JehanneOS/mksh
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
2,190 Commits 2 Branches 0 Tags 6.4 MiB
5386859b3f
Commit Graph

2190 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
tg
712a3cc064 overhaul the random stuff ;-) 2004-05-24 19:06:55 +00:00
tg
bcd702035d file strlfun.c was initially added on branch tg. 2004-05-24 17:43:42 +00:00
tg
b8d0872109 save us from including GNU GPL'd content 2004-05-24 17:32:11 +00:00
tg
aff9c430e6 free AC_HEADER_STAT 2004-05-24 17:25:43 +00:00
tg
165d1927be free AC_LANG_C{,PLUSPLUS} - we don't do OS/2 2004-05-24 17:22:48 +00:00
tg
f30bb19422 un-GPL-ify the sys/wait.h check 
I still won't say I speak m4, but I'm getting closer,
although not as close as with sirc/wiki perl...
 2004-05-24 17:19:10 +00:00
tg
2745b4381b no twice sys/wait.h 2004-05-24 17:15:37 +00:00
tg
d2da772b39 further simplification; for mirbsdksh, GNU autoconf is 
only a portability, no Makefile-generating, tool
 2004-05-24 16:35:08 +00:00
tg
916f673e8a shutup autoconf 2004-05-24 16:26:27 +00:00
tg
c3babfb63d * mirbsdksh is now maintained in the HEAD branch again 
* remove aclocal.m4-local copy of the GNU GPL'd AC_TRY_RUN macro
  (makes no difference, except maybe on OS/2)
 2004-05-24 16:24:42 +00:00
tg
e727889c89 sync with latest ksh.1tbl 2004-05-23 13:03:17 +00:00
tg
ffe14a1de4 further minor manpage fixes 2004-05-23 13:01:35 +00:00
tg
b051c6e857 * merge OpenBSD 3.5-current 
* retain local changes to ls and md5 (md4, hint hint)
* fix up ksh (limits - we still have 4096 bytes input buffer;
  random usage)
* fix manpages (mdX, ksh)
 2004-05-23 12:47:01 +00:00
tg
daf9f30b2f cope with GNU groff's incapability to properly escape [] chars 2004-04-29 19:06:24 +00:00
tg
4f8d5a47a4 correct pointer type, noticed by gecko2 (Andreas G.), thanks 2004-04-29 18:58:16 +00:00
tg
1c11edfc64 null commit for version bump to mirbsdksh 1.6 2004-04-27 20:00:23 +00:00
tg
ce9d26a8fb * mirbsdksh and mirosksh != rksh 
* use arc4random familiy for randomness, if exists
* feed back randomness on reseed and variable assignments
 2004-04-27 19:59:57 +00:00
tg
10f86cd408 sync with configure.in on ksh_maintainer branch 2004-04-27 19:30:39 +00:00
tg
c4fec215ff back out; it slows down (interactive) work on slow machines 
quite a bit, and users can seed $RANDOM in their ~/.profile
by using RANDOM=$(dd if=/dev/prandom count=1 2>/dev/null |\
cksum | while read a b; do echo $a; done) instead.
 2004-04-26 18:38:20 +00:00
tg
b77a4be013 if ksh invoked as sh, don't seed with /dev/urandom 
(because even a 4-byte read slows configure/make down a lot)
 2004-04-26 18:28:07 +00:00
tg
5860828e51 oops, committed to wrong branch accidentally: 
Log message:
seed $RANDOM from /dev/urandom at startup, too.
 2004-04-26 18:24:39 +00:00
tg
7834ecbe3c document /dev/urandom use (ksh only) 2004-04-26 18:20:45 +00:00
tg
461c2a76e3 bump to mirbsdksh 1.5 2004-04-24 21:45:37 +00:00
tg
b04855f181 sync with ksh_maintainer tag 2004-04-24 21:45:00 +00:00
tg
5f7c1fdd91 these files aren't needed in-tree 2004-04-24 21:29:49 +00:00
tg
deedab0128 add Mac OSX notes (tested on Angelo Laub's iBook) 
add pre-installed ksh/bash note for various systems
(Solaris, Interix/SFU 3.5 in particular; I didn't
 try even without bash on Mac OSX)
 2004-04-23 23:06:07 +00:00
tg
0db78036f8 blame myself here ;) 2004-04-17 01:00:24 +00:00
tg
67ccae311e hello, reality check? 2004-04-17 00:51:19 +00:00
tg
4fcecac715 bump to 1.4 2004-04-17 00:50:28 +00:00
tg
8d3634d008 throw out a bunch more of 0x60 characters throughout code and comments 2004-04-17 00:47:20 +00:00
tg
72426f16e0 pull in latest configure 
reorder sections in build script a bit, because strip isn't always GNU

should build on interi^WSFU 3.5 now
 2004-04-17 00:37:08 +00:00
tg
a10bbd0c7f mirbsdksh-1.3 
tested on
* Linux shell.thinknerd.de 2.4.25-grsec #1 Mon Mar 29 08:50:10 CEST 2004 i686 GNU/Linux
* SunOS stinky 5.8 Generic_108528-21 sun4u sparc
  -> SHELL=ksh LDFLAGS=-ldl ksh ./BuildGNU.sh
  (using GNU bash for preparing should work, too)
* MirBSD odem.66h.42h.de 7 Kv#7uA9-20040323 GENERIC#358 i386
 2004-04-07 18:43:42 +00:00
tg
95ae34eb0c mkstemp check is there, use it 2004-04-07 17:24:40 +00:00
tg
85a77b6191 pull in ksh_maintainer branch: mkstemp() checks 2004-04-07 17:22:48 +00:00
tg
138338c10d build process improvements 
builds with -Werror -Wall and without -Wno-unused now
GNU now uses improvements and mkstemp, too
correct version variable for sh(1)
 2004-04-07 17:14:13 +00:00
tg
bc29832339 sync with ksh.1tbl 
sh.1tbl is now derived from it
 2004-04-07 16:56:19 +00:00
tg
455051d58e another fix round 2004-04-07 16:54:01 +00:00
tg
a8f426cf13 whoops. even better. 2004-04-07 16:48:57 +00:00
tg
22d9964850 man page fixes 2004-04-07 16:38:51 +00:00
tg
1387701f49 * more sections to strip 
* manpage fixes

tested on SuSE GNU today.
 2004-03-22 19:03:53 +00:00
tg
9981f691d1 in case someone wants to build a more small, posixlike sh... 2004-03-21 02:49:39 +00:00
tg
83b5247f5d d'oh; when specifying CPPFLAGS, gcc should use them. 
while here, restructure the file.

With these changes, mirbsdksh-1.2 passes all expectations and tests
on Walgrind GNU/Linux (-current version); thanks to cnuke@ (Josef
"Jupp" Söntgen) to having me a test account to validate this.
 2004-03-21 02:47:58 +00:00
tg
be20ad9892 better generation of config.h.in (if needed) 
please note this is still only used by BuildGNU.sh, not by
the MirOS build process
 2004-03-21 02:32:48 +00:00
tg
899ddca642 This is mirbsdksh-1.2, passing all expected (all but three) tests 
on MirOS BSD-current, and failing like 37 on Walgrind GNU/Linux
(probably some of these are environment-dependant):
FAIL bksl-nl.t:blsk-nl-ksh-1
FAIL bksl-nl.t:blsk-nl-ksh-2
FAIL eglob.t:eglob-bad-1
FAIL eglob.t:eglob-bad-2
FAIL eglob.t:eglob-infinite-plus
FAIL eglob.t:eglob-nomatch-1
FAIL eglob.t:eglob-match-1
FAIL eglob.t:eglob-case-1
FAIL eglob.t:eglob-case-2
FAIL history.t:history-basic
FAIL history.t:history-e-minus-1
FAIL history.t:history-e-minus-2
FAIL history.t:history-e-minus-3
FAIL history.t:history-e-minus-4
FAIL history.t:history-e-minus-5
FAIL history.t:history-list-1
FAIL history.t:history-list-2
FAIL history.t:history-list-3
FAIL history.t:history-list-4
FAIL history.t:history-list-5
FAIL history.t:history-list-6
FAIL history.t:history-list-7
FAIL history.t:history-list-r-1
FAIL history.t:history-list-r-2
FAIL history.t:history-list-r-3
FAIL history.t:history-subst-1
FAIL history.t:history-subst-2
FAIL history.t:history-subst-3
FAIL history.t:history-subst-4
FAIL history.t:history-subst-5
FAIL integer.t:integer-base-2
FAIL integer.t:integer-base-4
FAIL regress.t:regression-50
FAIL regress.t:regression-52
FAIL regress.t:regression-54
FAIL unclass2.t:env-prompt
FAIL version.t:version-1
 2004-03-21 02:12:51 +00:00
tg
9fe69adfe8 add (c) and "fix" build on GNU systems (tested) 
add MirOS to version
 2004-03-21 01:51:26 +00:00
tg
847035b24e note that public domain does not exist where this OS comes from 
(doesn't change much, though)
 2004-03-21 01:27:40 +00:00
tg
00d06ba328 Improvements from the "ksh_maintainer" tag (development branch): 
- use game-of-life
- add configure script (public domain, thus: miros licence)
  to make it portable
 2004-03-21 01:25:05 +00:00
tg
6c4e45f6e8 we don't have alpha, and stuff like that belongs 
into conf-end.h anyways (autoconf annoys)
 2004-03-21 01:19:44 +00:00
tg
a1f46dc93a the last sync with OpenBSD seems to have introduced more manpage errors 2004-03-21 00:50:40 +00:00
tg
4171cbc1e1 these files don't belong into mirbsdksh, either 2004-03-21 00:44:35 +00:00