On Wed, 24 May 2006, Stephen Thorne wrote:
> Thanks for the patch, but it fails to apply, Can you check the md5sum?
> 8da52ebce2725b4056a4e56a657b8a65 is what I get...
>
> Can you check that ranlib on solaris has the same (broken) behaviour
> as gnu ranlib? gnu ranlib will exit with an exit code of 0 if you pass
> it -c. Can you check that solaris ranlib does so too?
>
Hmm, I'm kinda lost then. You're saying that the problem with gnu ranlib
is that it returns 0 when ran with -c. I have GNU ranlib and it exits
with 9 when ran with -c.
I know that there are versions of BSD like systems that replaced ranlib
with ar, so the ranlib is actually script that has 'exit 0' in it. I have
such ranlib on Solaris 8. Actually any SunOS 5.x has such ranlib.
I think you have such ranlib on your system as well. If this is the case
it means that ar will do all the job that ranlib was supposed to do as
well. So in case of such system running ranlib -c or ranlib -qwe would
have same effect on library - no effect at all.
I actually created the configure.in patch to fix the code that looks wrong
to me:
AC_MSG_CHECKING(for OSX ranlib)
if $RANLIB -c 2>/dev/null; then
RANLIB_FLAGS=
AC_MSG_RESULT(no)
else
RANLIB_FLAGS='-c'
AC_MSG_RESULT(drat - yes)
fi
I'm not sure what your shell interpreter returns on the first test, but
mine always fails this test (no matter if I run ranlib -c or not).
The correct way with GNU lib would be:
AC_MSG_CHECKING(for OSX ranlib)
(eval $RANLIB -c) 2>&5
if test $? -ne 0; then
RANLIB_FLAGS=
AC_MSG_RESULT(no)
else
RANLIB_FLAGS='-c'
AC_MSG_RESULT(drat - yes)
fi
Hope this help understanding what's wrong with ranlib. I'd be glad if
you'll add this patch to configure.in, because I'm kinda frustrated with
darcs already.
> darcs annotate has a really annoying output. but it shows you who
> wrote every line.
After Trent's message about Mercurial I've read lots of info on it and it
really looks much better to me in numerous ways (most important it has
diff features similar to CVS and change sets (i.e. patches) are easily
searched and diffed).
I'd vote for moving to Mercurial from darcs. The main idea of
decentralized repository would stay.
Please (everyone), take a look at Mercurial's home page and see if you
would like working with it more than with darcs:
http://www.selenic.com/mercurial/wiki/index.cgi
Stas.
> --
> Stephen Thorne
>
> "Give me enough bandwidth and a place to sit and I will move the world."
> --Jonathan Lange
>
> _______________________________________________
> netrek-dev mailing list
> netrek-dev at us.netrek.org
> http://mailman.us.netrek.org/mailman/listinfo/netrek-dev
>