lp:ubuntu/karmic/mksh

Related source package recipes

No recipes using this branch. (?)

Create packaging recipe

Related snap packages

Related bugs

Related blueprints

19. By Thorsten Glaser on 2009-10-10

Exclude dietlibc on powerpc/ppc64 and sparc (LP: #381332)

18. By Thorsten Glaser on 2009-10-08

* Merge from Debian unstable. Remaining Ubuntu changes:
  - debian/mksh.postinst: hardcode debconf question if mksh should
    be installed as /bin/sh to false (see LP: #348872) but retain
    debconf infrastructure so that, if this was set to true in old
    installations, it will correctly be removed on upgrade
  - debian/rules: remove “-combine” option from Build.sh invocation
    to work around a gcc bug (see LP: #375604)

17. By Thorsten Glaser on 2009-06-11

* Merge from Debian unstable (LP: #385887). Remaining Ubuntu changes:
  - disable debconf mksh/sh
  - do not use -combine for build
  - skip dietlibc on powerpc and sparc (LP: #381332)

16. By Thorsten Glaser on 2009-05-28

* debian/mksh.postinst: hardcode debconf question if mksh should
  be installed as /bin/sh to false (LP: #348872) but retain debconf
  stuff so that if this was set to true in previous packages, it
  will correctly be removed on upgrade
* debian/rules: remove “-combine” option from Build.sh invocation
  (LP: #375604)

15. By Luca Falavigna on 2009-01-10

debian/control: build-depend on locales instead of locales-all,
Ubuntu does not provide the latter.

14. By Thorsten Glaser on 2008-12-14

* Reword package description, avoid things not of interest to
  a Debian user; Closes: #505882
* Switch from patching copyright to merely prepending information
* Remove all patches, as they have been integrated upstream; add
  ‘-DMKSH_BINSHREDUCED’ to CPPFLAGS to enable the functionality
* New upstream version; complete ChangeLog:
  - [tg] Add check for naming the output file “scn” instead of “a.out”
    or “a.exe” when compiling scn.c, for Haiku, from Adam “replaced” Hoka
  - [tg] Rewrite utf_backch macro and x_bs2 function into a combined
    x_bs3 function for the Emacs editing mode, to optimise them and get
    rid of the use of __typeof__ (suggested by Anders “ragge” Magnusson
    after the problem was spot by replaced) and one of the uses of the
    statements-as-expressions feature
  - [ahoka] Add mirtoconf check for nice(3), missing on Haiku
  - [tg] Remove all uses of the statements-as-expressions feature by
    rewriting the source code accordingly and optimising some parts
  - [tg] Recognise nwcc (Nils Weller’s C compiler) in Build.sh
  - [tg] If <strings.h> exists, pull it in for strcasecmp(3)
  - [tg] Welcome QNX/Neutrino; work around broken /bin/ed
  - [tg] Simplify, shorten, speed up PS1 in dot.mkshrc
  - [tg] Remove some dead code courtesy of scan-build native runs
  - [tg] Add some casts to prevent LLVM+Clang warnings
  - [tg] Work around llvm-gcc-4.2.1 -Wformat pickyness
  - [tg] Add new Build.sh option ‘-llvm’ (clang, llvm-gcc)
  - [tg] Speed up mirtoconf if ‘-DMKSH_ASSUME_UTF8’ is set
  - [tg] Add the workaround for Debian #492377 into the main mirtoconf
    function (hiding gcc errors during the configure phase) because
    Gentoo has similar scanner issues; reported by Hanno Böck
  - [tg] If an MKSH_SMALL has arc4random(3), skip the rand(3)/srand(3)
    fall-back altogether to shrink size, also removed need for time(3)
  - [tg] Fix alias expansion recursion check if the word to be expanded
    is immediately followed by end of input, add test case; spotted by
    Michael Hlavinka in pdksh and mksh; RedHat #474115
  - [tg] Fix string/wdstring confusion preventing bashiop (&>foo) to
    work inside a function, add regression test, limit to 99 fds
  - [tg] change regression tests to ‘set -U’ or ‘set +U̲ instead of
    ‘set -o utf8-hack’, as well as query using ‘$-’ ipv ‘$(set +o)’
    (easier and more reliable)
  - [tg] Add comment to regression tests which can fail on slow machines
    or Cygwin environments due to timing issues
  - [tg] Remove a lot of superfluous casts, improve type cleanliness
  - [tg] Insert a couple of /* CONSTCOND */ for lint
* New source package and build structure:
  - distfile is now packaged *inside* the orig tarball instead of repak-
    kaged, similar to PostgreSQL
  - debian/rules “patch” target takes care of it, as per Policy Manual
  - debian-specific things (debian/ directory and all patches) are kept
    in a publically accessible CVS repository
  - VCS-CVS and VCS-Browser fields have been added to debian/control
  - source is extracted to ./mksh and built, using relative paths to the
    source code, in ./build/full and ./build/small; both are .cvsignore’d
  - most files now carry an RCS ID
  - paths in mksh.install, mksh.manpages, rules have been adjusted
* word-wrap debian/control lines to 80c
* add build dependency on cpio (for distfile extraction)
* for non-pbuilder testsuite runs (needed on gnubber), rename the ./manual
  file tested for to ./attended to clarify its purpose
* debian/rules: share code instead of duplicating it
* debian/rules: fix testsuite use of $? and $x (gmake wants $$? and $$x)
* debian/rules: use sensible stamp files and new paths, coming along with
  the new “patch” target
* remove the ‘-Q’ option from Build.sh invocation, which is gone

13. By Thorsten Glaser on 2008-09-19

* Apply upstream changeset 10048D15ABE2EA76C75:
  - Bring back automatic -o posix setting if the shell is invoked
    as “sh” or “-sh” (unless compiled with MKSH_SMALL), add
    regression test
  - If -o posix is set, do not keep file descriptors created via
    I/O redirection, as Korn Shells do, private; add regression
    test; Debian Closes: #499139; reported by Markus Schaber and
    Agustin Martin Domingo
* Change mksh-internal version number to document aberration from
  pristine source
* Add DM-Upload-Allowed control field to prepare for future updates
* Add workaround to “posix-mode-2” regression test to cope for GNU
  getopt idiocy to not stop argument parsing upon encountering the
  first non-flag argument: “ln -s mksh -sh” tries to parse -s and -h

12. By Thorsten Glaser on 2008-04-11

* New upstream release; changelog:
  - Move a portability define from sh.h to the setmode.c helper, as it’s
    only needed there, and we want to use the latter from MirMake as well
  - SECURITY: when spawning mksh on a new terminal, for example with
    “sudo mksh -lT/dev/ttyC7”, flush all of that tty’s I/O first
  - dot.mkshrc: ensure “ls” is no alias, don’t hardcode its path
* As dash won’t be the default /bin/sh without the current alternative
  handling / debconf mechanism in lenny (as per the mail from Martin
  Zobel-Helas), there is no need to act regarding our debconf scripts
  and /bin/sh ability, so I think this Closes: #469675

11. By Thorsten Glaser on 2007-10-25

* New upstream release; summary of changes:
  - Checks for symbol declarations are compile time checks, not link
    time checks; fixes (optional) arc4random on AIX
  - Widen the range for array indicēs to the entire unsigned 32-bit
    integer range (enough for ino_t on BSD); wrap numbers outside of
    that range into it for simplicity (e.g. foo[-1] = foo[4294967295])
  - Fix an internal error when using a pipeline as co-process
  - Relax requirement on compilation environment to provide certain types
  - Optimise some more for size (struct padding; functions → macros, …)
* Integrate galician translation, Closes: #447947

10. By Thorsten Glaser on 2007-09-11

* New upstream minor release (R31b); summary of changes:
  - Fix a syntax error in Build.sh checking for TenDRA
  - Fix typo (blsk → bksl) in check.t test naming
  - Autoscan for uint32_t, u_int etc. presence
  - Fix some memory leaks, mostly by NetBSD® via OpenBSD
  - The “unset” builtin always returns zero, even if the variable was
    already unset, as per SUSv3 (reported by Arkadiusz Miskiewicz via
    pld-linux and oksh)
  - In tab-completion, escape the question mark, reminded by cbiere@tnf
  - Fix a busy-loop problem, Debian #296446 via oksh
  - Fix a few display output problems in the build script
  - Shut up some gcc warnings on Fedora; beautify some code
  - Support OSF/1 with gcc2.8, thanks to Jupp Schugt
  - Fix gcc4 detection of __attribute__() on non-SSP targets
* debian/control: sync description with that of packages for other OSes
* debian/menu: Apps → Applications, as per Lintian
* debian/rules: do not run the testsuite with script on Debian GNU/HURD,
  because some translators seem to be unable to cope with the chroot
* arc4random.c: use uint<n>_t consistently, helps compiling on OSF/1