On 05/11/2015 07:55 PM, Joseph Myers wrote: > On Sat, 9 May 2015, Bernd Edlinger wrote: > >> But maybe you would like it better if we update, for instance, to: >> automake-1.14 _and_ autoconf-2.69 ? > > Updating to current automake and autoconf release versions (but still > using git versions of the toplevel scripts, not those from particular > releases) is a good thing
Agreed - but that seems to require additional source changes, which is beyond my knowledge (and need). Besides that: wouldn't such commits be of better quality when starting from identical old (1.11.6) versions? > - remembering that toplevel is shared with the > binutils-gdb and newlib-cygwin repositories Just curious: I do remember this from the old CVS days, but given that gcc has its own Subversion repository, how is that organized now? > (unfortunately some files > sometimes get out of sync, especially in newlib-cygwin which people > commonly don't update when changing the other repositories), so if > subdirectories in those repositories aren't updated at least ensure that a > mixture of versions (toplevel using a newer version than some > subdirectories) works. I'm all fine to omit this top-level downgrades and regenerate without the "--add-missing --copy --force-missing" automake flags. Feels more obvious anyway, otherways "--enable-maintainer-mode" should use them already. > (If libtool gets involved in the update, remember about reverting libtool > commit 3334f7ed5851ef1e96b052f2984c4acdbf39e20c from the version of > libtool code used.) Indeed a libtool patch (for aix) is the driving one behind this automake sync. But as far as I understand, libtool-originating files are slightly modified in gcc anyway, so I plan to pick this aix patch only, not a general update. Thanks! /haubi/
