autoconf-patches
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: problem in Makefile.maint


From: Stepan Kasal
Subject: Re: problem in Makefile.maint
Date: Mon, 22 Aug 2005 14:38:12 +0200
User-agent: Mutt/1.4.1i

Hello,

On Mon, Aug 22, 2005 at 12:26:54AM -0700, Paul Eggert wrote:
> >  GZIP_ENV = \
> >    "--no-name --best$$( \
> > -    case $$(gzip --help) in \
> > +    case $$(gzip --help 2>&1) in \
> >      (*--rsyncable*) echo ' --rsyncable';; \
> 
> Do you know of any gzip version that has --rsyncable but where --help
> outputs to stderr?  If not, then the patch is unnecessary.

I agree that my patch is equivalent to

> > +    case $$(gzip --help 2>/dev/null) in \

But with the original code, the help text was printed on my terminal,
and I had to find out about how GNUmakefile includes Makefile.maint, etc.
(Before I realized that I'm indirectly using Makefile.maint,
I was afraid that this problem could affect all users of 2.60.)

I don't think that when a developper occassionally runs "make distcheck"
then he should be punnished this way.

There are many solutions:
1) gzip --help 2>&1
2) gzip --help 2>/dev/null
3) remove GNUmakefile from CVS 
4) Implement a warning somewhere into Makefile.maint, which would state
that for Autoconf developers should upgrade to development gzip.
(Remember, the problem appears with the latest release from
ftp.gnu.org/gnu/gzip/!)

Have a noce day,
        Stepan




reply via email to

[Prev in Thread] Current Thread [Next in Thread]