a user reported this issue here:
http://bugs.gentoo.org/show_bug.cgi?id=92349
if you enable the PGRP_PIPE option, this simple example will fail:
$ ( exec 31; : `echo3` )
bash: 3: Bad file descriptor
disabling the PGRP_PIPE option and everything works as it should ... verified
that Debian works
dont know why i didnt forward this earlier ...
http://bugs.gentoo.org/show_bug.cgi?id=69407
user has attached many patches but again these are out of my league :)
-mike
___
Bug-bash mailing list
Bug-bash@gnu.org
some compilers (like Darwin's gcc-4) will attempt to locate a shared library
in the lib search path before searching for a static one ... the default
behavior of gcc on linux is to search all library paths for static or shared
versions of a library and use whichever it finds first (defaulting
in the bash 3.1 release, new options were added: -i -q -x
however, the doc/bashref.texi and doc/bash.1 files were not updated
to document these fun features
i've attached an updated patch based upon the ulimit.patch from Fedora
which documents these three options and adds two new ones ...
any
On Mon, Dec 12, 2005 at 11:51:24AM -0700, Nelson H. F. Beebe wrote:
Machinetype:Sun W40z (4 CPUs, 2400 MHz AMD64 Opteron, 8GB RAM);
GNU/Linux Red Hat Enterprise Linux AS 4
Remote gcc version: gcc (GCC) 3.3.5
Remote g++ version: g++ (GCC) 3.3.5
Configure environment:
On Tue, Dec 13, 2005 at 10:00:52PM -0500, George Sherwood wrote:
Description:
Let expression parsing appears to be broken in bash 3.1. When
using a simple expression such as let a=(5+3) I get this error:
bash: syntax error near unexpected token `('
seems to be most assignments that
On Wed, Dec 21, 2005 at 10:35:11AM -0600, Perry Smith wrote:
I'm compiling bash-3.1 using IBM's compiler. I'm on AIX 5.3. I did
just the normal ./configure followed by make. Everything is fine
until the final link. It fails because it can not find isnan.
On Thu, Dec 22, 2005 at 09:24:06AM -0500, Chet Ramey wrote:
Bash-Release: 3.1
Patch-ID: bash31-001
Bug-Description:
There are parsing problems with compound assignments in several contexts,
including as arguments to builtins like `local', `eval', and `let', and
as multiple assignments in
On Friday 13 January 2006 21:16, Chet Ramey wrote:
Someone who can reproduce this is going to have to gdb bash, attach to
the shell producing the bad behavior, and find out what _rl_screenwidth
and _rl_term_autowrap are set to.
[EMAIL PROTECTED] 0 ~ $ gdb bash
(no debugging symbols found)
On Friday 13 January 2006 22:02, Mike Frysinger wrote:
On Friday 13 January 2006 21:16, Chet Ramey wrote:
Someone who can reproduce this is going to have to gdb bash, attach to
the shell producing the bad behavior, and find out what _rl_screenwidth
and _rl_term_autowrap are set to.
[EMAIL
On Friday 13 January 2006 12:11, Tim Waugh wrote:
Description:
If bash has argv[0] as '-/bin/bash' it does not become a login
shell.
Repeat-By:
exec -l /bin/bash
shopt
Fix:
patch works for me using screen / login shells
thanks
-mike
On Wednesday 18 January 2006 18:34, Jeff Chua wrote:
GNU bash, version 3.1.5(1)-release
sh -c echo -n ok returns -n ok.
works correctly for me:
[EMAIL PROTECTED] 0 ~ $ sh -c echo -n ok
[EMAIL PROTECTED] 0 ~ $
-mike
___
Bug-bash mailing list
On Sunday 29 January 2006 17:25, Bob Proulx wrote:
The bash manual documents this as Patterns to be brace expanded take
the form of an optional PREAMBLE, followed by either a series of
comma-separated strings or a sequnce expression between a pair of
braces, followed by an optional POSTSCRIPT.
On Sunday 29 January 2006 19:23, William Park wrote:
Let's see...
a-{b{d,e}}-c
a-{bd,be}-c
i'm pretty sure the commas are consumed in the expansion
side note, this also fails:
$ echo {a}{b,c}
{a}{b,c}
-mike
___
Bug-bash mailing list
On Sunday 29 January 2006 20:08, William Park wrote:
On Sun, Jan 29, 2006 at 07:33:14PM -0500, Chris F.A. Johnson wrote:
On Sun, 29 Jan 2006, William Park wrote:
Let's see...
a-{b{d,e}}-c
a-{bd,be}-c
a-bd-c a-be-c
It looks okey, I think.
Except that b{d,e} expands
if i'm simply unaware of such bash features, please feel free to yell at me :)
when working on misc bash scripts, i semi-frequently find the need to do some
sort of string processing along the lines of say changing a string to all
upper case or lower case ... if 'tr' was a bash builtin, then
we've been using a patch in Gentoo for sometime which adds support
for /etc/inputrc as a fallback after $INPUTRC and ~/.inputrc ... i couldnt
seem to find anything in the archives where someone proposed this be added to
readline, but maybe i just missed it ?
-mike
On Saturday 04 February 2006 18:35, Dmitry V. Levin wrote:
I believe this idea comes from Debian and is used also in other distros
including ALT and Owl, see e.g.
http://cvsweb.openwall.com/cgi/cvsweb.cgi/Owl/packages/bash/readline-5.1-de
b-alt-inputrc.diff
Yes, it looks not yet submitted.
On Sunday 05 February 2006 00:41, Paul Jarc wrote:
[EMAIL PROTECTED] wrote:
[kaneda-ogorasek]~$if [ ! -a /etc/passwd ] ; then echo si ; fi
-a is used as a binary and operator here. The expressions ! and
/etc/passwd are nonempty strings, so both are considered true, and
the overall result
On Saturday 04 February 2006 18:35, Dmitry V. Levin wrote:
On Sat, Feb 04, 2006 at 03:27:19PM -0500, Mike Frysinger wrote:
we've been using a patch in Gentoo for sometime which adds support
for /etc/inputrc as a fallback after $INPUTRC and ~/.inputrc ... i
couldnt seem to find anything
attached is a patch to address these issues with rlfe:
- trying to build rlfe on uClibc fails because it lacks sys/stropts.h
- pty.c uses close() but doesnt include unistd.h
- rlfe.c uses OpenPTY() but doesnt include extern.h
- rlfe.c uses wait() but doesnt include sys/wait.h
- rlfe/configure
On Monday 13 February 2006 15:40, Chet Ramey wrote:
Yuri Karlsbrun wrote:
I build bash v3.1 for cross-target. I configured bash like this:
CC=my_target_gcc AR=my_target_ar RANLIB=my_target_ranlib
bash-3.1_src/configure/configure --host=my_target
--target=my_target
hmm ... seems my e-mail client crashed when i sent the last message ...
On Monday 13 February 2006 19:31, Yuri Karlsbrun wrote:
You just cannot execute cross-compiled code on native platform...
yes, i know that, that isnt what i asked you
Also mksignames utility is built using native headers
not sure if this is a bug or feature ... take this little snippet:
testit() {
local foo=$(false) ; echo $?
foo=$(false) ; echo $?
}
when we run the code, the output is:
0
1
rather than intuitive:
1
1
-mike
___
Bug-bash mailing list
not a regression as bash-2.05 / bash-3.0 also barf on this ... i imagine
someone has already filed this, but i couldnt seem to find it in the mailing
lists ...
foo=$( #'
echo hi)
now the neat thing is that with bash-3.1 (unlike older versions), the
open/close parens are detected properly when
On Friday 14 April 2006 22:52, Chet Ramey wrote:
Mike Frysinger wrote:
not a regression as bash-2.05 / bash-3.0 also barf on this ... i imagine
someone has already filed this, but i couldnt seem to find it in the
mailing lists ...
foo=$( #'
echo hi)
This has already been fixed
On Saturday 15 April 2006 19:55, Herculano Einloft wrote:
$ if $(echo string /dev/null); then echo true; fi
true
This should be a syntax error, since
$ if; then echo true; fi
bash: syntax error near unexpected token `;'
the first command statement is more like:
if :; then echo true; fi
to dig up this thread yet again :)
http://lists.gnu.org/archive/html/bug-bash/2006-03/msg6.html
a user reported this with bash-3.0.17 with a good test case:
http://bugs.gentoo.org/130955
and i'm able to reproduce this here by setting my PS1:
PS1='$(echo Strange \[\e[0;32m\]Prompt\[\e[0m\] )'
ignoring the fact that i can pass in variables to gawk using the '-v' option,
i'm wondering if this is a bug in how bash expands variables to pass to
programs ... i couldnt pick out anything under EXPANSION, but that's probably
just because i missed it ;)
take for example:
$ foo=a b c
$ gawk
On Wednesday 03 May 2006 21:42, Mike Frysinger wrote:
the proposed hack:
http://lists.gnu.org/archive/html/bug-bash/2006-03/msg5.html
seems to work for this test case ...
but then seems to break another one:
export LC_ALL=C
PS1='\[\e[0;33m\]\u\[\e[0m\] '
printf a foo.txt
cat foo.txt
On Thursday 04 May 2006 00:44, Paul Jarc wrote:
Mike Frysinger [EMAIL PROTECTED] wrote:
$ foo=a b c
$ gawk 'BEGIN {foo='${foo}'}'
gawk: BEGIN {foo=a
gawk:^ unterminated string
This is normal. man bash:
# Word Splitting
# The shell scans the results of parameter
On Thursday 04 May 2006 11:08, Mike Stroyan wrote:
A little more bash syntax can quote newlines for awk.
this is when you start using gawk -v foo=$foo ...
i was using gawk as an example of my variable expansion question, not as a way
to figure out how to pass a variable into gawk
-mike
On Thursday 04 May 2006 11:37, Paul Jarc wrote:
Mike Frysinger [EMAIL PROTECTED] wrote:
On Thursday 04 May 2006 00:44, Paul Jarc wrote:
What do you mean by fail? What do you want to happen in this case?
i meant gawk hates it ... not bash
Ok, and what about the second question
On Thursday 25 May 2006 18:56, Cai Qian wrote:
echo -e hello \004world | cat /dev/null
will print out nothing
It suggests that cat has not seen EOF (004) generated by echo.
you piped the output to /dev/null ... how does that suggest anything ?
looks to me like you're evaluating with the
On Thursday 13 July 2006 14:06, Cheltenham, Christopher J wrote:
What can I do for this error?
not that this has anything to do with bash, but you could try installing the
library bash is complaining about:
ld.so.1: bash: fatal: libiconv.so.2: open failed: No such file or
directory
-mike
this little bit of code doesnt work right:
foo() { echo ${1:-a{b,c}} ; }
$ foo
a{b,c}
$ foo 1
a}
tested with bash-3.1.17
-mike
pgp0Oi7rbI6UV.pgp
Description: PGP signature
___
Bug-bash mailing list
Bug-bash@gnu.org
On Wednesday 06 September 2006 05:04, Andreas Schwab wrote:
[EMAIL PROTECTED] (Paul Jarc) writes:
Mike Frysinger [EMAIL PROTECTED] wrote:
this little bit of code doesnt work right:
foo() { echo ${1:-a{b,c}} ; }
Brace expansion happens before parameter expansion (man bash,
EXPANSION
looks like the latest release is missing just 1 last thing in the ulimit
documentation with the new rlimit features ... patch attached
-mike
pgpYzDtWR3t3W.pgp
Description: PGP signature
add missing docs for -e and -r options
--- builtins/ulimit.def
+++ builtins/ulimit.def
@@ -24,7 +24,7 @@
On Friday 13 October 2006 10:56, Jim Gifford wrote:
Been trying to compile swig with the current bash 3.2 have ran into
several issues that I have been able to fix except for this one.
this has been reported already:
http://lists.gnu.org/archive/html/bug-bash/2006-10/msg00046.html
-mike
On Friday 13 October 2006 11:31, Chris Clayton wrote:
The configure script from kdelibs-3.5.5 can't be run under bash 3.2. It
runs fine under bash 3.1. When I run the script I get:
this has been reported already:
http://lists.gnu.org/archive/html/bug-bash/2006-10/msg00046.html
-mike
On Saturday 14 October 2006 15:25, Kyle Sallee wrote:
Execution of kdebase-3.5.5/configure bash errors.
Best guess is line 39698:
ac_cv_maxpathlen=`sed 's#KDE_HELLO ##' conftest.out`
Bash does not recognize the final backtick?
this has been reported already:
On Tuesday 03 April 2007, [EMAIL PROTECTED] wrote:
type -p CMD normally has a non-zero return status if CMD is not
found. But if CMD is an alias, the return status is zero even though
nothing is printed to stdout.
by definition, this is how -p is supposed to work ... my guess is you want
On Friday 13 April 2007, Chet Ramey wrote:
West Stephen-QSW000 wrote:
Configuration Information [Automatically generated, do not change]:
Machine: powerpc
OS: linux-gnuspe
Compiler: ppc_85xx-gcc
Compilation CFLAGS: -DPROGRAM='bash' -DCONF_HOSTTYPE='powerpc'
On Friday 20 July 2007, Barbarian Mama wrote:
how can know subnet address of the machine using bash scripting .. if i use
ifconfig it only displays ipaddress and subnet mask. I you have to
calculate subnet address then you have to do binary AND operation on
ipaddress AND subnet mask to know
On Friday 24 August 2007, Chet Ramey wrote:
READLINE PATCH REPORT
=
Readline-Release: 5.2
Patch-ID: readline52-007
is this one going to be released as a bash patch as well ? i see readline
patches 5/6 match a bash patch,
On Friday 24 August 2007, Eric Blake wrote:
According to Mike Frysinger on 8/24/2007 3:03 PM:
Readline-Release: 5.2
Patch-ID: readline52-007
is this one going to be released as a bash patch as well ? i see
readline patches 5/6 match a bash patch, but not this one ...
bash32-025 has
On Friday 24 August 2007, Chet Ramey wrote:
BASH PATCH REPORT
=
Bash-Release: 3.2
Patch-ID: bash32-020
Bug-Reported-by: Ian A Watson [EMAIL PROTECTED]
Bug-Reference-ID:
On Saturday 25 August 2007, Chet Ramey wrote:
Mike Frysinger wrote:
a side note ... if you change any of BASH_{ARGC,ARGV,LINENO,SOURCE}
before setting a readonly variable, bash will not spit out the error
message about the variable being readonly ...
(UID=1)
-bash: UID: readonly
the top level Makefile has a race condition in the readline subdir due to how
it's been architected ...
start with the bash deps:
$(Program): ... $(LIBDEP) ...
LIBDEP expands to:
LIBDEP = ... $(READLINE_DEP) $(HISTORY_DEP) ...
and the DEP vars expant to the LIBRARY vars:
$(READLINE_LIBRARY):
On Monday 27 August 2007, Christian Boon wrote:
i want to cross compile bash-3.2 for my pxa255 arm processor and its
working although i can't get job control working.
cross-compiling bash is known to be broken as it'll mix your host signal defs
into the target binary
-mike
signature.asc
On Tuesday 28 August 2007, Christian Boon wrote:
Christian Boon wrote:
Chet Ramey wrote:
Christian Boon wrote:
Hello,
i want to cross compile bash-3.2 for my pxa255 arm processor and its
working although i can't get job control working.
It tells you it won't be able to:
checking
On Tuesday 18 September 2007, Pratiksha Powar wrote:
There was no network problem, I could access this url through my web
browser.
It was instead a yum configuration problem.
There wasn't any proxy information in the yum.conf file.
I had to add the parameters: proxy, proxy_username and
On Wednesday 26 September 2007, Mike Frysinger wrote:
or perhaps i want to take an arg list, append a string, and run a command
on it ... but i cant pass it straight as it may be too large, so i need to
xargs it ... so i'd do something like:
echo ${@/%.moo/.foo$'\000'} | xargs -0 rm -f
On Monday 01 October 2007, retiredff wrote:
I have several functions in my /etc/profile (Mac OSX 10.4.9). I can use the
functions at the commandline, however inside of scripts I receive an error.
I'll use an example of a function I have called cecho that echo's a string
in a color that is
On Wednesday 06 February 2008, Alexander Renn wrote:
Now I'm having troubles on this version:
GNU bash, version 3.1.17(0)-release (i386-portbld-freebsd6.2)
this isnt the latest version available. please retest with the latest.
-mike
signature.asc
Description: This is a digitally signed
On Monday 11 February 2008, [EMAIL PROTECTED] wrote:
Your attention is drawn to the current thread, a small sample of which is:
Newsgroups: comp.unix.shell
Subject: Re: Dr. Evil typed sleep 666; rm -rf /
Date: 11 Feb 2008
bash is the only shell that will run rm if you interrupt
sleep with
the configure script goes through the effort of properly setting up
LDFLAGS_FOR_BUILD, but then the Makefile.in turns around and ignores it.
instead it incorrectly sets it to $(LDFLAGS). attached patch by Takashi
YOSHII should fix things up.
-mike
signature.asc
Description: This is a
On Tuesday 08 April 2008, Soumyadip Das Mahapatra wrote:
I (Soumyadip Das Mahapatra, India, Calcutta), being a frequent user of bash
shell, have got it lacks an advanced feature dynamic command completion. As
for example, latest ides ( e.g- Netbeans for Java) support dynamic
auto-completion.
On Thursday 22 May 2008, Brendan Oakley wrote:
With some effort and help, I have Bash 3.1 built and mostly working
under OS/2. This uses gcc 3.3.5 with kLIBC 0.6.3, rather than the
old EMX tools. 3.2 needs a bit more work, and some of the patches
might need some more polish. My questions:
On Thursday 22 May 2008, Bob Proulx wrote:
If you still want to modify the source that is okay. You are free to
do so. Grab the source and modify it as you desire. But in order to
get help from other people you would need to motivate them as to why
you are one of the 0.01% of the people
On Friday 23 May 2008, Brendan Oakley wrote:
I would like to see them included, so the crucial part of my question
is whether it would be proper to submit any incremental patches for
inclusion even while the port itself is, in some ways, incomplete; or
whether I should have everything right
On Wednesday 06 August 2008, kinda wrote:
hi friends.,
my ascii file is as shown
1 41.18324 117.2673 ***
2 41.18324 117.2799 12.78
3 41.18324 117.2925 ***
299859 37.65457 122.0885 16.8
299860 37.65457 122.1011 16.668
Configuration Information:
Machine: x86_64
OS: linux-gnu
Compiler: x86_64-pc-linux-gnu-gcc
Compilation CFLAGS: -DPROGRAM='bash' -DCONF_HOSTTYPE='x86_64' -
DCONF_OSTYPE='linux-gnu' -DCONF_MACHTYPE='x86_64-pc-linux-gnu' -
DCONF_VENDOR='pc' -DLOCALEDIR='/usr/share/locale' -DPACKAGE='bash' -DSHELL -
On Wednesday 11 February 2009 23:38:10 Rolf Brudeseth wrote:
I would like to propose a new command for bash:
ca [path]
It returns the canonical path based on the current working directory and
entered path.
If the current working directory has been traversed through a symbolic
link, then
On Sunday 15 February 2009 23:19:28 Jan Schampera wrote:
Mike Frysinger wrote:
there is any way to get the last file that created that is fomat is
*.sql
why not just use `ls` and one of its sort options ? the ls man page
documents how to sort by creation time
Without looking
On Sunday 15 February 2009 23:39:03 Paul Jarc wrote:
Mike Frysinger vap...@gentoo.org wrote:
the op wasnt asking for the time, they were asking for the last created
file. and the ls man page talks how to sort by ctime.
ctime is the time when the inode was last modified, not (necessarily
this issue was posted before against bash-3.2 and the sub-makefiles were
fixed, but for some reason the top level makefile in bash-4.0 is still broken.
this fix is of course from Takashi YOSHII.
-mike
--- bash/Makefile.in
+++ bash/Makefile.in
@@ -143,7 +143,7 @@
CCFLAGS_FOR_BUILD =
the internal_wstrmatch() prototype should be moved to lib/globl/strmatch.h as
more than just strmatch.c uses this function (smatch.c uses it as well).
-mike
move prototypes around so they show up when needed
--- a/lib/glob/strmatch.c
+++ b/lib/glob/strmatch.c
@@ -25,9 +25,6 @@
#include
old issue that wasnt resolved, and the fix in question still works:
http://lists.gnu.org/archive/html/bug-bash/2007-08/msg00081.html
-mike
signature.asc
Description: This is a digitally signed message part.
bash-4.0 mishandles this code while bash-3.2_p48 and earlier work fine:
echo $(echo \;)
-mike
signature.asc
Description: This is a digitally signed message part.
On Sunday 22 February 2009 16:49:01 Chet Ramey wrote:
Mike Frysinger wrote:
the internal_wstrmatch() prototype should be moved to
lib/globl/strmatch.h as more than just strmatch.c uses this function
(smatch.c uses it as well).
No. If you look at how smatch.c is compiled, you'll see
On Sunday 22 February 2009 16:16:57 Chet Ramey wrote:
Mike Frysinger wrote:
bash-4.0 mishandles this code while bash-3.2_p48 and earlier work fine:
echo $(echo \;)
Try the attached patch.
seems to work, thanks
-mike
signature.asc
Description: This is a digitally signed message part.
previous versions of bash would happily accept negative values ( treated as a
signed integer and masked with like 0xff), but it seems some changes related
to option parsing has broken that
$ f(){ return -1; }; f
-bash: return: -1: invalid option
return: usage: return [n]
POSIX states that the
On Monday 23 February 2009 00:25:57 Jon Seymour wrote:
On Mon, Feb 23, 2009 at 4:03 PM, Mike Frysinger vap...@gentoo.org wrote:
previous versions of bash would happily accept negative values ( treated
as a signed integer and masked with like 0xff), but it seems some changes
related
On Monday 23 February 2009 07:50:30 Eric Blake wrote:
According to Mike Frysinger on 2/22/2009 10:03 PM:
previous versions of bash would happily accept negative values ( treated
as a signed integer and masked with like 0xff), but it seems some changes
related to option parsing has broken
On Monday 23 February 2009 08:48:32 Chet Ramey wrote:
Mike Frysinger wrote:
previous versions of bash would happily accept negative values ( treated
as a signed integer and masked with like 0xff), but it seems some changes
related to option parsing has broken that
$ f(){ return -1; }; f
On Monday 23 February 2009 15:16:21 Chet Ramey wrote:
Mike Frysinger wrote:
Are filenames beginning with a `-' useless because `rm' interprets
them as option arguments when, for instance, they're generated by the
expansion of `*'? Is `rm' broken for interpreting them as options?
I mean
i was going through the new features list in NEWS to see what cool things are
in here and i saw the new checkjobs option. is it just me or does this cause
a non-login bash to crash at exit which leads to an infinite loop / cpu
churning ?
$ gdb bash
(gdb) r
vap...@vapier 0:0 bash-4.0 $ shopt
On Monday 23 February 2009 18:05:26 Chet Ramey wrote:
Mike Frysinger wrote:
i was going through the new features list in NEWS to see what cool things
are in here and i saw the new checkjobs option. is it just me or does
this cause a non-login bash to crash at exit which leads
On Monday 23 February 2009 23:00:31 Chet Ramey wrote:
Mike Frysinger wrote:
On Monday 23 February 2009 18:05:26 Chet Ramey wrote:
Mike Frysinger wrote:
i was going through the new features list in NEWS to see what cool
things are in here and i saw the new checkjobs option. is it just me
On Tuesday 24 February 2009 21:14:46 Vincent Lefevre wrote:
On 2009-02-24 19:32:05 -0500, Paul Jarc wrote:
Vincent Lefevre vinc...@vinc17.org wrote:
I get the following errors with bash 3.2.39 under Debian/unstable:
bash -c 'alias a=echo OK 2
a
/dev/null a'
bash: line
On Wednesday 25 February 2009 09:06:21 Roman Rakus wrote:
It will be useful to have bash-4.0-patches available as it was for past
versions. When are you planning to do that? Thanks.
ive been adding them to Gentoo as Chet posts them ...
http://sources.gentoo.org/app-shells/bash/files/bash-4.0-*
seems there's a way to get bash to report exit values greater than 255 ...
since it requires certain key presses, things in between ... means a key
press rather than typing literally ...
$ true
$ echo 'enter
ctrl+c
$ echo $?
128
$ echo 'enter
ctrl+c
$ echo 'enter
ctrl+c
$ echo $?
128
$ true
$
On Wednesday 25 February 2009 16:21:47 Chet Ramey wrote:
Yep, it's a bug. Try the attached patch; it works for me.
this introduces a bug of it's own though :/. you can no longer use
ctrl+c to escape from unbalanced quotes.
- type: echo '
- hit enter
- hit ctrl+c over and over
On Thursday 26 February 2009 03:25:50 Sven Mascheck wrote:
On Wed, Feb 25, 2009 at 01:20:50PM -0500, Mike Frysinger wrote:
seems there's a way to get bash to report exit values greater than 255
...
you will find the special error values
in shell.h, for instance
#define EX_SHERRBASE
the documentation shows that for here documents, the word must be right
after the operator (although it doesnt really spell it out). not sure if that
should be made explicit and to have bash reject it, or to fix up this issue so
it works again ...
at any rate, this style usage, while
On Thursday 26 February 2009 17:29:18 Chet Ramey wrote:
Mike Frysinger wrote:
the documentation shows that for here documents, the word must be
right after the operator (although it doesnt really spell it out). not
sure if that should be made explicit and to have bash reject
On Thursday 26 February 2009 17:36:12 lehe wrote:
I installed bashdb through Synaptic Package Manager of my Ubuntu 8.10 and
it was installed into /usr/share/bashdb/, /usr/bin/, /usr/lib/ etc. Not
sure these could be found by bash.
How to know if DEBUGGER_START_FILE is defined in pathnames.h
On Monday 02 March 2009 00:22:15 Ray Parrish wrote:
but it would be nice if the man pages would at least mention things like
this parameter has to be quoted to work or use a * on the end of the
path to activate the --recursive option. It took me hours to find that
out with the ls command, see
code that uses case statements in a subshell and then uses comments causes
bash to trigger a parsing error:
$ cat test.sh
echo $(case a in (a) echo ok ;; # comment
)
$ sh ./test.sh
./test.sh: line 1: unexpected EOF while looking for matching `)'
./test.sh: line 3: syntax error: unexpected end of
On Tuesday 03 March 2009 15:28:28 Chet Ramey wrote:
Mike Frysinger wrote:
code that uses case statements in a subshell and then uses comments
causes bash to trigger a parsing error:
$ cat test.sh
echo $(case a in (a) echo ok ;; # comment
)
$ sh ./test.sh
./test.sh: line 1
On Tuesday 03 March 2009 17:14:16 Chet Ramey wrote:
Mike Frysinger wrote:
On Tuesday 03 March 2009 15:28:28 Chet Ramey wrote:
Mike Frysinger wrote:
code that uses case statements in a subshell and then uses comments
causes bash to trigger a parsing error:
$ cat test.sh
echo $(case
we use extdebug in our scripts to get the extended BASH_ARGC and BASH_ARGV
info so that when an error does occur, we can construct and display a useful
backtrace of where things fell apart and how the script got there. this has
significantly cut down time otherwise spent on manual debugging.
On Tuesday 10 March 2009 07:42:29 Angel Tsankov wrote:
I want to pass the output from one command (e.g. find) to some other
command so that each path (output by the first command) gets into a
distinct positional parameter of the second command. How can I do this if
some paths contain spaces?
On Friday 13 March 2009 23:03:47 Chet Ramey wrote:
Cameron Pulsford wrote:
I am running bash 4.0.10 (installed through macports)
hw = Macbook 2,1 os = mac os x 10.5.6
Compiled with gcc 4.0.1
A lot of times after a long vim session when I return to bash, my cursor
is not on a new line,
On Tuesday 17 March 2009 05:36:05 ipif wrote:
Stephane Chazelas wrote:
2009-03-12, 06:47(-07), ipif:
[...]
sh-3.2# bash
bash: command_substitute: cannot duplicate pipe as fd 1: Bad file
descriptor
(currently I'm using bash as init)
[...]
In my opinion this problem might
http://sources.gentoo.org/viewcvs.py/*checkout*/portage/main/trunk/bin/isolated-
functions.sh?content-type=text%2Fplain
checkout the dump_trace() func at the top ... that might help. or it might
confuse. g'luck!
-mike
signature.asc
Description: This is a digitally signed message part.
On Thursday 02 April 2009 21:33:40 Brandon F wrote:
When I do traceroute in bash
traceroute is not part of bash
I am always getting
12-215-11-193.client.mchsi.com as the third or fourth site. I want to know
how to clear this from my route list. So it will bounce off of a differant
site.
On Monday 06 April 2009 21:24:35 jida...@jidanni.org wrote:
Gentlemen, -x's reporting should just pass the Chinese right back.
$ set -x; export LC_ALL=$LANG; echo 中文
+ export LC_ALL=zh_TW.UTF-8
+ LC_ALL=zh_TW.UTF-8
+ echo $'\344\270\255\346\226\207'
中文
i dont think that's UTF-8. unicode
On Tuesday 07 April 2009 15:10:16 Special Sauce wrote:
From: anton
To: bug-bash@gnu.org
Subject: Cursor starts inside prompt
Configuration Information [Automatically generated, do not change]:
Machine: x86_64
OS: linux-gnu
Compiler: gcc
Compilation CFLAGS: -DPROGRAM='bash'
1 - 100 of 302 matches
Mail list logo