[09:40] <jtaylor> can I query who moved something from main to universe?
[09:40] <jtaylor> given the changelog gmm is obsolete in universe as we have getfem++
[09:40] <jtaylor> but instead of removing it it was just dumped on us it seems
[09:41] <cjwatson> err, we don't generally apply that much thought
[09:42] <cjwatson> if component-mismatches says to move it and there are no reverse deps in main, we move it, pretty much
[09:42] <cjwatson> we don't spend time seeing whether it should be removed first
[09:42] <jtaylor> k
[09:43] <cjwatson> I don't believe there's any audit trail of override changes except for the date; at any rate, back in January there was no way for Launchpad to know who did the move
[09:43] <jtaylor> I just wanted to check if there might have been reason for why it still exists before I file for removal
[09:43] <cjwatson> (because at that point removals were done from a shared account rather than over the API)
[09:43] <cjwatson> but I think it's still true that there's no audit trail of that
[09:44] <cjwatson> er s/removals/override changes/
[11:48] <jtaylor> why do we have two glade sources?
[11:48] <jtaylor> 3.8 and 3.14
[12:40] <shadeslayer> hm, question, when upstream uses autotools, should they release a tarball after running autogen.sh  or is autogen.sh run during the build?
[12:41] <shadeslayer> *supposed to be run during the build
[12:41] <jtaylor> normally autogen or autoreconf is run by upstream
[12:41] <jtaylor> and the output put into the tarballs
[12:42] <jtaylor> then the user does not need autofoo installed to build
[12:42] <shadeslayer> hm .. ok, then upstream's tarball is most certainly broken
[12:42] <jtaylor> for packaging its often wise to run autoreconf during build
[12:42] <jtaylor> dh-autoreconf helps with that
[12:42] <shadeslayer> yeah, I was already looking at that :)
[12:43] <shadeslayer> it's just that upstream released a tarball withour running autogen.sh
[12:43] <shadeslayer> so I'll have to repack it
[12:43] <jtaylor> repack?
[12:43] <jtaylor> running autogen should be enough for the package
[12:45] <shadeslayer> okay
[12:45] <jtaylor> unless of course the tarball does not have the autotools files in it
[12:45] <jtaylor> which makes the source tarball utterly useless
[12:45] <shadeslayer> autotools files?
[12:46] <jtaylor> configure.ac Makefile.am etc
[12:46] <shadeslayer> ah, that it does have
[12:46] <shadeslayer> the only issue is that upstream forgot to run autogen.sh
[12:46]  * shadeslayer is from the CMake world
[12:47] <shadeslayer> I've only used autotools extensively once, and that too worked in the first go, so don't have alot of experience ;)
[12:55] <shadeslayer> jtaylor: btw when you said running autogen should be enough for the package, did you mean I should run autogen.sh inside the build using debian/rules ?
[12:55] <jtaylor> autogen.sh normally just does autreconf + some flags
[12:55] <jtaylor> dh-autoreconf does that too
[12:55] <jtaylor> so with dh-7 rules its just --with autoreconf
[12:56] <shadeslayer> actually, this autogen.sh has some other stuff as well
[12:56] <shadeslayer> http://paste.ubuntu.com/1249416/
[12:56] <jtaylor> dh-autoreconf can also execute a script
[12:56] <shadeslayer> ah ok
[12:56] <jtaylor> override_dh_autoreconf: dh_autoreconf -- ./autogen.sh
[12:56] <shadeslayer> right :)
[13:00] <shadeslayer> thanks :)
[13:01] <shadeslayer> oh whee .. stupid autogen.sh also runs configure, and the only way to fix it is to patch it out
[13:01] <jtaylor> don'T use autogen then
[13:01] <jtaylor> autoreconf should be enough in most cases
[13:02] <shadeslayer> actually, I need intltoolize --automake --copy --force || exit 1
[13:02] <shadeslayer> for translations
[13:02] <shadeslayer> I guess I could just run all of the commands manually via rules and skip configure
[13:02] <jtaylor> autoreconf should include that, but I have to check
[13:03] <shadeslayer> doesn't I think
[13:04] <jtaylor> then the configure.ac should be buggy
[13:04] <jtaylor> hm no it does not seem included
[13:45] <shadeslayer> jtaylor: http://paste.ubuntu.com/1249498/ < does that look sane?
[13:47] <jtaylor> the exit 1 is unnecessary
[13:47] <jtaylor> else its fine
[13:48] <jtaylor> please add a newline between the targets
[13:48] <jtaylor> makes it more readable
[13:48] <shadeslayer> yeah, fixed that locally :)
[17:21] <cjwatson> shadeslayer: dh_autoconfigure -> dh_auto_configure in the diff above
[17:22] <cjwatson> and really that override_dh_autoreconf is busted, because it won't allow dh-autoreconf to keep track of changed files correctly
[17:22] <cjwatson> I would suggest writing a script called something like debian/autogen with the commands you need, and then dh_autoreconf -- debian/autogen
[17:23] <cjwatson> or
[17:23] <cjwatson> autogen:
[17:23] <cjwatson>         autoreconf -fi
[17:23] <cjwatson>         glib-gettextize --copy --force
[17:23] <cjwatson>         intltoolize --automake --copy --force
[17:23] <cjwatson> and then 'dh_autoreconf -- debian/rules autogen
[17:23] <cjwatson> '
[17:24] <cjwatson> which BTW are the two suggestions for this situation in dh-autoreconf(1)
[17:24] <cjwatson> then you shouldn't need the override_dh_clean
[17:46] <shadeslayer> oh hmm
[17:46] <shadeslayer> cjwatson: thanks for the suggestions, I'll fix it :)
[18:29] <ESphynx> awesome Ecere has been acceptd into Debian Experimental =) thanks everyone