Re: chmod

2006-05-29 Thread Jim Meyering
Vasek Potocek [EMAIL PROTECTED] wrote: I think I have found some bug in chmod. ... touch a mkdir b touch c cd b chmod a-x ../* makes a segmentation fault. The versions are: chmod (GNU coreutils) 5.93 Fedora Core 5 on a i686 (2.6.16-1.2111_FC5) glibc 2.4 Thank you for the report.

Re: OSF/4.0D strtold

2006-05-29 Thread Paul Eggert
* c-strtod.m4 (gl_C99_STRTOLD): Use a link test rather than a compile test, for Tru64 4.0D. Thanks; I installed that on both gnulib and coreutils trunk. Jim, OK if I put it into coreutils b5_9x? Yes. Thanks again. Done, and you're welcome.

Re: 5.96: chgrp/basic failure on OpenBSD

2006-05-29 Thread Paul Eggert
Ralf Wildenhues [EMAIL PROTECTED] writes: With 5.96, I get one test failure on i386-unknown-openbsd3.8. Possibly you're suffering from this problem, noted in tests/chgrp/basic: # The following no-change chgrp command is supposed to update f's ctime, # but on OpenBSD, it appears to be a

Re: 5.96: chgrp/basic failure on OpenBSD

2006-05-29 Thread Ralf Wildenhues
Hi Paul, * Paul Eggert wrote on Mon, May 29, 2006 at 09:54:04AM CEST: Ralf Wildenhues [EMAIL PROTECTED] writes: With 5.96, I get one test failure on i386-unknown-openbsd3.8. Possibly you're suffering from this problem, noted in tests/chgrp/basic: # The following no-change chgrp

Re: AIX grep limitation: coreutils build failure

2006-05-29 Thread Paul Eggert
Ralf Wildenhues [EMAIL PROTECTED] writes: I don't mind the plug, but unfortunately, things are not so simple: The `grep -c' apparently keeps going, and just omits counting the line that was too long (as it happens, the LTLIBOBJS and the LIBOBJS substitutions are in different conf$$subs.sed

Re: Missing variable quoting in m4 files

2006-05-29 Thread Roland Illig
Jim Meyering wrote: Paul Eggert [EMAIL PROTECTED] wrote: ... That patch masks the actual error, which is that the variable was used without being defined. The underlying bug was fixed in a different way, as described in

Question about md5sum

2006-05-29 Thread Arseny Solokha
Hello! I have a question about md5sum 5.2.1. For example, I'm in deirectory /home/arseny and I want to get hash of file foo.bar. I have to write md5sum /home/arseny/foo.bar. But I consider that it would be more natural to look file firstly in current directory. For example command would be cd

Question about md5sum

2006-05-29 Thread Arseny Solokha
Hello! I have a question about md5sum 5.2.1. For example, I'm in deirectory /home/arseny and I want to get hash of file foo.bar. I have to write md5sum /home/arseny/foo.bar. But I consider that it would be more natural to look file firstly in current directory. For example command would be cd

Re: Question about md5sum

2006-05-29 Thread Eric Blake
I have a question about md5sum 5.2.1. 5.2.1 is several years old; I would consider upgrading. The latest stable version is 5.96, and a 5.97 will probably arrive within the next month or so. For example, I'm in deirectory /home/arseny and I want to get hash of file foo.bar. I have to write

5.96: 3 test failures on HP-UX 11.23

2006-05-29 Thread Ralf Wildenhues
I get 3 failures on hppa2.0w-hp-hpux11.23: misc/close-stdout, touch/no-create-missing, and help-version (groups), see the verbose output below. Note that I also get the warning below. On ia64-hp-hpux11.23, the same tests fail. On hppa2.0w-hp-hpux11.11, I get no test failures, and no such

Re: AIX tests failures

2006-05-29 Thread Albert Chin
On Sun, May 28, 2006 at 05:52:45PM +0200, Ralf Wildenhues wrote: And yes, that happens with GCC as well. They simply forget the final `\0', I guess. As also mentioned here[2], the strndup declaration is #ifdef'ed away behind a _LINUX_SOURCE_COMPAT define. I guess Autoconf should have a

Re: AIX tests failures

2006-05-29 Thread Albert Chin
On Sun, May 28, 2006 at 07:58:24PM +0200, Jim Meyering wrote: I have to confess that I wonder if it's worth trying to work around bugs in AIX 4. Is it still officially supported? Is it used by many? I haven't had access to such a system for a few years now, and no one has been building

Re: Support bytesize comparison in sort

2006-05-29 Thread Mart Somermaa
What is the general consensus on adding the '--human-readable-bytesize' otpion to sort? The initial response seemed positive and I personally crave for the feature for 'du -hs | sort -h'. ___ Bug-coreutils mailing list Bug-coreutils@gnu.org

Re: coreutils-5.96 doesn't build (out of the box)

2006-05-29 Thread Sam Sirlin
Date: Sun, 28 May 2006 10:55:23 +0200 From: Ralf Wildenhues [EMAIL PROTECTED] * Sam Sirlin wrote on Sun, May 28, 2006 at 01:20:04AM CEST: Here's the full output of the bad configure (autoconf 2.59 produced a working configure) I still have no idea what's going on in

Re: coreutils-5.96 doesn't build (out of the box)

2006-05-29 Thread Ralf Wildenhues
* Sam Sirlin wrote on Mon, May 29, 2006 at 10:41:03PM CEST: From: Ralf Wildenhues [EMAIL PROTECTED] /bin/bash --version echo $PATH type mkdir expr dirname basename dirname is apparently the issue. I had an old dirname shell script lying around from 1989 in my path