Date: Tue, 15 Feb 2000 21:13:22 +1100 (EST)
From: Ben Elliston [EMAIL PROTECTED]
cc: [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED],
[EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED]
+uname -m = `(uname -m) 2/dev/null || echo unknown`
+uname -r =
Date: Sun, 27 Feb 2000 02:27:56 -0800 (PST)
From: Martin Buchholz [EMAIL PROTECTED]
I know of no way to quote a non-paired quote character
In the current autoconf draft, @BKL@ and @BKR@ expand to left and
right brackets; that should get you want you want. (Those names
should get
From: Akim Demaille [EMAIL PROTECTED]
Date: 28 Feb 2000 09:16:00 +0100
I'm very OK with changing them. I buy any suggestion of names.
Currently there are four guys:
s/@BKL@/[/g
s/@BKR@/]/g
s/@DLR@/$/g
s/@PND@/#/g
Here's an idea: why don't we
From: Akim Demaille [EMAIL PROTECTED]
Date: 28 Feb 2000 18:16:13 +0100
Paul s/@:@/[/g s/@:@/]/g s/@S|@/$/g s/@%:@/#/g
Can we consider these guys are unlikely enough to avoid using @@?
No. E.g. grep's configure.in says:
LIBOBJS=`echo $LIBOBJS|sed 's/\.o /\$U.o
From: Akim Demaille [EMAIL PROTECTED]
Date: 27 Mar 2000 17:25:20 +0200
my proposal is to support `mktemp -d', and only `mktemp -d', no
juggling with mktemp for each file.
If the system on which Autoconf runs does not support `mktemp -d',
then (umask 077 mkdir $$).
That
From: Akim Demaille [EMAIL PROTECTED]
Date: 26 Apr 2000 18:41:32 +0200
If you concentrate the measure on this very script, the performance
penalty is frightening:
It certainly is. This is a performance bug in mawk.
I observed the bug in mawk 1.3.3.
To work around the problem
sts. Here is a patch; it uses _GNU_SOURCE instead of
_XOPEN_SOURCE since I think that is better from an autoconf point of
view. I'm CC'ing this to the maintainers of autoconf and fileutils as
they are using this code too.
2000-05-02 Paul Eggert [EMAIL PROTECTED]
* m4/largefile.m4 (AC_SYS
like to fix largefile.m4 so that it didn't depend on $host_os
and AC_CANONICAL_HOST, but that's a larger project
2000-05-03 Paul Eggert [EMAIL PROTECTED]
* m4/largefile.m4 (AC_SYS_LARGEFILE): Define _XOPEN_SOURCE to
be 500, instead of _GNU_SOURCE to be 1, to work around glibc
From: Per Bothner [EMAIL PROTECTED]
Date: 08 May 2000 06:53:13 -0700
I do not see 4.3BSD as a reason to not use functions in shell scripts.
I can see good arguments on both sides of this question. A similar
question is whether an application should work with KR C.
It used to be that
From: Akim Demaille [EMAIL PROTECTED]
Date: 22 May 2000 10:34:41 +0200
I am not against giving a chance to `[!...]', but it has to be
heavily tested. Is there anybody who knows more?
I believe negated character ranges were added to /bin/sh around the
same time as shell functions.
From: Akim Demaille [EMAIL PROTECTED]
Date: 23 May 2000 17:15:16 +0200
expr has a bad reputation
Its syntax is awkward -- was that what you were thinking about? But I
haven't had portability problems with it, so I don't know what you
want to document. Perhaps this is because I
From: Akim Demaille [EMAIL PROTECTED]
Date: 24 May 2000 10:41:53 +0200
Akimexpr has a bad reputation
Paul Its syntax is awkward -- was that what you were thinking about?
Nope, I was referring to the portability. And precisely about the
syntax, is `match string expr'
If the file's size is sufficiently small and if the locale is C (which
is the case here), you should be able to compute the file's size with
this portable script:
set X `ls -l "$file"`
case "$6" in
[0-9]*)
size=$6;;
*)
# We are on a non-POSIX host and the group was omitted,
# or
On Sun, Jun 18, 2000 at 01:44:14PM -0700, Paul Eggert wrote:
set X `ls -l "$file"`
case "$6" in
[0-9]*)
size=$6;;
*)
# We are on a non-POSIX host and the group was omitted,
# or the user overflowed into the group.
size=$5;;
esac
From: Bernard Dautrevaux [EMAIL PROTECTED]
Date: Thu, 6 Jul 2000 17:26:48 +0200
under UN*X, '//1' should give '/',
No, under some older flavors of Unix, leading // is a special path
name: it refers to a "super-root" and is used to access other
machines' files. Leading ///, , etc.
From: Akim Demaille [EMAIL PROTECTED]
Date: 24 Jul 2000 10:56:01 +0200
What would be the problem with enabling the use of the fourth
parameter of AC_CHECK_FUNCS? AC_CHECK_FUNC(func, yes, no, includes)?
That sounds right to me. I didn't quite follow your patch, but I
assume the
From: Greg McGary [EMAIL PROTECTED]
Date: 24 Jul 2000 02:18:59 -0700
Also, surely trying twice for each function won't work in general.
For example, suppose a library function is available only in
non-bounded-pointers mode, and we are building in bounded-pointers
mode.
From: Akim Demaille [EMAIL PROTECTED]
Date: 25 Jul 2000 09:41:52 +0200
Paul, may I ask you what you think about this?
It should be OK to set LC_COLLATE to "C", and also all the other LC_*
variables: LC_CTYPE, LC_TIME, LC_NUMERIC, LC_MONETARY, and
LC_MESSAGES.
It shouldn't be necessary
From: Alexandre Oliva [EMAIL PROTECTED]
Date: 27 Jul 2000 14:16:14 -0300
On Jul 24, 2000, Akim Demaille [EMAIL PROTECTED] wrote:
What would be the problem with enabling the use of the fourth
parameter of AC_CHECK_FUNCS? AC_CHECK_FUNC(func, yes, no,
includes)?
Even
From: Alexandre Oliva [EMAIL PROTECTED]
Date: 27 Jul 2000 21:50:47 -0300
the proposed changes don't make autoconf unusable for pre-C99.
Not unusable, but it will certainly break some code.
autoconf's behavior wouldn't change for existing autoconf input files.
So I don't see how it
From: Akim Demaille [EMAIL PROTECTED]
Date: 31 Jul 2000 12:14:49 +0200
my proposal always defines _GNU_SOURCE and _ALL_SOURCE
I suggest always define __EXTENSIONS__ too. Solaris uses that symbol
to define all extensions that are not incompatible with whatever
standard is in use.
From: Simon Josefsson [EMAIL PROTECTED]
Date: 01 Aug 2000 17:54:30 +0200
Now, as it happens, sometimes CC running in AFS (a distributed
filesystem) have problem with the locking mechanism for this file
and hang indefinitely saying:
SunWS_cache: Information: Database is locked,
From: Akim Demaille [EMAIL PROTECTED]
Date: 01 Aug 2000 16:15:22 +0200
Here is a satisfying starting point? OK to commit?
Not yet, unfortunately, and partly this is due to inadequate research
on my part; please see below. (Sorry about that.)
First, some preliminary comments:
-
Date: Fri, 11 Aug 2000 12:30:27 -0400 (EDT)
From: "John David Anglin" [EMAIL PROTECTED]
Possibly `(exit $?); exit' could be simplified to just `exit'.
Yes, that's correct. Sorry, I missed that in my earlier scan.
From: Jim Meyering [EMAIL PROTECTED]
Date: 11 Aug 2000 20:11:25 +0200
I saw the earlier thread (on the autoconf list) on this topic where
someone found that `?' and \| were not portable. Has anyone found a
system on which \{m,n\} is not portable?
Most likely, seds that don't
Date: Sun, 13 Aug 2000 10:30:21 +0300 (IDT)
From: Eli Zaretskii [EMAIL PROTECTED]
I don't see any work-arounds suggested for these problems.
Does anyone know how do you express \(foo\)\{0,1\} portably?
There's no simple, general substitute for \(foo\)\{0,1\}, but you can
generally
Date: Tue, 29 Aug 2000 10:15:13 -0400 (EDT)
From: Pavel Roskin [EMAIL PROTECTED]
grep-2.4.2 is sufficient to reproduce the problem. No need to upgrade libc
- glibc-2.1.1 is fine.
$ grep --version
grep (GNU grep) 2.4.2
[Copyright etc skipped]
$ echo foo infile
$ grep
Date: Tue, 29 Aug 2000 13:27:10 -0400 (EDT)
From: "Thomas E. Dickey" [EMAIL PROTECTED]
unless I missed a response, so far all that's been demonstrated is that a
newer version of GNU grep doesn't behave the same as other versions of
grep.
I don't even see where that has been
Date: Tue, 29 Aug 2000 14:20:12 -0400 (EDT)
From: Pavel Roskin [EMAIL PROTECTED]
I haven't checked the older versions, but this behaviour of GNU grep is
weird (it may or may not be a bug, dependent on the standard):
$ echo foo |./grep -E ' *+'
foo
It is not a bug. GNU grep
Date: Tue, 29 Aug 2000 19:07:03 -0400
From: Thomas Dickey [EMAIL PROTECTED]
It is not a bug. GNU grep extends the semantics of regular
expressions so that 'x*+' is equivalent to '(x*)+'. POSIX does not
I wouldn't call it an extension, since it breaks some existing scripts
Date: Thu, 31 Aug 2000 14:41:48 -0400 (EDT)
From: Pavel Roskin [EMAIL PROTECTED]
If Paul said it's
ok to assume `+' is not a meta-character in `grep', I believe it.
You shouldn't believe it. The answer is Plan 9.
OK, but Plan 9 grep is not the same as traditional grep at all.
Date: Thu, 31 Aug 2000 14:41:48 -0400 (EDT)
From: Pavel Roskin [EMAIL PROTECTED]
If Paul said it's
ok to assume `+' is not a meta-character in `grep', I believe it.
You shouldn't believe it. The answer is Plan 9.
OK, that's news to me. But you should be careful: Plan 9 grep
Russ Allbery [EMAIL PROTECTED] writes:
I see software that uses ANSI C function prototypes but tests for
varargs.h vs. stdarg.h and chooses the right variadic argument syntax
accordingly. Are there really ANSI C compilers that don't have stdarg.h?
No, but there were compilers that had
From: Akim Demaille [EMAIL PROTECTED]
User-Agent: Gnus/5.0807 (Gnus v5.8.7) XEmacs/21.1 (Channel Islands)
The failure is extremely famous, and there is an incredible amount of
news in gnu.bug.utils about this but in the context of grep. It is
simply that some locales have more
From: Akim Demaille [EMAIL PROTECTED]
Date: 25 Oct 2000 12:14:12 +0200
/tmp % echo "Xfoo" | nostromo 12:08
pipe sed '/^X\(.*[^/]\)\/\/*[^/][^/]*\/*$/{ s//\1/p; q; }
pipe quote /^X\(\/\/\)[^/].*/{ s//\1/p; q; }
pipe quote
From: David Morgan [EMAIL PROTECTED]
Date: Thu, 09 Nov 2000 08:12:01 -0800
| expr 'a' : '\(b\)' echo failure
|
| Gives:
| 0
Result is wrong, $? right.
Wrong explanation - no characters match so the output of 0 is correct
(At least I think so)
The POSIX standard says:
Date: Thu, 09 Nov 2000 15:24:07 -0800
From: David Morgan [EMAIL PROTECTED]
Sorry, I've lost context. Why is this a problem? If expr exits with
nonzero status, then AS_DIRNAME will fall back on sed, right?
It will use the output of the expr and then append the output of the
From: Akim Demaille [EMAIL PROTECTED]
Date: 10 Nov 2000 10:23:25 +0100
But our troubles could be worse than this, since `expr' is not used
only in AS_DIRNAME. For instance any CVS configure contains
ac_optarg=`expr "x$ac_option" : 'x[^=]*=\(.*\)'`
Ouch. Can you rewrite that as
From: Akim Demaille [EMAIL PROTECTED]
Date: 16 Nov 2000 10:45:57 +0100
We do agree `/foo[\/]bar/' is a single pattern which matches the
string `foo/bar', right? Then, why do you use the char class here?
Because, if my admittedly fallible memory serves, some older hosts
mishandle
From: Akim Demaille [EMAIL PROTECTED]
User-Agent: Gnus/5.0807 (Gnus v5.8.7) XEmacs/21.1 (Channel Islands)
But the question remains open :)
I certainly recall reports of hosts where fopen (..."wb") did not work.
However, I don't recall which hosts they were.
I briefly looked around the net
From: Akim Demaille [EMAIL PROTECTED]
Date: 29 Nov 2000 12:29:09 +0100
it's not clear to me whether there is a difference between the
empty as a literal or as the result of an evaluation.
I don't see any difference in the standard's wording. They should
be treated the same.
could you
From: Akim Demaille [EMAIL PROTECTED]
Date: 30 Nov 2000 09:48:35 +0100
Pavel I don't know how many times we rely on similar
Pavel constructs. There may be more serious problems. To fix them, we
Pavel should never use $? after _any_ assignments.
That's really bad in theory, I hope its
From: Matthew Whitworth [EMAIL PROTECTED]
Date: Tue, 5 Dec 2000 14:45:19 -0800 (PST)
I have examined several packages that use getopt_long(), including
several GNU packages, and none of them autoconfiscate this function
the same way. Actually, few of them even autoconfiscate it at all
--
From: Akim Demaille [EMAIL PROTECTED]
User-Agent: Gnus/5.0807 (Gnus v5.8.7) XEmacs/21.1 (Channel Islands)
| nawk: Tempcell list is curdled
| nawk: Source line number 7
The stock nawk on Debian unstable works fine with this, and there is
not `curdled' in the man page. Maybe the author can
From: Akim Demaille [EMAIL PROTECTED]
Date: 08 Dec 2000 14:34:35 +0100
| #include assert.h
| /* Override any gcc2 internal prototype to avoid an error. */
| #ifdef __cplusplus
| extern "C"
| #endif
| /* We use char because int might match the return type of a gcc2
|builtin and
From: Lars Hecking [EMAIL PROTECTED]
Date: Thu, 18 Jan 2001 11:06:40 +
I need to implement a configure check for waitpid(). While this by
itself is trivial, are there any known abnormalities/non-POSIXisms
I need to consider?
Not as far as I know, if you confine yourself to the
From: Assar Westerlund [EMAIL PROTECTED]
Date: 06 Mar 2001 17:51:45 +0100
If you're writing code that
depends on this and that has to be able to run on systems where you
cannot guarantee that rename isn't atomic, atomicity has to be
accomplished in some other way.
Agreed, but I've been
From: Alexandre Oliva [EMAIL PROTECTED]
Date: 10 Apr 2001 08:49:15 -0300
I hate Perl. Really. I mean it.
I wouldn't go that far. But perhaps I'm biased:
* I worked down the hall from Larry Wall when he invented Perl.
* In 1993 Stott Parker and I published one of the first academic
From: "Tim Van Holder" [EMAIL PROTECTED]
Importance: Normal
Date: Tue, 10 Apr 2001 19:17:17 +0200
Anyways, I've tried a slightly changed format:
$ ./config.status --version
GNU Autoconf config.status 2.49e
configured by ../autoconf/configure, generated by GNU Autoconf 2.49e,
Cc: [EMAIL PROTECTED]
From: Akim Demaille [EMAIL PROTECTED]
Date: 10 Apr 2001 19:03:14 +0200
User-Agent: Gnus/5.0808 (Gnus v5.8.8) XEmacs/21.1 (Cuyahoga Valley)
"Paul" == Paul Eggert [EMAIL PROTECTED] writes:
Paul Scheme by far is the best choice for this kind of applicat
From: Bob Proulx [EMAIL PROTECTED]
Date: Tue, 10 Apr 2001 16:22:01 -0600
Python causes more polarization than any other language.
Clearly you haven't been in the trenches of past language wars. :-)
But if the principal maintainer prefers Perl, that settles the matter.
From: Ben Pfaff [EMAIL PROTECTED]
Date: 20 Apr 2001 11:21:01 -0400
How about jigsaw puzzle pieces that fit together. (Has that been
used elsewhere?)
Yes, but we could do it better.
Here's a suggestion: take the logos from the Plan of St Gall.
This is a famous architectural diagram from
From: Harlan Stenn [EMAIL PROTECTED]
Date: Sun, 13 May 2001 14:05:36 -0400
What version(s) of automake should be used with this candidate?
I have been using recent CVS snapshots of automake which I got from
http://sources.redhat.com/automake/. Currently I'm using a snapshot
that I took
From: Keith Bostic [EMAIL PROTECTED]
Date: Tue, 22 May 2001 11:00:31 -0400 (EDT)
A feature I would like to see in autoconf is a way to find out
how to typedef to a specific size on a system.
I would say that any such scheme should mimic the C99 size scheme, as
it can be implemented on
From: Akim Demaille [EMAIL PROTECTED]
User-Agent: Gnus/5.0808 (Gnus v5.8.8) XEmacs/21.1 (Cuyahoga Valley)
I'm for config/.
No name is likely to achieve universal consensus, but config works
for me.
From: Russ Allbery [EMAIL PROTECTED]
Date: 28 May 2001 11:05:08 -0700
Lars Hecking [EMAIL PROTECTED] writes:
What's the rationale behind using inttypes.h, and not stdint.h?
inttypes.h is more widespread on pre-C99 systems, particularly Solaris,
since IIRC it was present in an
From: Lars Hecking [EMAIL PROTECTED]
Date: Tue, 29 May 2001 10:33:14 +0100
I'm unfamiliar with the C99 Standard, but did it adopt sys/types.h,
No.
or define any relationship between sys/types.h and stdint.h/inttypes.h?
Only indirectly. E.g. POSIX says sys/types.h defines size_t, and
From: Akim Demaille [EMAIL PROTECTED]
Date: 31 May 2001 09:32:23 +0200
* acgeneral.m4 (_AC_INCLUDES_DEFAULT_REQUIREMENTS): Include
stdlib.h.
From Paul Eggert and Lars Hecking.
The change itself looks good as far as it goes, but:
* That ChangeLog entry has a typo
both their time
and yours.
So, how about this documentation patch to address the matter? It
implements the above suggestion, along with mentioning the problems
with echo a bit more clearly (and fixing the manual to follow its
own new advice about echo).
2001-06-01 Paul Eggert [EMAIL PROTECTED
From: Akim Demaille [EMAIL PROTECTED]
Date: 11 Jun 2001 20:36:50 +0200
Russ If it isn't, autoconf may need to check explicitly if
Russ inttypes.h and sys/types.h can be included at the same time,
Russ a la the existing checks for time.h and sys/time.h.
Akim Agreed.
Akim Paul, Jim? On
From: Akim Demaille [EMAIL PROTECTED]
Date: 11 Jun 2001 23:07:29 +0200
I wanted to have some feedback from Jim and Paul... If you two guys
don't have time to give a look at it, than I'll apply it on Wednesday.
I read through http://sources.redhat.com/ml/autoconf/2001-06/msg9.html
and
From: Eric Siegerman [EMAIL PROTECTED]
Date: Mon, 11 Jun 2001 21:19:22 -0400
Paul Eggert wrote:
the Unix tradition is that code that uses 'vfork' can be safely
changed to use 'fork' if you like.
If this is a tradition, it's a bad one. Recall that with vfork,
the child's memory
From: Akim Demaille [EMAIL PROTECTED]
Date: 12 Jun 2001 09:56:38 +0200
So, are you actually saying that *any* #include foo.h should be
checked?
No, just if there's a good reason for it. For example, AC_HEADER_STAT
need not bother to wrap the '#include sys/types.h' inside '#if
From: Russ Allbery [EMAIL PROTECTED]
Date: 12 Jun 2001 05:58:31 -0700
I've yet to see *any* system that didn't have sys/types.h.
I found 'AC_CHECK_HEADERS(sys/types.h)' in the configure.in files for
a2ps, kaffe, and rpm. So it's plausible that some people have found
such systems. (It
From: Akim Demaille [EMAIL PROTECTED]
Date: 12 Jun 2001 15:37:47 +0200
Well, too late, here is a proto monster patch.
Unfortunately, if you wrap every include of sys/types.h, even for
checks that apply only to POSIX-like systems, then you'll end up
raising other questions. For example, in
From: Eric Siegerman [EMAIL PROTECTED]
Date: Tue, 12 Jun 2001 15:02:58 -0400
you can get away with:
#define a b
(or the equivalent using true functions) only if a()'s behaviour
is a strict subset of b()'s. That's not the case with
fork/vfork, in either direction.
But it is
From: R. Kuhlmann [EMAIL PROTECTED]
Date: Tue, 12 Jun 2001 17:01:08 +0200
#define vfork to fork if necessary. to vfork if necessary.
There's an echo in that part of the ChangeLog...
In the light of the previous message, it seems to me that the patch
should not alter
From: [EMAIL PROTECTED]
Date: Wed, 13 Jun 2001 00:07:55 +0200
I don't think the uses are portable.
Wait, why not?
The latest POSIX draft says this:
The vfork() function shall be equivalent to fork(), except that
the behavior is undefined if the process created by vfork()
From: [EMAIL PROTECTED]
Date: Wed, 13 Jun 2001 11:52:36 +0200
I don't see why the AC_FUNC_WAIT3 test relies on the parent running at the
same time
You're right; I got confused by the 'sleep' in the child, and assumed
that it was checking for a race condition. But on second thought that
From: Paul D. Smith [EMAIL PROTECTED]
Date: 13 Jun 2001 15:03:41 -0400
pe However, this point is somewhat academic. Nobody uses wait3
pe anymore
GNU make uses wait3() if waitpid() doesn't exist, as well.
Sorry, I missed that, because I was looking only for instances of
From: Eric Siegerman [EMAIL PROTECTED]
Date: Wed, 13 Jun 2001 17:29:52 -0400
On Tue, Jun 12, 2001 at 07:42:47PM -0700, Paul Eggert wrote:
From: [EMAIL PROTECTED]
Date: Wed, 13 Jun 2001 00:07:55 +0200
* fork(), that should be a fork() if possible, but if not available,
a vfork
Date: Fri, 15 Jun 2001 15:13:49 -0400 (EDT)
From: Pavel Roskin [EMAIL PROTECTED]
Either I'm missing something obvious or the current scheme with AU_DEFUN
should be reworked.
You're not missing anything obvious that I can see, unfortunately.
+If @file{vfork.h} is found, define @code{HAVE_VFORK_H}. If a working
+@code{vfork} is found, define @code{HAVE_WORKING_VFORK}. Otherwise,
+define @code{vfork} to be @code{fork} for backward compatibility.
You need to append with previous versions of @command{autoconf} here.
+In case you
From: Alexandre Oliva [EMAIL PROTECTED]
Date: 17 Jun 2001 19:20:13 -0300
On Jun 11, 2001, Akim Demaille [EMAIL PROTECTED] wrote:
It's very easy to set a guard against such compilers. But once we
detected the compiler cannot evaluate such expressions, what do we do?
Would it work
That looks good to me.
From: Matt Watson [EMAIL PROTECTED]
Date: Mon, 25 Jun 2001 14:55:26 -0700
My vote too. How often does configure come across a system which doesn't
have a macro defining the byte order?
Fairly often. Solaris 8 has two macros (_BIG_ENDIAN and
_LITTLE_ENDIAN), but autoconf 2.50 doesn't know
From: Akim Demaille [EMAIL PROTECTED]
Date: 23 Jun 2001 17:51:54 +0200
Paul, this patch was sent later, and you did not comment it.
Sorry, I didn't see it. I don't think that patch is necessary.
Here are some more detailed comments:
+ AC_DEFINE(ac_vfork, vfork, [Define to `vfork' if it
From: Andrej Borsenkow [EMAIL PROTECTED]
Date: Thu, 28 Jun 2001 13:58:35 +0400
It checks only for specific option for IRIX and _FILE_OFFSET_BITS and
_LARGE_FILES macros. That is not enough; e.g. on our system:
bor@itsrm2% getconf LFS_CFLAGS
-D_FILE_OFFSET_BITS=64 -D_LONGLONG -Kll64
From: Dale E Martin [EMAIL PROTECTED]
Date: Thu, 5 Jul 2001 14:21:36 -0400
Would it make sense for autoconf to have a macro to define all of
the available native types for the language it supports? Or is this
a fairly unique bit of information to require?
In C99, the right way to see
From: Ed L Cashin [EMAIL PROTECTED]
Date: 04 Jul 2001 18:57:13 -0400
Ed I've found that saying test -r is not reliable on Solaris.
Huh?
Sorry, I think I was confusing test -e with test -r. I'll let you
know if I learn more.
Yes, test -e does not work with Solaris 8 /bin/sh. This
From: Paul J. Menchini [EMAIL PROTECTED]
Date: Fri, 20 Jul 2001 12:50:26 -0400 (EDT)
I discovered that every non-directory file in
/ was gone; in particular, the symbolic links /bin, /lib, and /sys (in
addition to /tmp and some local symbolic links) and the files /boot,
/kadb and /vmunix
Martin == Martin Baulig [EMAIL PROTECTED] writes:
Martin Everything works just fine there except that I
Martin needed to make changes which won't work with autoconf 2.1x
Martin anymore.
Could you please be more specific here? Perhaps we can suggest ways
for you to make the changes so that
From: Adam J. Richter [EMAIL PROTECTED]
Date: Mon, 30 Jul 2001 00:21:43 -0700
Am I duplicating someone else's effort?
Not as far as I know.
If not, does this look good? Should I develop a patch?
It looks good to me; perhaps other people on the autoconf list (more
expert than I)
' as root but the shell-script as
non-root? That might explain matters if the source is mounted via NFS
without root-over-the-wire privileges.
It does seem to me that 'configure' should exit if this error is
found, as it's a serious one. Here's a proposed (and untested) patch.
2001-07-31 Paul Eggert
From: Paul J. Menchini [EMAIL PROTECTED]
Date: Wed, 1 Aug 2001 14:05:55 -0400 (EDT)
From: Paul J. Menchini [EMAIL PROTECTED]
Date: Fri, 20 Jul 2001 12:50:26 -0400 (EDT)
I discovered that every non-directory file in
/ was gone; in particular, the symbolic links /bin, /lib, and /sys
From: Akim Demaille [EMAIL PROTECTED]
Date: 04 Aug 2001 17:04:17 +0200
| * m4/init.m4 (AM_INIT_AUTOMAKE): Likewise.
| * m4/missing.m4 (AM_MISSING_HAS_RUN, AM_AUX_DIR_EXPAND): Likewise.
These are from Automake. Is it an update of Autoconf's copy, or a
patch to forward to
Date: Mon, 13 Aug 2001 10:19:19 -0600 (MDT)
From: Richard Stallman [EMAIL PROTECTED]
+The @file{INSTALL} file is free documentation; the Free Software
+Foundation gives unlimited permission to copy, distribute and modify it.
There are two possibilities. If the file is big enough, it
From: Akim Demaille [EMAIL PROTECTED]
Date: 13 Aug 2001 15:06:03 +0200
Paul The biggest pain of generating this patch was coming up with the
Paul dates, i.e. which years to put into the copyright notices. I
Paul did this by inspecting the CVS history.
My guess is that you wrote a tool
this resolves the problem for 'INSTALL'.
2001-08-14 Paul Eggert [EMAIL PROTECTED]
* doc/autoconf.texi, doc/install.texi: Add a copyright notice
to the INSTALL file.
--- doc/autoconf.texi-1.498.0.1 Sun Aug 12 04:34:23 2001
+++ doc/autoconf.texi Wed Aug 15 01:13:38 2001
-- I
can help with a patch if you like.)
standards.texi is next on my list of things to do. As far as I know,
that's the last file that is missing a proper copyright notice in the
autoconf distribution.
2001-08-14 Paul Eggert [EMAIL PROTECTED]
* make-stds.texi: Add a copyright notice
From: Richard Stallman [EMAIL PROTECTED]
CC: [EMAIL PROTECTED]
Date: Wed, 15 Aug 2001 17:07:12 -0600 (MDT)
The copyright notice and permission notice should not go in a separate
section. They should go in the standard place, on the page after the
title page if there are such things, or
From: Akim Demaille [EMAIL PROTECTED]
User-Agent: Gnus/5.0808 (Gnus v5.8.8) XEmacs/21.4 (Artificial Intelligence)
OTOH, is it really the purpose of the GCS to include such
information? That pertains to autoconf.texi IMHO.
The GCS is talking about a more general issue: whether to use `#if'
From: Richard Stallman [EMAIL PROTECTED]
Date: Thu, 16 Aug 2001 15:30:47 -0600 (MDT)
Please put the copyright notice and permission notice at the top of
the file.
OK, I moved them to the top of the INSTALL file.
PS to autoconf: I finally got remote CVS working from my home machine
to the
Thanks for the suggestion. How about the following patch instead?
I tried to tighten up the wording a bit.
2001-08-16 Paul Eggert [EMAIL PROTECTED]
* autoconf.texi (Function Portability): Add snprintf.
This rewords a suggestion by Kevin Ryde.
ndex: autoconf.texi
From: Akim Demaille [EMAIL PROTECTED]
User-Agent: Gnus/5.0808 (Gnus v5.8.8) XEmacs/21.4 (Artificial Intelligence)
Currently I see two means to have them: introduce quadrigraphs for
space and tab, or/and introduce a quadrigraph marking the eol, but
which is to be removed in the actual
in the neighborhood (e.g. I would hardly call
snprintf Classical -- perhaps because I've been around too long :-).
Thanks again for the suggestion.
2001-08-17 Paul Eggert [EMAIL PROTECTED]
* doc/autoconf.texi (Function Portability): Mention snprintf,
following up on a suggestion
From: Paul Eggert [EMAIL PROTECTED]
Date: Tue, 21 Aug 2001 21:08:39 -0700 (PDT)
I don't fully understand all the issues involved with line smashing,
but I tend to agree that it's relatively important to remove multiple
Sorry, I meant to write relatively *un*important.
blank lines from
From: Akim Demaille [EMAIL PROTECTED]
Date: 27 Aug 2001 13:23:47 +0200
is special: I've been bitten by this AWK program:
gsub (@comment.*, @t@, line);
which of course had the result of prefixing with @, suffixing with @.
I backlashed it and it works, but do we want to keep sth
From: Akim Demaille [EMAIL PROTECTED]
Date: 29 Aug 2001 13:09:46 +0200
Paul The is my own invention, but the t came from the source
Paul code of the ALGOL68C compiler, written by Steve Bourne (of
Paul Bourne shell fame),
Is there is any place where I could actually make my education
From: Paul Townsend [EMAIL PROTECTED]
Date: Tue, 18 Sep 2001 00:09:40 EST
Isn't using strerror in a check for POSIX sort of like mixing apples
and oranges.
Yes, just as ISC mixed them.
Is this a check for the strerror function, a check for a POSIX library
or both?
The latter.
If it's
1 - 100 of 2019 matches
Mail list logo
