[03:36] <Noskcaj> Can someone review/upload https://code.launchpad.net/~noskcaj/+junk/geoclue-2.0 please? Ubuntu needs to change the source package name due to the number of packages we have that need the old API
[03:51] <xnox> Noskcaj: has that been done in debian already?
[03:52] <xnox> Noskcaj: geoclue 2.0 is in experimental, why do we need it in ubuntu?
[03:55] <xnox> Noskcaj: it would be easier to review, if your branch was based on top of lp:debian/experimental/geoclue
[04:44] <Noskcaj> xnox, I didn't know about the experimental lp branch. darkxst requested that we change source package due to API breaks
[04:44] <Noskcaj> Should i change it?
[05:42] <Logan_> Halp. Why isn't dh-autoreconf working to fix the ppc64el FTBFS on this library? https://launchpadlibrarian.net/161336888/buildlog_ubuntu-trusty-ppc64el.libapache-mod-encoding_0.0.20021209-10ubuntu1_FAILEDTOBUILD.txt.gz
[05:43] <Logan_> https://launchpadlibrarian.net/161336929/buildlog_ubuntu-trusty-arm64.libapache-mod-encoding_0.0.20021209-10ubuntu1_FAILEDTOBUILD.txt.gz <-- And it also didn't update config.{sub,guess} for arm64?
[06:02] <infinity> Logan_: Looks like it's not relibtoolising.  Perhaps the aclocal include is hand-crafted.
[06:02] <Logan_> Hmm.
[06:03] <infinity> Logan_: And autoreconf won't update config.{sub.guess} on projects that don't use automake.
[06:03] <Logan_> It uses automake.
[06:03] <infinity> Oh, it does.  Curious.
[06:04] <infinity> Anyhow, I'd just back out the autoreconf change there, stick to autotools-dev, and hand-patch occurences of the libtool elf64ppc bits.
[06:04] <infinity> Logan_: See, eg: http://launchpadlibrarian.net/161285599/rhythmbox_3.0.1-1ubuntu7_3.0.1-1ubuntu8.diff.gz
[06:05] <Logan_> Did autoreconf not work on rhythm box?
[06:05] <Logan_> *rhythmbox
[06:05] <Logan_> Because I think that's a more futureproof solution than hand-patching the libtool stuff.
[06:11] <darkxst> xnox, geoclue2 is needed by a number of 3.10 apps such as gnome-maps (which currently broken in archive), gnome-clocks, gnome-settings-daemon/g-c-c
[06:21] <infinity> Logan_: autoreconf is definitely the more future-proof solution, and if you want to fix the upstream package to autoreconf correctly, please do.  I'm just being time/effort pragmatic here in some cases.
[06:22] <Logan_> infinity: I don't mind losing the TIL on libapache-mod-encoding... wanna fix it? ;P
[06:23] <Logan_> Lemme look into autoreconf for rhythm box.
[06:23] <infinity> I wouldn't bother, if I were you.  A new upstream will (at some point) pick up the new libtool, and the delta can go away.
[06:23] <infinity> The Debian maintainers once tried to autoreconf and ended up backing it out too. :P
[06:24] <Logan_> Which package are you referring to?
[06:24] <infinity> rhythmbox
[06:25] <Logan_> Oh, okay.
[06:25] <Logan_> My request still stands, in that case. :P
[06:25] <infinity> Yeah, I'll fix up mod-encoding.
[06:25] <Logan_> Many thanks!
[06:25] <Logan_> I was going to say, libapache-mod-encoding isn't being updated upstream any time soon. The last release was in 2002. :P
[06:36] <infinity> Logan_: Also, you forgot to run update-maintainer(1) on this upload.  Probably true for some others too.
[06:36] <Logan_> Wait, I always run that...
[06:36] <infinity> Oh, wait.  No.  That was me starting from the Debian base.  Nevermind.
[06:36]  * Logan_ wipes the sweat off his forehead.
[06:43] <Noskcaj> infinity, Logan_ Is there anything i can do to help with arm64/ppc64el ?
[06:43] <Logan_> Fix FTBFSes! http://qa.ubuntuwire.org/ftbfs/ppc64el.html long lists yay
[06:43] <Logan_> and http://qa.ubuntuwire.org/ftbfs/arm64.html
[06:44] <Logan_> Hmm, libass. Sounds like my kind of library.
[06:45] <Noskcaj> ok, and convince corsac and/or xfce upstream to update their autotools stuff
[06:45] <Noskcaj> lol
[06:45] <Noskcaj> i think there's a merge for it too
[06:47] <Logan_> Noskcaj: I'd recommend looking at my uploads to see how I fix stuff.
[06:47] <Noskcaj> will do
[06:47] <Logan_> (If you want to help out.)
[06:48] <Noskcaj> yeah, i've got nothing else to do
[06:49] <Logan_> Go outside!
[06:49] <Logan_> Don't be a loser like me.
[06:49] <Logan_> Better yet, go outside AND fix ppc64el failures.
[06:50] <Logan_> With a Nexus 7 running Ubuntu Touch or something.
[06:53] <Noskcaj> I already have my pajamas on, so outside doesn't really work, and my brother won't let me put ubuntu on his nexus
[06:53] <Logan_> Hmm, that does pose a conundrum.
[06:54] <Noskcaj> :)
[06:54] <Noskcaj> How would i fix https://launchpadlibrarian.net/154958763/buildlog_ubuntu-trusty-arm64.xfce4-battery-plugin_1.0.5-2ubuntu1_FAILEDTOBUILD.txt.gz ?
[06:55] <Noskcaj> It doesn't look like the standard autotools fix
[06:56] <infinity> It's not.
[06:56] <Logan_> Hmm.
[06:58] <infinity> Logan_: mod-encoding happy.
[06:58] <Logan_> From a quick Google, it looks like you might need #include <unistd.h>.
[06:58] <Logan_> infinity: Yayness.
[06:59] <Noskcaj> ok
[07:00] <Logan_> Which is there, but inside an #ifdef HAVE_SYSCTL conditional block.
[07:00] <Logan_> Hmmmm.
[07:00] <Logan_> ...and checking for sysctl... no
[07:00] <Logan_> I feel like Sherlock.
[07:01] <Logan_> Let's see how it's checking for sysctl, as arm64 should presumably have that.
[07:02] <Logan_> AC_CHECK_FUNCS([sysctl]) ... well that's not helpful.
[07:02] <Logan_> infinity: Any insight? :P
[07:02] <infinity> Dunno.  Retrying it with glibc 2.18 before I dig.
[07:04] <infinity> Nope, still fails.  Curious.
[07:04] <Logan_> I think it has something to do with the sysctl check.
[07:05] <infinity> Well, yes.  But curious that it fails.
[07:06] <Noskcaj> Just wondering, would https://launchpadlibrarian.net/160701912/buildlog_ubuntu-trusty-ppc64el.lxsession_0.4.9.2-0ubuntu6_FAILEDTOBUILD.txt.gz just need a rebuild, since it has autoreconf already?
[07:06]  * Logan_ looks.
[07:06] <infinity> Noskcaj: No.
[07:06] <Logan_> No. The autoreconf failed. I can fix it.
[07:06] <Logan_> We'll either want to specify subdir-objects as an option or remove -Werror.
[07:07] <Logan_> That's thanks to a new automake, not the ppc64el architecture.
[07:08] <infinity> I'd remove Werror, that's actively hostile to the point of dh_autoreconf.
[07:08] <infinity> You want errors to halt the build, but not every new warning.
[07:08] <Logan_> I always have trouble deciding which to do.
[07:09] <infinity> Logan_: A good rule of thumb for Werror (for both C code and autotools) is that upstream should develop with it on, but distributions should ship with it off.
[07:09] <Logan_> Gotcha. I'll keep that in mind for the future.
[07:09] <infinity> Logan_: Upstreams should try to release code as clean as they can, but making me stop and fix every warning on every rebuild with a slightly different toolchain is insanity.
[07:14] <infinity> Logan_: That said, when an upstream is very fastidious about the cleanliness of their code *and* is very open to patch submissions (the latter being a big deal), I don't mind fixing and pushing the patch.
[07:15] <infinity> But it's a huge time sync to do it for every project every few months.
[07:15] <infinity> sink*
[07:15] <Logan_> Noskcaj: Uploaded a new lxsession with a fix.
[07:15] <Noskcaj> Logan_, thanks
[07:15] <infinity> Oh, hrm, midnight happened here and I didn't even notice.
[07:15] <infinity> Happy 2014.
[07:16] <Logan_> So unfestive.
[07:16] <Noskcaj> happy 2014 infinity
[07:16] <Logan_> It's been 2014 here for two hours, 15 minutes and counting. :P
[07:17] <Noskcaj> 18 hours, 16 min here
[07:17] <infinity> I had been planning to celebrate on the east coast.  So, yeah, I noticed when you ticked over, not when I did. :P
[07:17] <Logan_> yay east coast
[07:17] <Noskcaj> what is the normal error for just needing autoreconf?
[07:17] <Logan_> also CDBS needs to die a long and miserable death
[07:18] <Logan_> Noskcaj: for ppc64el? missing *.so
[07:18] <infinity> Noskcaj: For ppc64el, it's generally that autoconf complains that the linker doesn't support shared libraries.
[07:18] <Logan_> during dh_install
[07:18] <Logan_> or that
[07:18] <Logan_> mostly the one I mentioned
[07:18] <Noskcaj> ok, thanks
[07:18] <infinity> Noskcaj: Which can fail early, or can make it all the way to dh_install before noticing there's no .so
[07:18] <Logan_> it often doesn't fail early, unfortunately
[07:18] <Logan_> I'd say 90% of the packages I've fixed so far fail at dh_install
[07:18] <infinity> (But in either case, you'll see the autoconf message in configure saying "checking if ld supports shared libraries... no"
[07:35] <Logan_> Noskcaj: you can throw a debrief at me if you want to see if you're doing things right, since you can't test locally
[07:35] <Logan_> friggin autocorrect
[07:35] <Logan_> debdiff*
[07:44] <Noskcaj> Logan_, For anything complex, will do
[09:57] <JackYu> happy new year, every one!
[12:48] <xnox> darkxst: right, there are 10 packages using geoclue at the moment, it would be nice to test-rebuild all ten of them against geoclue 2.0 and get them ported if porting is needed & simply sync over geoclue.
[12:48] <xnox> darkxst: and please email ubuntu-devel about it, instead of doing it on irc.
[12:49] <xnox> (one might be a duplicate) so 9 in total only.
[12:49] <xnox> darkxst: http://paste.ubuntu.com/6672991/
[12:55] <xnox> darkxst: oh, since it's DBus there might be much more users....
[12:55] <xnox> Laney: is the code search active to do a geoclue search?
[12:57] <xnox> darkxst: is it just me, or geoclue 2.0 API offers a lot less than 0.x did ?
[13:00] <Laney> xnox: http://162.213.35.4/
[13:15] <xnox> darkxst: yeah there are a few more.
[13:19] <xnox> darkxst: it's best to not diverge from debian here, and request for geoclue 2.x to be packaged in-parallel to 0.x series there as well and have matching source package names.
[13:49] <xnox> Laney: http://ubuntu-codesearch.surgut.co.uk/ =)
[13:52] <Laney> neat
[16:21] <alkisg> Happy new year to all
[16:21] <alkisg> stgraber: can we sync LTSP 5.5 from unstable so that we have more time in testing/fixing things? Or do we have to wait until it migrates to testing?
[16:37] <labsin> Hi all, I'm trying to cross-compile my ubuntu touch app.
[16:38] <labsin> I've found the mailinglist post [Cross-compile with CMake from SDK Apps to Unity8/Mir]https://lists.launchpad.net/ubuntu-phone/msg05556.html
[16:39] <labsin> but I can't seem to get it right...
[20:24] <darklight_> Is someone working on https://bugs.launchpad.net/ubuntu/+source/unity/+bug/775434 ? Having the meta+"numbers" shortcut hardcoded is making unity a living hell for me
[21:17] <Noskcaj> Does arm64 have to use gles2 like armhf does?
[21:54] <infinity> Noskcaj: Not currently, though we may have to make that switch in the future.
[21:55] <Noskcaj> ok, then i trust my evas merge slightly more.
[22:02] <darkxst> xnox, geoclue 2 is a completely new API that is not at all compatible with the old version
[22:10] <darkxst> and there will be rdepends that use features not in geoclue-2
[22:30] <xnox> darkxst: ok. then it really should be introduced as a new src package in both debian and ubuntu. Can you ask for it to be a new src package in Debian?
[22:30] <xnox> darkxst: since it's already packaged as same src in experimental....
[22:31] <darkxst> I filed http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=733901
[22:53] <Noskcaj> What do i have to do to fix http://paste.ubuntu.com/6675509/ ? It only occurs with autoreconf turned on
[22:54] <xnox> Noskcaj: that doesn't look like a complete log. look truncated on the top.
[22:55] <Noskcaj> xnox, Does it need to have the stuff above?
[22:56] <xnox> Noskcaj: yeah.... the error is on the top, and then debhelper cat config.log to standard output which is a big pile of poo =)
[22:57] <Noskcaj> http://paste.ubuntu.com/6675520/ is the entire build
[22:57] <xnox> Noskcaj: line 129/130
[22:58] <Noskcaj> ok. Now i will do a lot of googleing to find what is wrong.
[22:59] <xnox> Noskcaj: IT_PROG_INTLTOOL macro was unknown at the time you did autoreconf, so probably simply a built-dependency is missing of a package which provides that autoconf macro....
[23:00] <Noskcaj> so intltool?