necessary expansion.
 1275   (Viktor SZÉPE)
 1276 
 1277   Note that *hostname*#441 as well as *smtp-hostname*#564
 1278   will now undergo IDNA expansion if IDNA is supported.
 1279 
 1280   And *from*#430 and *sender*#541 are now verified at `set'#268
 1281   time, not when used.  (Viktor SZÉPE)
 1282 
 1283 - The commit message in [d503bd82] is wrong, apologies to
 1284   Paride Legovini.  The test(1) operator "-n" appeared in Seventh
 1285   Edition UNIX, not V8 as falsely claimed.
 1286 
 1287 ChangeLog
 1288 ^^^^^^^^^
 1289 
 1290 - Our `addrcodec'#142 parser chokes on lesser constructs.
 1291 
 1292 - Presence of command-line MTA arguments without *expandargv*#417
 1293   are now a hard error.  It was my fault that this was not the
 1294   default from the very start.  (Viktor SZÉPE)
 1295 
 1296 - Seen on the mutt bug tracker, we also still have had problems
 1297   with time settings that cross 32-bit boundaries.  As that is
 1298   in parts induced by the C standard, now implement those parts on
 1299   our own, and be super careful in general.  (Joseph Bisch)
 1300 
 1301 - The `~@'#317 command escape did not shell-unquote the user input
 1302   again and was thus a bit broken; message attachments also work
 1303   again.
 1304 
 1305 - Support custom headers from the command line via -C#60.
 1306   And *customhdr*#403 is verified upon `set'#268 time.
 1307 
 1308 - The simple builtin HTML viewer now supports <blockquote>
 1309   elements, which many web mailers, most notably gmail, use for
 1310   citation.  (Peter J. Holzer)
 1311 
 1312 git(1) shortlog: Paride Legovini (1), Steffen Nurpmeso (66)
 1313 
 1314 v14.9.6 ("Marsh tit abiding a snow storm"), 2017-12-05
 1315 ------------------------------------------------------
 1316 
 1317 A bugfix release which fixes four serious and three other bugs.
 1318 A few new features came in, too.
 1319 
 1320 Many thanks go to Ralph Corderoy who reported an issue that was
 1321 caused by a terrible, terrible word reversal that i managed to
 1322 produce in December 2016, and which caused the v14.9.x series to
 1323 not MIME encode (non-address) content of address header fields!
 1324 
 1325 Credits, in order of commit appearance: Thomas Dickey,
 1326 Andreas Baumann, Erich Eckner, Gaetan Bisson, Solar Designer, Cág,
 1327 Ivan Tham, Ralph Corderoy and Doug McIlroy.
 1328 
 1329 We welcome Andreas Baumann, Erich Eckner, Solar Designer and Cág
 1330 in THANKS.
 1331 
 1332 NOTES, ChangeLog (packager-affine)
 1333 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 1334 
 1335 - After USB stick loss the authors OpenPGP key has been switched to
 1336 
 1337     pub   4096R/1883A0DD 2017-11-30 [expires: 2027-11-28]
 1338           Key fingerprint =
 1339             EE19 E1C1 F2F7 054F 8D39  54D8 3089 64B5 1883 A0DD
 1340     uid                  Steffen Nurpmeso <steffen@sdaoden.eu>
 1341 
 1342 - $TMPDIR#636 no longer honoured for root runs.  (Solar Designer)
 1343 
 1344 - *mime-encoding*#471 defaults to quoted-printable again.  (Cág)
 1345 
 1346 - We _can_ MIME encode even header fields which contain addresses.
 1347   Thanks to Ralph Corderoy we now also _do_ so again!
 1348 
 1349 ChangeLog
 1350 ^^^^^^^^^
 1351 
 1352 - ***#336 now uses *ifs*#443 when splitting.
 1353 
 1354 - Freezing *ttycharset*#600 via -S#79 also survives using or
 1355   setting any of $LC_ALL#617, $LC_CTYPE#618 and $LANG#619 during
 1356   program startup.
 1357 
 1358 - New `local'#126 command modifier to localize changes.
 1359   Yet supported only for `set'#268, i.e., we have gained
 1360   macro-local variables.
 1361 
 1362 - `vexpr'#302 now supports a BASE#number notation for integers,
 1363   like 16#AFFE as an alternative to 0xAFFE.
 1364 
 1365   Hint: variable settings can most often use several bases, too,
 1366   e.g., i have "set mime-counter-evidence=0b1111".
 1367 
 1368 - Very simple form of *quote-chars*#521 to adjust our knowledge of
 1369   what actually is to be treated as a quote character.
 1370 
 1371 - *mime-counter-evidence*#470 deep inspection (bit four) has
 1372   been improved for the sole cases of quoting or displaying
 1373   a message.  So messages with less than 25% of control characters
 1374   and such will now be displayed (made printable).  This is yet
 1375   not configurable nor do we have a way to easily access a message
 1376   with more than that.  (Doug McIlroy)
 1377 
 1378 git(1) shortlog: Steffen Nurpmeso (44)
 1379 
 1380 v14.9.5 ("Marsh tit engaged with a peanut"), 2017-10-21
 1381 -------------------------------------------------------
 1382 
 1383 A bugfix release which fixes two bugs which were cast in stone.
 1384 A few compatibility improvements (AlpineLinux, Solaris).
 1385 And minor features.
 1386 
 1387 Apologies to Jörg Schilling, a git bug i think it was who caused
 1388 joining of changesets, losing a credit, and it had been pushed to
 1389 [master] before the problem was realized.
 1390 
 1391 Credits, in order of commit appearance: Jörg Schilling,
 1392 Doug McIlroy, Random832, Nick Stoughton and Ivan Tham.
 1393 
 1394 We welcome Nick Stoughton and Ivan Tham in THANKS.
 1395 
 1396 NOTES, ChangeLog (packager-affine)
 1397 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 1398 
 1399 - New OPT_USE_PKGSYS option can be disabled to not automatically
 1400   pick known package system paths (pkg(7), OpenCSW, schily).
 1401   (Jörg Schilling)
 1402 
 1403 ChangeLog
 1404 ^^^^^^^^^
 1405 
 1406 - The software indeed _never_ dealt with iconv(3) output character
 1407   set errors (as opposed to invalid input character set byte
 1408   sequences etc.) for the main message body!
 1409   And I have missed that when i tweaked our iconv layer a bit!
 1410 
 1411 - Fixed a race condition with sigsuspend(2) that i could only see
 1412   on OpenBSD.  config.h offers n_SIGSUSPEND_NOT_WAITPID, by the
 1413   way, which saves some systemcalls and did not run races, but
 1414   noone adjusts this file.
 1415 
 1416 - Message list specifications gained two new colon modifiers, one
 1417   can now "search :Ll" to find "Mailing lists"#10.
 1418   The *headline*#433 format %T now also uses L and l rather than
 1419   S and L accordingly.
 1420 
 1421   New `addrcodec'#142 subcommand `skinlist' acts like `skin'
 1422   but stores in *!*#348 *^ERR*#350-EXIST if the address is
 1423   one of the known "Mailing lists"#10.
 1424 
 1425 - `echo'#181 family now supports `vput'#129 and *!*#348 error
 1426   storage, offering some kind of printf(1) experience, almost.
 1427 
 1428 git(1) shortlog: Steffen Nurpmeso (35)
 1429 
 1430 v14.9.4 ("(5th anniversary) Marsh tit"), 2017-09-18
 1431 ---------------------------------------------------
 1432 
 1433 This is an update feature release but which also ships a furious
 1434 number of bug fixes, about six of which were pretty serious.  It
 1435 also applies overall trimming, and improves configuration time
 1436 compatibility on macOS.
 1437 
 1438 Thanks to Alexander Harm there is now a macOS Homebrew package.
 1439 
 1440 Credits, in order of commit appearance: Paul Vojta, Daniel Lublin,
 1441 Alexander Harm, Norman Ramsey, Viktor Szépe, Rich Salz,
 1442 David Čepelík, Ralph Corderoy, Stéphane Chazelas, Aharon Robbins,
 1443 Ken Hornstein.
 1444 
 1445 We welcome Daniel Lublin, Alexander Harm, David Čepelík and
 1446 Stéphane Chazelas in THANKS.
 1447 
 1448 NOTES, ChangeLog (packager-affine)
 1449 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 1450 
 1451 - Out-of-tree builds have become possible via the new
 1452   make-emerge.sh script:
 1453 
 1454   $ cd /tmp && mkdir build && cd build &&
 1455     ~/src/nail.git/make-emerge.sh &&
 1456     make tangerine DESTDIR=.ddir
 1457 
 1458   We now have a `citron' make target which is like `tangerine' but
 1459   does not run the tests.
 1460 
 1461 - Configuration with OPT_AUTOCC honours $CC=cc. (Norman Ramsey)
 1462 
 1463 - SSL/TLS configuration has been revamped (again) in order to
 1464   support new possibilities of OpenSSL (and LibreSSL) without
 1465   ending up and introducing more and more variables.
 1466 
 1467   Instead we now have *ssl-config-pairs*, a comma-separated
 1468   list of all options.  With e.g. OpenSSL 1.1.xx this will be
 1469   directly passed through to SSL_CONF_cmd(), so there _anything_
 1470   can be passed, otherwise we use a builtin parser to map.
 1471   The new *ssl-features*#? states what is supported.  E.g.:
 1472 
 1473     if [ "$ssl-features" =% +ctx-set-maxmin-proto ]
 1474       wysh set ssl-config-pairs='\
 1475           CipherList=TLSv1.2:!aNULL:!eNULL:@STRENGTH,\
 1476           Curves=P-521:P-384:P-256,\
 1477           MinProtocol=TLSv1.1'
 1478     else
 1479       wysh set ssl-config-pairs='\
 1480           CipherList=TLSv1.2:!aNULL:!eNULL:@STRENGTH,\
 1481           Curves=P-521:P-384:P-256,\
 1482           Protocol=-ALL\,+TLSv1.1 \, +TLSv1.2'
 1483     endif
 1484 
 1485   OpenSSL v1.1.xx also introduces an interesting and neat idea to
 1486   centralize SSL/TLS configuration of (all) programs in a single
 1487   file.  This can be driven via *ssl-config-file* and the new
 1488   *ssl-config-module* variables, several entries per program
 1489   are allowed, see *ssl-config-module* for an example.
 1490 
 1491   New manual section "Encrypted network communication"#13.
 1492 
 1493 - Variables set or unset via -S#79 are now frozen until program
 1494   startup is complete.
 1495 
 1496 ChangeLog
 1497 ^^^^^^^^^
 1498 
 1499 - Historical behaviour of *askcc*#366 / *askbcc*#367 has been
 1500   reintroduced.  (Norman Ramsey)
 1501 
 1502   A new *asksend*#368 variable will show a final header summary
 1503   and allows reentering compose mode.  Set by default.
 1504 
 1505   POSIX mirrors *ask* onto *asksub*#370, so dropped" the former.
 1506 
 1507 - `~^'#319 no longer normalizes header names to titlecase.
 1508 
 1509 - We no longer generate charset=binary MIME parameters.
 1510   This was introduced on 2013-01-02 and was i think owed to
 1511   file(1)s -i output as i failed to find any other reference.
 1512   (Normal Ramsey)
 1513 
 1514 - *mime-alternative-favour-rich*#469 now also works for handlers
 1515   installed via *pipe-TYPE/SUBTYPE*#500. (Viktor Szépe)
 1516 
 1517 - v14.9.* series did not generate In-Reply-To: headers!
 1518 
 1519 - `alias'#143 now supports high-bit bytes and semicolon.
 1520   Expect that at some later time the input must be valid according
 1521   to the locale, though. (Norman Ramsey)
 1522 
 1523 - Combinations of *record*#527 could crash because of an
 1524   unterminated variable function argument list. (Norman Ramsey)
 1525 
 1526 - New command `readall'#252 loads an entire file into a variable.
 1527 
 1528   *signature*#548 has been obsoleted.
 1529 
 1530 - `vexpr'#302 now supports negative arguments for the substring
 1531   subcommand and adds trim, trim-front and trim-end subcommands.
 1532 
 1533 - `!'#133 can be used in send mode.
 1534 
 1535 - `~A'#320, `~a'#321, `~I'#330 and `~i'#331 will henceforth expand
 1536   \t and \n only if *posix*#513 is set.
 1537   Please use `set'#268 instead (with `wysh'#130, until v15).
 1538 
 1539 - New "The mime.types files"#35 type marker: @q ("quiet").
 1540 
 1541 git(1) shortlog (edited): Steffen (Daode) Nurpmeso (90)
 1542 
 1543 v14.9.3 ("Crested tit nibbling sunflower seeds"), 2017-08-03
 1544 ------------------------------------------------------------
 1545 
 1546 This is a bugfix release but which ships some improvements, too.
 1547 It silently replaces both of v14.9.1 v14.9.2 from earlier this
 1548 week, which were broken or not entirely fixed.
 1549 
 1550 Credits, in order of commit appearance: Felix Fontein, Paul Vojta,
 1551 Ralph Corderoy, Christos Zoulas, Gavin Troy, Gaetan Bisson.
 1552 Thanks, Coverity.com.
 1553 
 1554 We welcome Christos Zoulas in THANKS.
 1555 
 1556 Apologies to Viktor Szépe for the false spelling of his name in
 1557 the v14.9.0 announcement.
 1558 And to Gaetan Bisson for not giving credit for [14fbce97]!
 1559 
 1560 NOTES, ChangeLog (packager-affine)
 1561 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 1562 
 1563 - fakeroot support was blindly taken from Debian and broken.
 1564   (Gavin Troy, Gaetan Bisson)
 1565 
 1566 - Base64 output was broken for cases which involved iconv(3).
 1567   The data is not lost, you can read it with S-nail v14.9.0 and
 1568   above, and save it somewhere.  I know of no other base64 decoder
 1569   which reads those things correctly, though.  We now have tests.
 1570   Along this i fixed an iconv(3) error which likely caused
 1571   stateful decoding (like, e.g., for ISO-2022-JP) to fail because
 1572   of an unnecessary reset of the iconv(3) state machine.
 1573   Thanks to Gaetan Bisson for mentioning this issue!
 1574 
 1575 ChangeLog
 1576 ^^^^^^^^^
 1577 
 1578 - In compose mode the MLE allows empty lines again.
 1579 
 1580 - We no longer require a writable $HOME#616.  Due to false code
 1581   flow (but but but: with correct comment) a non-writable $HOME
 1582   entry in /etc/password (i.e., from getpwuid(3)) would lead to
 1583   a crash.  (Felix Fontain; Ralph Corderoy)
 1584 
 1585 - Two faulty string operations slipped into the IMAP code,
 1586   resulting in a crash and a "is-same-host" test that would fail
 1587   for IMAPS connections like `save'#265 or `copy'#165 because of an
 1588   implicit IMAP protocol for the target of those operations (thus
 1589   IMAP != IMAPS).  (Paul Vojta)
 1590 
 1591 - The MLE tab-expansion will now automatically append a "/" if
 1592   there is only one possible expansion and that is a directory,
 1593   saving the user one <TAB>.  (Christos Zoulas)
 1594 
 1595   The shell expression parser had a bug regarding understood
 1596   metacharacters (;|&), which in turn could cause an infinite loop
 1597   in the MLE tab-expansion for, e.g., "move &9 +<TAB>", because
 1598   the "&" would never have been stepped over.
 1599 
 1600 - New `~I'#330 command escape is like `~i'#331 but does not append
 1601   a newline.
 1602 
 1603 - `localopts'#217 gained an optional second argument.
 1604   It is now possible to specify that any macro `call'#151ed
 1605   will have localopts enabled, and it is possible to fixate the
 1606   setting so that it cannot be reverted.
 1607 
 1608 - *@*#356 should now act completely compatible to the sh(1)ell,
 1609   thus obsoleting my hysteric warnings in the v14.9.0 announcement.
 1610 
 1611 - The `Lreply'#219, `reply'#258, `Reply'#256 series as well as
 1612   `mail'#221 now manage the error status *!*#348.
 1613   I.e., there are now errors like *^ERR*#350-DESTADDRREQ,
 1614   ^ERR-NODATA, ^ERR-PERM and similar.  It is not perfect yet,
 1615   because $DEAD#614 may have been written (with *save*#536) or not,
 1616   for example.
 1617 
 1618   `Lreply' and `reply' have been rewritten rather completely
 1619   indeed.  They join Reply-To: and Mail-Followup-To: dependent on
 1620   the context (i.e., *reply-to-honour*#533, *followup-to-honour*#425,
 1621   see "Mailing lists"#10 for the picture), and if they did, use
 1622   this list as the receivers exclusively.  It now honours
 1623   *recipients-in-cc*#526 even for such addressees.  (And now i wonder
 1624   whether i should have credited Paul Vojta for that.)
 1625 
 1626   Also `Lreply' would have crashed for mails with Reply-To: but
 1627   without *reply-to-honour*#533 set.  We now have a test.
 1628 
 1629   Note *replyto* is obsoleted in favour of *reply-to*#532.
 1630 
 1631 v14.9.0 ("Long-tailed tit"), 2017-07-16
 1632 ---------------------------------------
 1633 
 1634 This is a major feature release which took about ~22 months (24
 1635 less two) of development to complete, and which imposed massive
 1636 changes under the hood, but also quite a lot of user visible
 1637 changes, including some **backward incompatibilities**.
 1638 As usual, "s-nail -d" will show obsoletion warnings.
 1639 
 1640 We gain noticeable improvements regarding scriptability and its
 1641 reliability, but also for interactive use cases, especially
 1642 notable to users is our completely new M(ailx)L(ine)E(ditor) that
 1643 supports rather real tabulator expansion and program-mode-context-
 1644 sensitive key bindings.
 1645 
 1646 We now support macros with arguments, which can be `shift'ed,
 1647 a `return' status can be used, and a `vexpr' multiplexer offers
 1648 some arithmetic and string operations.  `commandalias'es are
 1649 recursive, further command modifier prefixes, like `ignerr', give
 1650 a hand that we otherwise could not offer.  In compose-mode the new
 1651 `~^' command escape allows some message and attachment access, and
 1652 can be used, e.g., to implement things like custom headers, and
 1653 has been especially designed for scripted access via the new
 1654 *on-compose-splice* and *on-compose-splice-shell* hooks.
 1655 
 1656 S-nail will move (more or less) backward-incompatibly to sh(1)ell
 1657 compatible argument quoting (documented in "COMMANDS"), and an
 1658 increasing number of commands do support this already: new ones
 1659 exclusively, some old ones have either been switched (like
 1660 `localopts'), others -- noticeably `set' -- can be switched to the
 1661 new syntax with a `wysh' command modifier prefix.  E.g.:
 1662 
 1663   ? define __xv {
 1664     # Be careful to choose sh(1)ell-style on _entire_ line!
 1665      localopts yes; wysh set verbose; ignerr eval "${@}"; return $?
 1666     }
 1667   ? commandalias call echo boo-boo
 1668   ? commandalias xv call __xv
 1669   ? xv list
 1670   ? commandalias xv '\'call __xv
 1671   ? xv list
 1672 
 1673 Calling the latter `xv' for `list' will give more detailed command
 1674 information, including which kind of argument is used.
 1675 
 1676 I have not managed to implement the three features i have started
 1677 this development cycle for, these are thus subject to further
 1678 development, just like wysh for message-list argument commands to
 1679 support, e.g., negation, wysh for `if' and consorts, the --
 1680 terminator to finally overcome the ridiculous requirement to quote
 1681 entire shell commands filenames for commands like `pipe.
 1682 And and and.
 1683 
 1684 Credits, in order of commit appearance: Antonio Radici,
 1685 Aharon Robbins, Mike Frysinger, Predrag Punosevac, Michael Convey,
 1686 Hariskar, Rudolf Sykora, Martin Neitzel, Gavin Troy,
 1687 Salvatore Bonaccorso, Todd C. Miller, Sergey Matveev, Robert Elz,
 1688 Mantas Mikulėnas, Respiranto, Jens Schleusener, Walter Alejandro
 1689 Iglesias, Ralph Corderoy, David Levine, Lyndon Nerenberg,
 1690 Thomas Dickey, Afan, Justin Ellingwood, Ingo Schwarze,
 1691 Viktor Szépe, Gaetan Bisson, Juan RP, William Yodlowsky,
 1692 Hilko Bengen, Matthew Dillon, Colin Watson, Donald Mugnai,
 1693 Stephen Isard, Jürgen Daubert, Sven Neuhaus, trondd, Ismael Bouya,
 1694 Felipe Gasper, Paul Eggert, Dr. Werner Fink, Ken Hornstein,
 1695 Noel Chiappa, Random832, Doug McIlroy, Baptiste Daroussin,
 1696 Riccardo Ductor, Pietro Cerutti, Jörg Schilling, rain1, Xin LI.
 1697 
 1698 We welcome Antonio Radici, Mike Frysinger, Predrag Punosevac,
 1699 Michael Convey, Rudolf Sykora, Todd C. Miller, Robert Elz,
 1700 Jens Schleusener, Walter Alejandro Iglesias, Thomas Dickey, Afan,
 1701 Justin Ellingwood, Viktor Szépe, Juan RP, Matthew Dillon,
 1702 Colin Watson, Donald Mugnai, Sven Neuhaus, Ismael Bouya,
 1703 Felipe Gasper, Paul Eggert, Dr. Werner Fink, Ken Hornstein,
 1704 Noel Chiappa, Random832, Doug McIlroy, Baptiste Daroussin,
 1705 Riccardo Ductor, Pietro Cerutti, Jörg Schilling, rain1, and
 1706 Xin LI in THANKS.
 1707 
 1708 Apologies: Sergey Matveev.
 1709 Members of the Roff community which await progress.
 1710 
 1711 NOTES, ChangeLog (packager-affine)
 1712 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 1713 
 1714 * This release brings some backward incompatibilities, outlined
 1715   in the following.  Most users will not be affected, and we have
 1716   added a lot of compatibility cruft, but that will vanish in v15.
 1717   "$ s-nail -v"!
 1718 
 1719 * The configuration and build system has changed.
 1720   Packagers have received updated package files.
 1721 
 1722   o Anything which was WANT_xy before is now OPT_xy, and
 1723     compiled-in paths and values, like PREFIX or PAGER, have
 1724     gained a VAL_ prefix (thus VAL_PREFIX and VAL_PAGER).
 1725 
 1726     This is _not_ true for non-persistent or environmental values,
 1727     e.g., DESTDIR, CC, etc., and also not for the overwritable
 1728     program variables during configuration, e.g., $awk.
 1729 
 1730   o SENDMAIL -> VAL_MTA, SENDMAIL_PROGNAME -> VAL_MTA_ARGV0,
 1731     MAILSPOOL -> VAL_MAIL.
 1732     And NAIL -> VAL_MAILX, though this is still a lie.
 1733 
 1734   o The make system now needs config..build..install or
 1735     all..install or tangerine (config..build..test..install).
 1736     Some constants which some experts may want to fine-tune have
 1737     been moved to config.h.  Usual adjustments+doc via make.rc.
 1738 
 1739   o The `build' phase can be parallelized by setting the $MAKEJOBS
 1740     environment variable, e.g., "make MAKEJOBS=-j4 build".
 1741     Note this variable is not tracked in the configuration.
 1742     (Gaetan Bisson)
 1743 
 1744   o Unless DESTDIR is set an uninstallation script will be
 1745     installed along with the rest (see INSTALL file for more).
 1746 
 1747   o Set the new OPT_CROSS_BUILD to avoid feature runtime tests,
 1748     only compile- and link-availability will be tested.  (Juan RP)
 1749 
 1750   o VERBOSE is implemented straight, but must be given at
 1751     configuration time in order to become honoured.
 1752     (William Yodlowsky)
 1753 
 1754   o ADDCFLAGS / ADDLDFLAGS -> EXTRA_CFLAGS / EXTRA_LDFLAGS.
 1755 
 1756   o The LD_LIBRARY_PATH etc. building processes will skip any path
 1757     which contains the string "fakeroot".  (Hilko Bengen)
 1758 
 1759   o We honour a set $SOURCE_DATE_EPOCH#634 environment variable to
 1760     an extend that allows reproducible tests, which is why the
 1761     repository gained a [test-out] branch with some expected plain
 1762     text outputs.  (reproducible-builds.org; Colin Watson)
 1763 
 1764     The new *log-prefix*#455 variable aids in improving the
 1765     reproducibility of error messages.
 1766 
 1767   o These are upward compatible changes.
 1768 
 1769 * "make OPENSSL_API_COMPAT=0x10100000 all" should work.
 1770 
 1771 * Internal and environment variables are now explicitly _defined_
 1772   and _tracked_ after variable handling has been rewritten
 1773   completely.  Notes:
 1774 
 1775   o This means that, e.g., "$ password=NOT_SECRET s-nail" will
 1776     **NOT** work no more, since *password*#498 is an internal
 1777     variable!
 1778 
 1779   o But if you do, e.g., "? set TMPDIR=~/tmp", then this will
 1780     also be reflected in the program environment (it is an
 1781     environment variable) and thus affect child processes.
 1782 
 1783   o Therefore we no longer have `setenv' and `unsetenv'.
 1784 
 1785   o To integrate any other environment variable transparently
 1786     into our variable management, the new command `environ'#189
 1787     needs to be used, e.g., "? environ set NEWVAR=value" or
 1788     "? environ link EXISTINGVAR".
 1789 
 1790 - -H#68 and -L#71 have been decoupled:
 1791    it used to be -e#65 -L#71 instead!
 1792 
 1793 - *NAIL_{HEAD,TAIL}* have been obsoleted in favour of
 1794   *message-inject-head*#465 and *message-inject-tail*#466.
 1795 
 1796   *NAIL_HIST{FILE,SIZE}* have been obsoleted in favour of
 1797   *history-file*#436 and *history-size*#439.
 1798 
 1799   *NAIL_EXTRA_RC* has been obsoleted in favour of
 1800   *mailx-extra-rc*#459.
 1801 
 1802   *batch-exit-on-error* has been obsoleted by *errexit*#413, which
 1803   works just like the POSIX sh(1)ell "set -e" construct; the
 1804   `ignerr'#125 command modifier (`-' for command escapes in compose
 1805   mode, and see below) can be used to ignore command errors even
 1806   then.  (This will remain even if we at some later time will
 1807   support at least some of the sh(1) constructs which "swallow"
 1808   failures with set -e.)
 1809 
 1810   *bsdannounce* is obsolete, the feature is integrated in
 1811   *header*#432 as this is much more useful.  (This is however also
 1812   dependent upon the also new but well-known $POSIXLY_CORRECT#631
 1813   <> *posix*#513, but that is just how it is; these affect more
 1814   behaviour, and increasing.)
 1815 
 1816 - Colour support has been changed backward in- and upward (from
 1817   user interface side) compatibly, see the manual section
 1818   "Coloured display"#18.
 1819 
 1820   + New commands: `colour'#160 and `uncolour'#161.
 1821     You can define context-sensitive, terminal-capability-
 1822     sensitive settings, e.g.:
 1823 
 1824       if terminal && [ "$features" =% +colour ]
 1825         colour iso  view-header ft=bold,fg=magenta,bg=cyan
 1826         colour 256  view-header ft=bold,fg=208,bg=230 subject,from
 1827         colour mono view-header ft=bold
 1828         colour mono view-header ft=bold,ft=reverse subject,from
 1829       endif
 1830 
 1831   + The variable *colour-pager*#395 defines whether colour and font
 1832     attribute sequences should be generated when viewing something
 1833     in $PAGER#629.
 1834 
 1835   + Set the variable *colour-disable*#394 to turn colour off
 1836     without affecting established settings.
 1837 
 1838   + It is deduced via termcap(5) (see below) whether the terminal
 1839     supports colors, e.g., "$ s-nail -Stermcap=Co#256".
 1840     This is also true if we don't have termcap support.
 1841 
 1842   + Support for 256-colour terminals. (Gavin Troy)
 1843 
 1844 - `source'#280 series support shell pipes if the last character
 1845    of the "filename" ends with a vertical bar |, e.g.,
 1846 
 1847       ? source 'gpg -qd ~/.s-nailrc-private.gpg |'
 1848 
 1849 - Shell pipes are also supported as targets for `move'#232,
 1850   `copy'#165 etc., yet unfortunately not with via a sh(1)ell token
 1851   parser, so that the target still has to be a single argument.
 1852 
 1853       ? copy . '| cat; echo huhu'
 1854 
 1855 - Support for custom headers via the new `~^'#319 compose-mode
 1856   command escape and in addition, or alternatively, with the
 1857   internal variable *customhdr*#403, which also can be covered by
 1858   `localopts'#217.  (Sergey Matveev)
 1859 
 1860   + Support of $ORGANIZATION has been dropped.
 1861 
 1862   + Command escape `~e'#325 supports _any_ header.
 1863 
 1864   + Command escape `~^'#319 supports _any_ header.
 1865 
 1866 - New -:#55 command line option can be used to more easily select
 1867   which startup files should be loaded, e.g., -:/ loads none.
 1868   (Robert Elz)
 1869 
 1870 - `account'#140s and *folder-hook*#421s now have `localopts'#217
 1871    enabled by default.
 1872 
 1873 - A first simple form of compose-mode hooks has been implemented:
 1874   *on-compose-enter*#488, *on-compose-leave*#489 and
 1875   *on-compose-cleanup*#487 can be set to macros which get invoked
 1876   at appropriate times.
 1877   For the `resend'#261 series there is *on-resend-enter*#496 and
 1878   *on-resend-cleanup*#495: this is very likely to change once
 1879   true message access is possible even in this mode.
 1880 
 1881   An even more powerful mechanism is available via the also new
 1882   *on-compose-splice*#490 and *on-compose-splice-shell*#491 hooks.
 1883   These are executed in child processes and communicate with the
 1884   parent via their standard input and output, and therefore can
 1885   do anything and act as if they were the user.
 1886 
 1887   `localopts'#217 are enabled and cannot be disabled (and extend
 1888   until the message is sent).
 1889   (Jens Schleusener, Rudolf Sykora)
 1890 
 1891     ? set on-compose-splice=ocs
 1892     ? define ocs {
 1893       read ver
 1894       echo Splice protocol version is $ver
 1895       echo '~^header list'
 1896       read hl; vput vexpr es substring "${hl}" 0 1
 1897       if [ "$es" != 2 ]
 1898         echoerr 'Failed to read header list, bailing out'
 1899         echo '~x'
 1900       elif [ "$hl" @i!% ' cc' ]
 1901         echo '~^header insert cc Diet is your <mirr.or>'
 1902         read es; vput vexpr es substr "${es}" 0 1
 1903         if [ "$es" != 2 ]
 1904           echoerr 'Failed to insert Cc:, bailing out'; echo '~x'
 1905         end
 1906       end
 1907     }
 1908 
 1909 - "The .netrc file"#37
 1910 
 1911   + gained support for comments.
 1912     (Walter Alejandro Iglesias, Ralph Corderoy)
 1913 
 1914   + `netrc'#236 now has a "load" subcommand.
 1915 
 1916   + the new *netrc-pipe*#482 obsoletes OPT_AGENT and
 1917     *agent-shell-lookup*, and can be used to load an encrypted
 1918     .netrc file, e.g.:
 1919 
 1920       ? set netrc-lookup netrc-pipe='gpg -qd ~/.netrc.gpg'
 1921 
 1922     I.e., this is in usual .netrc syntax and thus possibly much
 1923     nicer than saying "? source 'gpg -qd ~/.credentials.gpg |'".
 1924 
 1925 - termcap(5) / terminfo(5) support has been changed backward in-
 1926   and upward (from user interface side) compatibly, please read
 1927   "On terminal control and line editor"#17.
 1928 
 1929   + OPT_TERMCAP is by default enabled.
 1930     The new, by default enabled, configuration option
 1931     OPT_TERMCAP_VIA_TERMINFO can be used to (try to) use
 1932     terminfo(5) instead.
 1933 
 1934   + The variable *termcap*#581 can be used to freely define or
 1935     override terminal capabilities, and *termcap-disable*#583 will
 1936     disable interaction with the chosen library, leaving only
 1937     *termcap* in charge.
 1938 
 1939     To use the so-called ca-mode on supporting terminals,
 1940     effectively turning S-nail into a fullscreen application,
 1941     *termcap-ca-mode*#582 must be set.
 1942 
 1943   + The built-in line editor has been rather completely rewritten
 1944     to be the Mailx-Line-Editor (OPT_MLE, default yes), and
 1945     supports wide glyphs (if possible), infinite line lengths
 1946     (2 GB) and more.  Tabulator expansion is no longer an option
 1947     (but needs fnmatch(3)).
 1948 
 1949   + Optionally (OPT_KEY_BINDINGS, default yes) it has become
 1950     possible to freely define key bindings for the MLE via the new
 1951     `bind'#149 and `unbind'#150 commands.  These key bindings can
 1952     make use of termcap(5) and/or terminfo(5) names.  The MLE will
 1953     install a set of default bindings (unless there is a set
 1954     *line-editor-no-defaults*#454), more so with OPT_TERMCAP,
 1955     i.e., try "? bind*".
 1956 
 1957     Sufficient support provided, one can now, e.g., type "p " and
 1958     then collect the message numbers to type, scrolling forward
 1959     and backward via key-bindings, without losing the line
 1960     content, then commit the final line.
 1961 
 1962   + OPT_EDITLINE and OPT_READLINE support have been dropped.
 1963     The new MLE should not miss anything.  Does it?
 1964     Tip: in an UTF-8 locale try "? !touch /tmp/hall{,öchen}" and
 1965     then autocomplete that: once, then ^Q, and again.
 1966 
 1967 - `source'#280 can be used in `call'#151ed macros.
 1968   What sounds so innocent replaced an entire machinery and got rid
 1969   of a brilliant idea of Kurt Shoens from the 70s, but which never
 1970   worked with Nail/Heirloom extensions, namely macros, and in the
 1971   right order.
 1972   Accompanying this -X#86 can (dig multiline arguments and can) be
 1973   used to define macros and run them etc.  Should work:
 1974 
 1975     $ s-nail -X'define x {' -Xversion -Xx -X'}' -X'call x'
 1976     $ s-nail -X'source \' -X'"echo version|"' -Xx
 1977 
 1978   Macros can be `undefine'#171d from within themselves, and re-
 1979   `define'#170d.  It is still not possible to define macros
 1980   from within macros, and/or have inner macros, not to talk
 1981   about local scoping or anything more sophisticated such.
 1982 
 1983 - -u#83 / $LOGNAME#622 ($USER) handling has been redefined,
 1984   and "-u USER" is now exactly the same as "-f %USER", and
 1985   $LOGNAME (and $USER) is actively set to the active user.  (Afan)
 1986 
 1987   $LOGNAME#622 is POSIX standardized and henceforth used and
 1988   preferred over $USER, which came from BSD.  (Todd C. Miller)
 1989 
 1990 - In the future (at least non-message-list) argument handling will
 1991   be changed backward-incompatibly to be sh(1)ell compatible (and
 1992   thus POSIX standardized), see "Shell-style argument quoting"#23.
 1993   New commands use it already today (`bind'#149, `colour'#160,
 1994   `headerpick'#208), some others (most importantly, `set'#268) can
 1995   be forced to do so via the new `wysh'#130 command prefix, as in:
 1996 
 1997     ? wysh set message-inject-tail=$'\n--steffen'
 1998     ? bind base $'\cA,\x61' 'echo control-A and small a'
 1999 
 2000 - We now actively manage *umask*#602: 0077 by default, but an
 2001   empty string will use the setting that is active upon startup.
 2002   Just like changes to (known) environment variables, this setting
 2003   will also be inherited by any child process.
 2004   (Walter Alejandro Iglesias)
 2005 
 2006 - Anything SENDMAIL / *sendmail*-ish has been renamed to *mta*#474,
 2007   *mta-arguments*#476, *mta-no-default-arguments*#477 and
 2008   *mta-argv0*#479.
 2009 
 2010   The reason is that in v15 we won't even have *smtp*: it is just
 2011   another form of MTA, and thus obsolete by itself.
 2012   Note that *mta-arguments* is now parsed via the shell-token
 2013   parser, so the following ends up exactly as desired.
 2014 
 2015     ? set mta-arguments='-t -X "/tmp/my log"'
 2016 
 2017   For now we support a hack that understands a file:// URL in
 2018   *mta*, too, but that is also the default if there is no protocol.
 2019   E.g.: "? set mta=smtp://a:b@xy.z"
 2020 
 2021 - The "spamd" *spam-interface*#568 is obsolete.  I haven't tested
 2022   it since my main machine died, it is error prone since it assumes
 2023   internals of the spamassassin wire protocol, and there never was
 2024   a speed improvement over "spamc".  (However it could react upon
 2025   the "is-spam" state of a message, which "spamc" doesn't allow.)
 2026 
 2027 - The new *inbox*#447 variable will henceforth be looked up when
 2028   searching for a primary system mailbox (as in "? File %"),
 2029   followed by the usual $MAIL#623 and compile-time defined local
 2030   mailspool search.  (Stephen Isard, Jürgen Daubert)
 2031 
 2032 - The semantic of -a#57 and `~@'#317 have been changed, and both
 2033   commands now use the same syntax:
 2034 
 2035     -a file[=input-charset[#output-charset]]
 2036 
 2037 - New "failinvaddr" keyword for *expandaddr*#415.
 2038 
 2039 - We finally "can" the so-called (by myself) "Dr. Problem" (a bit):
 2040   (Dr. Werner Fink)
 2041 
 2042     $ </dev/null s-nail -d:/ -sTrödler 'Dr. D. Iet <z@a.k>' 2>&1 |\
 2043       grep To:
 2044     s-nail: >>> To: "Dr. D. Iet" <z@a.k>
 2045 
 2046   This can be done via the new `addrcodec'#142, too, note this
 2047   supports multiple modes (and the `vput'#129 command modifier):
 2048 
 2049     $ echo 'addrcodec e Dr. Diet <to@fu.soj> Curd' | s-nail -#:/
 2050     "Dr. Diet Curd" <to@fu.soj>
 2051 
 2052 - All commands with the string "codec" in their name use different
 2053   argument quoting, namely none at all, please read
 2054   "Raw data arguments for codec commands"#25.
 2055   This means that `urlcodec'#298 (and `imapcodec'#661) has
 2056   slightly changed semantics.
 2057   And, while here: there is a new `shcodec'#270, too.
 2058 
 2059 - We gained "Command modifiers"#21: `\'#124 (avoid expansion of
 2060   `commandalias'#162es), `vput'#129 (store result in variable),
 2061   `ignerr'#125  (ignore an error of the following command, even
 2062   if the new *errexit*#413 is set), `wysh'#130 (use shell-style
 2063   arguments).
 2064 
 2065     $ echo 'vput cwd resvar;echo $resvar' | s-nail -#:/
 2066     /home/steffen/src/nail.git
 2067 
 2068   And the usual sh(1) stuff: `return'#263, `shift'#274, `eval'#191,
 2069   plus a `xcall'#306 stack-avoidance optimization (to be used in
 2070   place of a `call'#151 which would be the last called command).
 2071   And an "expr(1) like thing", yet simple, `vexpr'#302.
 2072 
 2073     $ echo 'vexpr + 1 2' | s-nail -#:/
 2074     00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000011
 2075     03 | 0x3 | 3
 2076     $ echo 'vput vexpr resvar + 1 2;echo $resvar' | s-nail -#:/
 2077     3
 2078 
 2079   We actually start walking (*?*#347, *^*#349).
 2080 
 2081     ? vput vexpr res regex 'bananarama' 'Bana(.+)' '\$1\$0'
 2082     ? echo $?/$^ERRNAME :$res:
 2083     1/NODATA ::
 2084     ? vput vexpr res iregex 'bananarama' '(.+)rama' '\$1\$0'
 2085     ? echo $?/$^ERRNAME :$res:
 2086     0/NONE :bananabananarama:
 2087 
 2088   The command `vpospar'#303 can be used to manage the stack of
 2089   positional parameters, i.e., much like "set --".
 2090   It also offers the possibility to save and restore the stack to
 2091   and from variables.  Etc.
 2092 
 2093   Btw., to contact the maintainer (make.rc variables
 2094   VAL_CONTACT_WEB and VAL_CONTACT_MAIL):
 2095 
 2096     ? echo $contact-web; eval mail $contact-mail
 2097 
 2098 - `if'#214 no longer performs automatic number conversion, we
 2099   use the explicit -lt, -gt etc. syntax of the sh(1).
 2100   Note: `if' will change to be almost identical to sh(1) if(1),
 2101   so please ensure proper test bracketing, even if it is less
 2102   convenient.
 2103 
 2104   Moreover, the default string comparison mode has changed to
 2105   case-sensitive, just like in the shell.  This is because in the
 2106   future this crux with trigger characters will vanish and `if'
 2107   etc. will simply slurp in already expanded shell tokens, it will
 2108   act like the shell in that respect.  We have modifiers, though,
 2109   yet only "@i" for case-insensitivity, also for regex matches:
 2110 
 2111     LC_ALL=C
 2112     i=`LC_ALL=C.utf8 s-nail -:/ -# -X '
 2113        \define cset_test {
 2114           \if [ "${ttycharset}" @i=% utf ]
 2115              \echo $LC_ALL
 2116              \xit 0
 2117           \end
 2118           \if [ "${#}" -gt 0 ]
 2119              \wysh set LC_ALL=${1}
 2120              \shift
 2121              \eval xcall cset_test "${@}"
 2122           \end
 2123           \xit 1
 2124        }
 2125        \call cset_test C.UTF-8 POSIX.utf8 POSIX.UTF-8 \
 2126           en_EN.utf8 en_EN.UTF-8 en_US.utf8 en_US.UTF-8
 2127    '`
 2128    [ $? -eq 0 ] && UTF8_LOCALE=$i
 2129 
 2130   Please note the `eval' in 'eval xcall cset_test "${@}".  This is
 2131   a difference of S-nail/mailx and the sh(1)ell that will remain,
 2132   as documented in "COMMANDS"#20: whereas the shell implements
 2133   a language and performs standardized expansions on the line
 2134   until finally the command is called, S-nail will decide the type
 2135   of command line parsing dependent on the seen command, and will
 2136   then perform a single expansion.  Therefore "${@}" will expand
 2137   to multiple arguments if $# is greater 0, but it will expand to
 2138   the empty string otherwise, which is not furtherly expanded away
 2139   since it is meaningless like it is in the shell: therefore $#
 2140   will be 1 (the empty string) not 0.
 2141 
 2142 - Using an explicit proto:// prefix should get you the desired
 2143   thing apart of *newfolders*#483, e.g.:
 2144 
 2145     ? File maildir:///tmp/x.mdir
 2146     ? copy * file:///tmp/x.mbox
 2147 
 2148 - New variable *record-files*#528 can be set to extend the meaning
 2149   of *record*#527.  *record-resent*#529 was there already.
 2150 
 2151 - New variable *ifs*#443 acts a bit like the sh(1)ell's $IFS for,
 2152   e.g., the new `read'#250 command.
 2153 
 2154   There is a `readctl'#253 command which can be used to manage
 2155   the active channel used by `read'#250.
 2156 
 2157 - The `~' alias for `call'#151 is gone.
 2158 
 2159 - `mimetype'#224 only allows specification of a single type per
 2160   call, on the other hand no need to quote that.
 2161 
 2162 - `mimeview'#226 must now be used explicitly to look at any
 2163   non-text MIME part, for normal display etc. purposes we only
 2164   support "copiousoutput"#649 MIME handlers.
 2165 
 2166 - New *socks-proxy*#567 can be used to proxy all network traffic
 2167   over a SOCKS5 proxy.  (Gaetan Bisson)
 2168 
 2169 ChangeLog
 2170 ^^^^^^^^^
 2171 
 2172 - The manual has seen another major overhaul, all the variables
 2173   are now documented in a single, sorted list, and many
 2174   clarifications should have been added.  I hope it has become
 2175   a better read.
 2176   (Predrag Punosevac, Michael Convey, Hariskar, Rudolf Sykora,
 2177   Respiranto, Thomas Dickey, Donald Mugnai)
 2178 
 2179 - To support RFC 1524 a.k.a. .mailcap files (see below) many
 2180   "trigger"-characters have been added for *pipe-TYPE/SUBTYPE*#500,
 2181   which may (rarely) affect existing values.
 2182   The .mailcap support itself is not yet implemented.
 2183 
 2184 - *mime-counter-evidence*#470 gained bit 4 (perform proper in-depth
 2185   content inspection as necessary; set to 0xE for all bits).
 2186   (Aharon Robbins)
 2187 
 2188 - Maildir paths are now created recursively as necessary.
 2189   (Justin Ellingwood)
 2190 
 2191 - -M#72 and -m#73 options have been added to enforce a special
 2192   send mode that will flag standard input / the given file with
 2193   the specified / detected MIME 'Content-Type:'.  This can be used
 2194   to directly send, e.g., HTML log output.
 2195   (Viktor Szépe, Ralph Corderoy)
 2196 
 2197 - Disallow symlinks on writable files.  Note this requires
 2198   O_NOFOLLOW support for the operating-system-call open(2), but
 2199   which has been standardized a long time ago.
 2200   (Matthew Dillon)
 2201 
 2202 - `retain'#262, `ignore'#215 etc. now differentiate in between
 2203   From (the From: header) and From_ (the MBOX ident).
 2204 
 2205   In fact we now have a new `headerpick'#208 command which
 2206   is a multiplexer for all retain and ignore lists used, call it
 2207   without arguments to see the current setting(s).
 2208   In v15 only `headerpick' and the standard-imposed wrappers
 2209   `retain' and `ignore' will remain, all other wrappers will
 2210   vanish.  Regular expressions can now be used if available:
 2211 
 2212     ? headerpick
 2213       headerpick type retain blahblahblah cc date from \
 2214         mail-followup-to message-id openpgp reply-to subject to \
 2215         user-agent
 2216       #headerpick type ignore currently covers no fields
 2217       #headerpick save retain currently covers no fields
 2218       headerpick save ignore '^Original-.*$' '^X-.*$' '^DKIM.*$'
 2219       headerpick forward retain cc date from list-id \
 2220         mail-followup-to openpgp reply-to subject to
 2221       #headerpick forward ignore currently covers no fields
 2222 
 2223 - `top'#290 has been rewritten completely, `Top'#289 is new.
 2224   It uses a built-in set of retain/ignore headers, but it is
 2225   possible to register a custom set via `headerpick'#208.
 2226   Also, *toplines*#598 has been extended a bit and the new
 2227   *topsqueeze*#599 variable may pimp your `top' experience.
 2228 
 2229     ? headerpick top retain add subject
 2230     ? top
 2231     [-- Message  1 -- 87 lines, 4791 bytes --]:
 2232     Subject: Re: I can't dist to myself
 2233 
 2234     I wrote:
 2235         3.22. bounce_delivered
 2236 
 2237 - `features' has been dropped, `version'#301 extended.
 2238 
 2239 - The *prompt*#515 handling has changed: we lost the capability to
 2240   expand \?, \@ and \$, instead new "private" variables *?*#347,
 2241   *account*#360, *mailbox-resolved*#457 and
 2242   *mailbox-display*#456 have been introduced, and the prompt
 2243   is completely shell expanded (thus twice with `wysh' or in v15),
 2244   as if dollar-single-quote quoted.  We do support the reverse-
 2245   solidus escaped bracket notation for embedding characters which
 2246   should not be counted when calculating the width of the prompt.
 2247   The `colour'#160 command has a slot for the prompt colour.
 2248   We gained *prompt2*#516 as a second level prompt.
 2249 
 2250     ? var prompt
 2251     wysh set \
 2252       prompt='?\${?}!\${!}[\${account}#\${mailbox-display}]? '
 2253 
 2254 - The filename "-" can be used as a receiver, e.g.,
 2255 
 2256     $ echo Hey,\ you | s-nail -:/ -Sexpandaddr -sUB -
 2257 
 2258 - The -s#80 command line option, the `~s'#339 command escape
 2259   as well as the corresponding slots of `~^'#319 will actively
 2260   strip [\r\n] from their value (Debian #419840).
 2261 
 2262 - New `read'#250 and `echoerr'#182 commands, mostly for
 2263   *on-compose-splice*#490.
 2264   But also `echon'#183 and `echoerrn'#184, which do not write
 2265   a trailing newline.
 2266 
 2267 - New variable *r-option-implicit*#525 may be helpful to those
 2268   who regulary need the functionality of the -r#78 command
 2269   line option.  (Felipe Gasper, Martin Neitzel)
 2270 
 2271 - By using new "pseudo-URLs" one can automatize the use of S/MIME
 2272   keys / (certificates / intermediate include certificates) with
 2273   passwords.  E.g., to drive bob@exam.ple, set
 2274   *smime-sign-cert-bob@exam.ple* to the private key / certificate
 2275   pair as usual, the password lookup will then be performed for
 2276   bob@exam.ple.smime-cert-key, bob@exam.ple.smime-cert-cert and
 2277   bob@exam.ple.smime-include-certs.
 2278   Like this the password can be stored in an encrypted .netrc file
 2279   when *netrc-lookup*#481 and *netrc-pipe*#482 are set, or it may
 2280   be stored in an encrypted resource file that has been loaded via
 2281   `source'#280 as a simple *password*#498 variable.
 2282 
 2283   Note that the prompting that happens as a last resort of
 2284   password lookup will still interfere with a possibly running
 2285   $PAGER#629 instance, dependent on the setting of *crt*#402, of
 2286   course.  Proper job control handling and recognizing that we are
 2287   running $PAGER when doing that prompt is a TODO for v15.  Sorry.
 2288 
 2289 - Some commands, like `set'#268, `help'#211, `list'#216,
 2290   `mlist'#227 etc., now react upon the setting of *verbose*#605
 2291   and(/or) *debug*#406.
 2292 
 2293 - `write'#305 uses iconv(3) as appropriate.
 2294 
 2295 - *mbox-rfc4155*#462 has first been dropped, and was then
 2296   reintroduced with different semantics.  Because, it can be
 2297   helpful if a messed up MBOX is read, in which case we henceforth
 2298   will warn you and point you to this:
 2299 
 2300     ? define mboxfix {
 2301       \localopts yes; \wysh set mbox-rfc4155;\
 2302         \wysh File "${1}"; \eval copy * "${2}"
 2303     }
 2304     ? call mboxfix /tmp/bad.mbox /tmp/good.mbox
 2305 
 2306   P.S.  Here you see how weird the current thing still is, in v15:
 2307 
 2308     ? define mboxfix {
 2309       localopts yes; set mbox-rfc4155; File "${1}"; copy * -- "${2}"
 2310     }
 2311 
 2312   And also in v15 we will not apply (proper) so-called MBOXO
 2313   quoting, but instead (simply MIME) re-encode mail messages.
 2314 
 2315 - `call_if'#152 is new and, different to "? ignerr call", silent
 2316   and not messing with the return status.
 2317 
 2318 - The new *smime-ca-flags*#552 and *ssl-ca-flags*#? can be used
 2319   to fine-tune X509_STORE_set_flags(3) a.k.a the X509 CA
 2320   certificate verification.
 2321 
 2322     ? set ssl-ca-flags=partial-chain
 2323     ? wysh set smime-ca-flags="${ssl-ca-flags}"
 2324 
 2325   Also, *ssl-curves*#? for TLSv1.3.
 2326 
 2327 - Socket connections use TLS S(erver)N(ame)I(ndication) as
 2328   appropriate (RFC 7817).
 2329 
 2330 - `alternates'#145 checks arguments and supports `vput'#129.
 2331   It by default no longer replaces but appends alternates, unless
 2332   *posix*#513 mode is active.  There is a new `unalternates'#146
 2333   command to remove alternates.
 2334 
 2335 - A new `charsetalias'#155 command.  (Pietro Cerutti, mutt#3925)
 2336 
 2337 - New commands `filetype'#195 and `unfiletype'#196: in the future
 2338   we will no longer know any builtin filetypes, in fact we already
 2339   simulate .gz etc. via the new mechanism as necessary:
 2340 
 2341     ? filetype \
 2342        bz2 'bzip2 -dc' 'bzip2 -zc' \
 2343        gpg 'gpg -d' 'gpg -e' \
 2344        gz 'gzip -dc' 'gzip -c' \
 2345        xz 'xz -dc' 'xz -zc' \
 2346        zst 'zstd -dc' 'zstd -19 -zc' \
 2347        zst.pgp 'gpg -d | zstd -dc' 'zstd -19 -zc | gpg -e'
 2348 
 2349 - `~<'#314 now offers a "- [HERE-delimiter]" mode for pasting etc.
 2350   (Ralph Corderoy)
 2351 
 2352 - `exit'#192 and `quit'#249 take an optional exit status.
 2353   (That is not fixated yet, though.)
 2354 
 2355 - We have a useful -h / --help output.  (Doug McIlroy)
 2356 
 2357 - *encoding* obsoleted in favour of new *mime-encoding*#471, which
 2358   now defaults to base64.
 2359 
 2360 - *allnet*#362 now works (broken since nail 10.00, 2002-09-29).
 2361 
 2362 Appendix
 2363 ^^^^^^^^
 2364 
 2365 The complete changelog of commits in between two versions OLD and
 2366 NEW can be inspected by using the git(1) `log' command:
 2367 
 2368   $ git log --reverse --topo-order --abbrev-commit OLD..NEW
 2369   # Only topic branch headers (--no-merges for content commits only):
 2370   $ git log --oneline --reverse --topo-order --merges OLD..NEW
 2371   # Same, but truly accessible:
 2372   $ git log --oneline --reverse --topo-order --merges --parents OLD..NEW |
 2373     while read c1 c2 c3 c4 c5 c6; do
 2374       printf "%-24s: \$ git log --oneline --no-merges %s ^%s\n" \
 2375         "${c6}" "${c1}" "${c2}";
 2376     done
 2377 
 2378 Entries for releases before v14.9.0 have been cut off and can be
 2379 found in the git(1) repository:
 2380 
 2381   v14.8.0 - v14.8.16: $ git show v14.8.16:NEWS
 2382   v13     - v14.8.5 : $ git show v14.8.5:NEWS
 2383   9.0     - 12.5    : $ git show heirloom:ChangeLog
 2384 
 2385 Also accessible via HTTPS?, just replace X.Y.Z accordingly:
 2386 
 2387   \https?://git.sdaoden.eu/cgit/s-nail.git/tree/NEWS?h=vX.Y.Y
 2388 
 2389 For even older releases you need to look into the [timeline]
 2390 branch, but no changelog has been administrated for them.
 2391 
 2392 # s-tm-mode