On Tue, Feb 08, 2005 at 01:52:49PM -0500, Dan Manthey wrote: > By the way, I generated the last patch and this one via `cvs diff -u', > which certainly captures all the needed information but I tried actually > applying the patch to a new cvs working copy and couldn't figure out how `patch -p0' tolerates CVS diffs. > +while grep -q "^CEOF$ac_eof\$" conf$$subs.sed; do -q is not portable. Use `>/dev/null 2>&1'. Test 116 fails. It looks like you need to add an `rm -f conf$$subs.sed' somewhere or store that file in $tmp. This is the relevant hunk from testsuite.log: +--- state-ls.before 2005-02-08 21:05:28.623045244 -0500 ++++ state-ls.after 2005-02-08 21:05:29.775731159 -0500 +@@ -1,5 +1,6 @@ + aclocal.m4 + autom4te.cache ++conf3238subs.sed + confdefs.h + configure + configure.ac I observed no performance regressions. Thank you for that. This seems ready to check in upon execution of your copyright papers. It is a fine patch; thanks for evolving it. It will be interesting to see which sed implementations can handle a substitution of, say, 12 KiB spread across twenty lines. _______________________________________________ Autoconf mailing list Autoconf@xxxxxxx http://lists.gnu.org/mailman/listinfo/autoconf
