[Bug 1908025] Re: export var=~root does not expand ~root to /root

2020-12-28 Thread Jilles Tjoelker
I agree that this is a useful feature (including that export a=$b does not split words and export a=* does not generate pathnames); FreeBSD sh has had it for a few years: https://cgit.freebsd.org/src/commit/?id=84fbdd8ca068ec63c6c9a931303d3811738128cb . It is specified by

[Bug 932067] Re: Minus associativity not correct in arithmetic expansions

2012-02-17 Thread Jilles Tjoelker
This was fixed between dash 0.5.5 and 0.5.6 by commit 9655c1ac5646bde1007ecba7c6271d3aa98f294b Date: Tue Mar 9 12:52:30 2010 +0800 [ARITH] Fix binary operator parsing -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu.

[Bug 915158] Re: Process is not closing

2012-01-28 Thread Jilles Tjoelker
Catching SIGCHLD in dash does not help in this case, since dash is already gone (execve'ed) by the time the SIGCHLD would come. -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/915158 Title: Process

[Bug 915158] Re: Process is not closing

2012-01-20 Thread Jilles Tjoelker
You can see the same problem with any shell with a command like (sleep 3 exec /bin/sleep 6) because most versions of /bin/sleep do not clean up zombies (here, a zombie from an inherited child process). Dash does not consider the existence of a background process a reason to fork for an

[Bug 880120] Re: read -r does not protect backslashes

2011-10-22 Thread Jilles Tjoelker
*** This bug is a duplicate of bug 259671 *** https://bugs.launchpad.net/bugs/259671 The backslash escape sequences are being processed by echo, not read. You can avoid this inconsistency by doing something like printf %s\n '000\n999' | dash -c ' read -r f ; printf %s\n $f ' ** This bug has

[Bug 782701] Re: Cant paste commands in the dash

2011-05-16 Thread Jilles Tjoelker
** Also affects: unity Importance: Undecided Status: New ** Changed in: dash (Ubuntu) Status: New = Invalid -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/782701 Title: Cant paste

[Bug 774447] Re: Dash text input does not get focus

2011-05-16 Thread Jilles Tjoelker
** Also affects: unity Importance: Undecided Status: New ** Changed in: dash (Ubuntu) Status: New = Invalid -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/774447 Title: Dash text

[Bug 772899] Re: Can't access applications that minimize to the tray

2011-04-29 Thread Jilles Tjoelker
This applies to the Unity desktop shell, not the /bin/dash shell. ** Also affects: unity Importance: Undecided Status: New ** Changed in: dash (Ubuntu) Status: New = Invalid -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to

[Bug 737105] Re: Dash being open prevents system from blanking screen

2011-04-19 Thread Jilles Tjoelker
This bug is not relevant to the /bin/dash shell (normally /bin/sh is a link to this). ** Also affects: unity Importance: Undecided Status: New ** Changed in: dash (Ubuntu) Status: New = Invalid -- You received this bug notification because you are a member of Ubuntu Bugs,

[Bug 718396] Re: strings not comparable with or as described in man page

2011-02-15 Thread Jilles Tjoelker
You need to quote the operators to distinguish them from redirections. It may be appropriate to mention this in the man page. For example: $ [ 2.6 2.7 ]; echo $? 0 $ [ 2.6 2.7 ]; echo $? 1 -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to

[Bug 249620] Re: dash does not support multi-digit file descriptors

2010-12-11 Thread Jilles Tjoelker
Upstream seems to have said no fix is planned: http://www.mail- archive.com/d...@vger.kernel.org/msg00200.html Furthermore, I have tried this command in various other shells and it does not work in zsh, ksh93 and mksh either. (ksh93 has a different, incompatible way to use fds greater than 9.)

[Bug 675185] Re: [Hardy SRU] dash bug causes mysqld_safe to spin at 100% CPU

2010-11-16 Thread Jilles Tjoelker
Be sure to verify that the script works with bash's non-POSIX mode as well. A not so well known feature of bash is that it enters POSIX mode when called as sh. This causes various subtle differences. Therefore, a #!/bin/sh script with /bin/sh - bash might behave differently from a #!/bin/bash

[Bug 675185] Re: [Hardy SRU] dash bug causes mysqld_safe to spin at 100% CPU

2010-11-16 Thread Jilles Tjoelker
Be sure to verify that the script works with bash's non-POSIX mode as well. A not so well known feature of bash is that it enters POSIX mode when called as sh. This causes various subtle differences. Therefore, a #!/bin/sh script with /bin/sh - bash might behave differently from a #!/bin/bash

[Bug 658838] Re: Cant redirect from file names containing Cyrillic с

2010-10-15 Thread Jilles Tjoelker
*** This bug is a duplicate of bug 422298 *** https://bugs.launchpad.net/bugs/422298 ** This bug has been marked a duplicate of bug 422298 dash interpreter don't handle some unicode characters correctly * You can subscribe to bug 422298 by following this link:

[Bug 313403] Re: Fix /bin/sh

2010-10-15 Thread Jilles Tjoelker
*** This bug is a duplicate of bug 259671 *** https://bugs.launchpad.net/bugs/259671 ** This bug has been marked a duplicate of bug 259671 dash: echo builtin interprets backslash escape sequences * You can subscribe to bug 259671 by following this link:

[Bug 268929] Re: Wrong escape character processing in dash

2010-10-15 Thread Jilles Tjoelker
*** This bug is a duplicate of bug 259671 *** https://bugs.launchpad.net/bugs/259671 ** This bug has been marked a duplicate of bug 259671 dash: echo builtin interprets backslash escape sequences * You can subscribe to bug 259671 by following this link:

[Bug 259671] Re: dash: echo builtin interprets backslash escape sequences

2010-10-15 Thread Jilles Tjoelker
The Debian dash maintainer considers this a WONTFIX: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=550399 It seems unlikely that Ubuntu will want to override this. ** Summary changed: - dash parses backslashes in single quotes + dash: echo builtin interprets backslash escape sequences **

[Bug 598279] Re: read builtin can't handle some synthetic files

2010-09-10 Thread Jilles Tjoelker
** Bug watch added: Debian Bug tracker #595063 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=595063 ** Also affects: dash (Debian) via http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=595063 Importance: Unknown Status: Unknown -- read builtin can't handle some synthetic files

[Bug 598054] Re: dash: Redirection operator should not truncate file

2010-09-03 Thread Jilles Tjoelker
This was fixed in upstream in git commit 09363426739388813a53d63716d15163f6c43caf, March 2010. -- dash: Redirection operator should not truncate file https://bugs.launchpad.net/bugs/598054 You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu.

[Bug 249620] Re: dash does not support multi-digit file descriptors

2010-08-22 Thread Jilles Tjoelker
You may want this, but allowing only fd 0-9 is not a POSIX violation. XCU 2.7 Redirection says the maximum fd for redirection is implementation-defined and must be at least 9. This is sufficient for most applications. -- dash does not support multi-digit file descriptors

[Bug 598279] Re: read builtin can't handle some synthetic files

2010-08-13 Thread Jilles Tjoelker
I consider this a kernel bug, it should be possible to read regular files one byte at a time (like dash's and zsh's read builtins do). -- read builtin can't handle some synthetic files https://bugs.launchpad.net/bugs/598279 You received this bug notification because you are a member of Ubuntu

[Bug 422298] Re: dash interpreter don't handle some unicode characters correctly

2010-05-28 Thread Jilles Tjoelker
This patch has been pushed to dash.git. However, applying it to an older version of dash could be useful so people can take advantage of this fix without the possible instability and possible subtle incompatibilities of a new dash version. -- dash interpreter don't handle some unicode characters

[Bug 340571] Re: Bash PPID is wrong when parent terminates

2010-05-18 Thread Jilles Tjoelker
POSIX says PPID should be set once at shell initialization, and I have not seen a shell which does otherwise. -- Bash PPID is wrong when parent terminates https://bugs.launchpad.net/bugs/340571 You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to

[Bug 268929] Re: Wrong escape character processing in dash

2010-05-16 Thread Jilles Tjoelker
You can avoid these problems by using the printf utility instead of echo. -- Wrong escape character processing in dash https://bugs.launchpad.net/bugs/268929 You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. -- ubuntu-bugs mailing list

[Bug 548902] Re: shift function does not work as expected.

2010-05-15 Thread Jilles Tjoelker
I think you should just cope with this (by not shifting more than $#), as there are other shells with the same behaviour. -- shift function does not work as expected. https://bugs.launchpad.net/bugs/548902 You received this bug notification because you are a member of Ubuntu Bugs, which is

[Bug 301225] Re: can't define custom exit function

2010-05-15 Thread Jilles Tjoelker
I think the real problem is in gdm here: if it wants to read the user's shell startup files, it should do so using the user's login shell, not /bin/sh. User shell startup files cannot be guaranteed to even parse in /bin/sh, and changing /bin/sh to make it work degrades sh -n for actual /bin/sh

[Bug 503761] Re: bash behaves different on symlinks than the GNU tools

2010-05-15 Thread Jilles Tjoelker
This bug should be tightened down to the completion only. The discrepancy between cd and pwd on one hand and other utilities on the other hand is inevitable (unless you disable the symlink magic using set -P). I consider the fact that /bin/pwd shows the path without symlinks a bug in GNU

[Bug 356624] Re: `time` crashes in a script

2010-05-15 Thread Jilles Tjoelker
This is because bash's time is a keyword, not a builtin command ('type time' will show). Therefore it will only be recognized if it is in the input literally and unquoted. -- `time` crashes in a script https://bugs.launchpad.net/bugs/356624 You received this bug notification because you are a

[Bug 474638] Re: shell script error if read -a is used

2010-05-14 Thread Jilles Tjoelker
Dash deliberately does not support arrays, so this is not a bug in dash. I suggest changing the #!/bin/sh line at the start of the script to #!/bin/bash. (Note that bash enters posix mode when it is called as 'sh', so it may be a good idea to add (set -o posix) 2/dev/null set -o posix to the

[Bug 259671] Re: dash parses backslashes in single quotes

2010-05-14 Thread Jilles Tjoelker
The issue is not with the single quotes, but with dash's echo builtin. Apparently, the way they are compiled and configured by default on Ubuntu, dash's echo builtin interprets backslash escape sequences while bash's does not. According to POSIX, this interpretation is permitted; with the XSI

[Bug 509128] Re: Inconsistent output from Bourne Shell type command

2010-05-14 Thread Jilles Tjoelker
I don't consider this a bug. Given you say Bourne shell, see this session with an original Bourne shell (Heirloom toolchest jsh, close to Solaris 10 /bin/sh): $ type ls ls is /bin/ls $ type ls ls is /bin/ls $ ls / COPYRIGHT dev homeproctmp bin

[Bug 139097] Re: dash 0.5.3-5ubuntu2 variable expansion error using local

2010-05-14 Thread Jilles Tjoelker
local will not be removed from dash, as it is used too much, even though it is not in POSIX. The underlying problem is related to a feature in bash which also applies to some POSIX special builtins (export and readonly). This feature is not in POSIX and strictly speaking conflicts with it. Dash

[Bug 239561] Re: Dash remove prefix expansion (${a#b}) doesn't work in double quotes

2010-05-14 Thread Jilles Tjoelker
This has been discussed at the Austin Group (POSIX fixing/development) recently, but no definitive interpretation has been issued yet. The current standard is rather vague to allow various implementations to comply. I agree that bash's behaviour is better here, but it is possible that the