Re: Broken makefile given Autoconf version mismatch

2006-04-14 Thread Stepan Kasal
Hello, On Thu, Apr 13, 2006 at 08:52:48PM +0200, Alexandre Duret-Lutz wrote: > Or can we tweak Autoconf to make its version more accessible? what would be wrong with parsing `autoconf --version' or `autom4te --version'? (At least as a fallback, when autom4te is not able to handle stdin.) echo

autoconf vs. stdin [was: Broken makefile given Autoconf version mismatch]

2006-04-14 Thread Eric Blake
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 According to Stepan Kasal on 4/14/2006 2:35 AM: > > The problem is that autom4te calls m4 with ` problem with a development version of m4: > There is a related problem: > > autom4te -l Autoconf-without-aclocal-m4 - > > still wouldn't work, because

Re: autoconf vs. stdin [was: Broken makefile given Autoconf version mismatch]

2006-04-14 Thread Stepan Kasal
Hello, On Fri, Apr 14, 2006 at 05:38:51AM -0600, Eric Blake wrote: > Why do we even try to support autoconf reading from stdin? it seems we have to fix *autom4te* to support reading from stdin, as the feature is needed for aclocal. As soon as we do it, `autoconf -' will work again. The fact tha

AC_CONFIG_LINKS and MSVC

2006-04-14 Thread David Byron
I'm having trouble getting autoconf to do what I want when configuring under cygwin for cccl/MSVC. I'm trying to change the behavior of: AC_CONFIG_LINKS(dest.h:source.h) MSVC doesn't seem to deal with reading through cygwin links, I think because the actual filename of the link is dest.h.lnk. W

Re: Broken makefile given Autoconf version mismatch

2006-04-14 Thread Paul Eggert
Stepan Kasal <[EMAIL PROTECTED]> writes: > --- lib/Automake/FileUtils.pm 2006-04-14 09:25:00.0 +0200 > +++ lib/Automake/FileUtils.pm 2006-04-14 09:45:44.0 +0200 > @@ -73,6 +73,9 @@ >$optional = 1 > if $file_name =~ s/\?$//; > > + return "-" > +if $file_name eq "-";

AC_CHECK_FUNCS vs. undeclared entry points

2006-04-14 Thread Eric Blake
While trying to compile CVS gettext on cygwin, I encountered: gcc -c -DLOCALEDIR=\"/usr/local/share/locale\" - DLOCALE_ALIAS_PATH=\"/usr/local/share/locale\" -DLIBDIR=\"/usr/local/lib\" - DBUILDING_LIBINTL -DBUILDING_DLL -DIN_LIBINTL -DENABLE_RELOCATABLE=1 - DIN_LIBRARY -DINSTALLDIR=\"/usr/local/

Re: AC_CHECK_FUNCS vs. undeclared entry points

2006-04-14 Thread Noah Misch
On Fri, Apr 14, 2006 at 02:53:30PM +, Eric Blake wrote: > It turns out that gettext was doing AC_CHECK_FUNCS([__argz_stringify]). Well, > on cygwin, the entry point __argz_stringify is defined as an alias to > argz_stringify, so the conftest.c program links just fine because it provides > its

Re: AC_CONFIG_LINKS and MSVC

2006-04-14 Thread Noah Misch
On Fri, Apr 14, 2006 at 10:11:39AM -0700, David Byron wrote: > I'm having trouble getting autoconf to do what I want when configuring > under cygwin for cccl/MSVC. I'm trying to change the behavior of: > > AC_CONFIG_LINKS(dest.h:source.h) > > MSVC doesn't seem to deal with reading through cygwin

RE: AC_CONFIG_LINKS and MSVC

2006-04-14 Thread David Byron
On Fri, Apr 14, 2006, Noah Misch wrote: > You can use the third argument to AC_CONFIG_COMMANDS to get > the variable declared in `config.status'. Probably this: > > AC_CONFIG_COMMANDS(foo, $config_cmd, config_cmd='$config_cmd') This did the trick. > AC_CONFIG_LINKS is not so widely used, an

Re: AC_CONFIG_LINKS and MSVC

2006-04-14 Thread Noah Misch
On Fri, Apr 14, 2006 at 03:43:39PM -0700, David Byron wrote: > AC_C_BIGENDIAN(ENDIAN=be,ENDIAN=le) > > doesn't work for me using MSVC. The test program #includes sys/params.h > which MSVC can't find. config.log has: > > conftest.c(33) : fatal error C1083: Cannot open include file: > 'sys/param.

Re: nested packages

2006-04-14 Thread Bob Rossi
> > I'm not interested in having it installed unless the package is > > being configured on it's own. (Unless the library install dir is the > > same as CGDB's installed dir --prefix=). > > OK, that should be fine. Thanks for the help. This seems to work wonderfully for me! Bob Rossi _

Re: AC_CHECK_FUNCS vs. undeclared entry points

2006-04-14 Thread Paul Eggert
Eric Blake <[EMAIL PROTECTED]> writes: > Should we change AC_CHECK_FUNCS to also check for a prototype? This is a longstanding issue with Autoconf, and as a practical matter I'm afraid the answer is no, for backward compatibility reasons. > Or should we just beef up the documentation of AC_CHECK