* Gary V. Vaughan wrote on Thu, Aug 18, 2005 at 03:28:00PM CEST: > Ralf Wildenhues wrote: > > >I have re-diffed this against current CVS HEAD, including changes for my > >complaints. Unless anybody complains, I will apply this soon and > >backport to branch-2-0, killing yet another impediment to 2.0. :) > > > * libltdl/m4/libtool.m4 (LT_OUTPUT): New macro to generate > > libtool before config.status. > > (_LT_CONFIG_COMMANDS): If LT_OUTPUT has been invoked, then > > generate code in config.status to call config.lt instead of > > generating libtool directly. > > * doc/libtool.texi (LT_INIT): Document LT_OUTPUT. > > * tests/early-libtool.at: New test. > > * Makefile.am (TESTSUITE_AT): Updated. > > * tests/testsuite.at: Updated. Define defaults for > > $ACLOCAL, $AUTOCONF. > > Excellent! I was just addressing your nits with patch-23 having > installed patched autoconf-2.59 and automake-1.9.6. Please go right > ahead...
FYI: Applied to HEAD and branch-2-0. The only thing I changed was _LTEOF -> _EOF in tests/early-libtool.at, because "the user" is not supposed to use _LTEOF in her configure.ac. The backport consisted exactly in removing all references to the new test suite. > I'll repost patch-23 presently. Hmm. I still don't like the massive source tree file reorganization. Can't we get around it? > Did you finish the branch-1-5 forward ports you wanted in before the > next 1.9h beta? If so, I think we are in pretty good shape to pump the > release once this patch and patch-23 are in branch-2-0! :-D I have tested using libltdl a bit. In another mail. And a TODO list in yet another mail (but it's really getting small :). I will also try to do 1.5.20 this weekend, for the ultra-stable lovers. Cheers, Ralf
