[05:42] morning [05:52] good morning desktoppers! [06:59] hey, morning andyrock and oSoMoN [08:03] moin [08:03] hey Laney willcooke [08:03] good morning desktopers [08:04] how was your w.e? and how is that new week looking? [08:04] hey Laney, willcooke [08:06] hey seb128 hey didrocks [08:06] good! [08:07] well on Saturday it rained all day more or less so we went to the library then the coffee shop [08:07] but yesterday was nice, allotment then pub with friends and board games [08:09] which board games in particular? [08:10] morning all [08:12] didrocks: https://boardgamegeek.com/boardgame/127024/room-25 then https://boardgamegeek.com/boardgame/73761/k2 [08:12] sorry, terminal froze up weirdly [08:13] didrocks: seb128: willcooke: how are you! [08:13] supposed to get to 21° today [08:13] I've just run out of disk space. Suspect my Monday morning backup script has gone wonky [08:13] on the target? [08:13] ah, room 25 :) [08:14] were you on of the traitors? :p [08:14] local hdd -> big F. off tar -> scp to local server -> rsync to AWS [08:14] (don't know K2) [08:14] Laney: doing well, thanks! supposed to get to 27° today :p better to run early… [08:14] someone else was the traitor [08:15] but at the start one of the contestents claimed falsely that he was [08:15] ahah [08:15] so we all suspected the wrong person [08:15] funny [08:15] * didrocks has horrible stories of people being killed at the second round while not being the traitor :p [08:15] (it was Rick's son being the traitor, and no-one suspected him at that time) [08:16] I fooled all of us [08:16] he* [08:16] :D [08:16] found the exit [08:16] marked it as dangerous [08:16] and we lost :p [08:16] clever [08:16] I wanted to push someone into the fire room [08:16] but didn't get the chance [08:37] * Laney relocates to town [08:40] * duflu wonders if there is ASCII art for teleportation [09:13] officially welcome duflu! :) [09:13] Trevinho, umm, thanks [09:17] evening Trevinho duflu [09:17] Hey seb128 [09:17] I still call this afternoon... [09:17] :-D [09:17] * Trevinho just got lunch... [09:17] if you wish :-) [09:18] but... hey seb128 :-D [09:18] what time is it for you? [09:18] seb128: how is the new life going? [09:18] 6pm? [09:18] seb128: 17... [09:18] oh ok, yeah, that's afternoon [09:18] life is good [09:18] had a nice relaxing w.e [09:19] and not so much sleep this night to balance it :p [09:21] eheh... Well, it has to be :-D [09:35] hey jamesh :) [09:35] hi willcooke [09:36] willcooke, sounds like I should email you status on Tuesdays and not Fridays (?) [09:37] duflu, yes please. Don't worry about being verbose, we can cover that in 1:1s, so something suitable for copy&paste to the wider team will be really helpful [10:02] * duflu fades into the night === JanC_ is now known as JanC [11:40] it looks like we are getting quite some reports about zesty gconf upgrade issue, postinst hitting an error [11:45] seb128, do you have a bug #? [11:46] jibel, bug #1688721 [11:46] bug 1688721 in gconf (Ubuntu) "package gconf2-common 3.2.6-3ubuntu7 failed to install/upgrade: subprocess installed post-installation script returned error exit status 128" [Undecided,Confirmed] https://launchpad.net/bugs/1688721 [11:46] thanks [11:46] yw [11:46] I wonder if that's the same as https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=760439 [11:46] Debian bug 760439 in gconf2-common "gconf2-common: error processing" [Important,Open] [11:47] and having to do with the perl binding missing (which was pointed as an issue before release on devel iirc) and debconf? [11:47] (random guess statement from the warnings in the dpkg log) === JamieBen_ is now known as JamieBennett [11:53] seb128, lot of reports mention chrome, maybe it's related [11:56] jibel, I would guess that chrome pulls in gconf which we don't install by default anymore otherwise [12:51] seb128, jibel: FYI starting with version 59 (to be released beginning of June) we will stop building chromium-browser with gconf (of course that doesn't affect upstream chrome packages) [12:58] oSoMoN, in this case it's chrome [13:50] hello desktoppers [13:50] hey kenvandine! [13:50] i'm building a xenial backport of gnome 3.24 in a ppa for the platform snap [13:50] but running into a gtk test failure only on xenial [13:51] ERROR:/build/gtk+3.0-dBZtvN/gtk+3.0-3.22.15/./testsuite/css/style/test-css-style.c:115:load_ui_file: assertion failed (error == NULL): colornames.css:120:42'rebeccapurple' is not a valid color name (gtk-css-provider-error-quark, 1) [13:51] which i think is related to this [13:51] Failed to open file “./test-css-style.gresource.xml”: No such file or directory [13:51] anyone familiar with those gresource files that can confirm that might be the cause? [13:52] full build log at https://launchpadlibrarian.net/320395481/buildlog_ubuntu-xenial-amd64.gtk+3.0_3.22.15-0ubuntu1_BUILDING.txt.gz [14:00] kenvandine, the zesty build has the same no such file so it's not likely the issue [14:01] seb128, ah... ok [14:01] * kenvandine should have thought to check for the warning :) [14:01] I wonder if it uses external files as a reference [14:01] oh... [14:02] kenvandine, try maybe reverting https://git.gnome.org/browse/gtk+/commit/testsuite/css/style/colornames.ui?id=c27b8b48 ? [14:02] rebeccapurple is defined in the test [14:02] just to see [14:02] I think those come from pango [14:02] try building that first [14:02] https://git.gnome.org/browse/gtk+/commit/?id=c27b8b48 [14:02] in fact [14:02] k, Laney got there first :p [14:02] I was looking at what provides the reference [14:02] oh... [14:02] sry :-) [14:02] damn ordering issue :) [14:02] ok, thanks! [14:03] I told you that deps might be needed [14:03] :p [14:03] seb128, indeed... [14:03] if that's true, please commit a BD version bump [14:03] sure [14:03] i'll push that to the ~ubuntu-desktop branch as well so it makes it in the next uplaod [14:03] upload [14:03] ya [14:03] but yeah [14:03] https://git.gnome.org/browse/pango/commit/?id=5aa230b1421b1a7dff8525125a5a22de48c5b6e9 [14:03] thanks guys [14:04] confirmed in https://bugzilla.gnome.org/show_bug.cgi?id=770653 [14:04] Gnome bug 770653 in .General "test-suite failure: testsuite/css/style/test-css-style.c: FAIL" [Normal,Resolved: wontfix] [14:04] "don't see a strong reason to bump the pango dependency, just to make a test pass." [14:04] such reference [14:04] not worth an upload for it [14:05] but no reason not to :) [14:05] nah [14:05] but we'll get it next time [14:05] mbiel reported it, he could have bumped the build-depends in Debian... [14:05] oh well [14:05] good work figuring it out team :-) [14:09] seb128, about the gconf2 bug you mentioned earlier, it's reproducible by installing chrome with ubuntu-software. It hangs on ucf (in gconf2-common postinst) [14:09] i'll update the report [14:10] zesty? [14:10] yes zesty [14:10] I didn't try on artful [14:12] nod [14:13] can you check/say if libgtk2-perl is installed? [14:13] I think that was the package they were doing stuff with just before release that seb128 referred to [14:16] Laney, it is installed [14:17] k [14:17] it's one for robert_ancell to look at most likely anyway ;-) [14:20] ximion: hey, got 2s to talk appstream for gstreamer elements? [14:20] do you think it makes sense for asgen to generate this? [14:20] try apt show gstreamer1.0-plugins-bad - I think everything we'd need is there [14:21] that'd be type=codec components [14:25] seb128, http://bugs.debian.org/863134 [14:25] Debian bug 863134 in wnpp "ITP: libayatana-indicator -- Ayatana Indicators panel applets shared library" [Wishlist,Open] [14:30] xnox, nice, see there are community people interested in some bits :-) [14:31] jibel, can you get a bt of the hang? does it do it only if installed from ubuntu-software? or is apt having the same issue? [14:34] Laney: jup [14:34] ximion: k, nice [14:35] Laney: would you want asgen to auto-generate this, or should we rather put a metainfo snippet in the gstreamer plugin package? [14:35] * ximion votes for the latter, we should even upstream that [14:35] I think asgen should make it [14:35] because... [14:35] (if it isn't already) [14:35] otherwise they have to have individual metainfo files [14:35] to deal with package splits and stuff [14:36] like what? [14:36] putting some .so files in one packages and some in another [14:36] ah, you man the individual plugins... yes, that would be necessary [14:36] dh_gst* deal with this for you already [14:36] but I think upstream should have that as metainfo files, to be distro-agnostic [14:37] I guess they could generate them [14:37] also... [14:37] the tag isn't really specified [14:39] Laney: yes, I noticed :D - I wanted to object to AppStream initially because I thought codes were only defined via their mimetypes, then I realized that at least in the example has a meaning [14:40] AFAICS it's not actually parsed currently [14:40] but it's not actually specified anywhere what that is [14:40] so it doesn't matter that much [14:41] but if we do something with PackageKit in the end then you probably want to be specified [14:41] I would still like to fix that in the spec somehow... But I am no expert on codecs [14:41] I think some of the examples in there are kind of wrong actually [14:42] audio/mpeg, mpegversion=(int)4, stream-format=(string){ adts, raw }, base-profile=(string)lc would be a thing that gstreamer can do something with [14:42] before I release AppStream 1.0 I need to read the whole thing from start to finish - the spec has changed a lot over time, and there are a few historical accidents in there [14:43] I guess hughsie and I thought "we of course want a codec component!" but didn't actually clearly specify it (as long as GStreamer is happy) [14:43] they just use it to match the package name atm [14:43] AFAICS [14:43] which is actually what I want it for too [14:43] since you get given the list of packages for a particular list of codec requests from PK anyway [14:52] ximion: have you thought about asgen for fonts? [14:55] jbicha: jup, it works [14:56] not only thought about it, but also worked with this madness for a while ^^ [14:56] the fonts have to ship a metainfo file though, results without it using heuristics were abysmal (hughsie came to the same conclusion at Fedora) [14:56] jbicha: search for Oxygen or Lato in GNOME Software ;-) [14:58] oh ok [14:59] oxygen is the only font where that's been upstreamed so far :( [14:59] jbicha: I have a pending blogpost about this to motivate people to write more metadata (it's really easy!) [15:00] that might be like saying "writing documentation is easy" :) [15:01] the tough part for me is "how do you describe a font?", especially if you want the description to be CC0 [17:53] Laney: you know what could have triggered https://bugzilla.gnome.org/show_bug.cgi?id=782942 ? [17:53] Gnome bug 782942 in general "Linking in Ubuntu/Debian errors with undefined reference to 'sincosf'" [Normal,Assigned] [17:56] last libtool is different from previous by https://pastebin.ubuntu.com/24625273/ as you can see in the new one link_all_deplibs is set to no, but also in the previous version it would have been ignored if no... === muktupavels_ is now known as muktupavels === JanC_ is now known as JanC [19:50] willcooke, hi [19:52] hi robert_ancell - brwy [20:24] Trevinho: how do I make that happen? [20:28] night all o/ [20:29] Trevinho: I tried building the distro package in an sbuild schroot, and I also tried building clutter-gtk in jhbuild with updated glib and g-i [20:29] all worked [23:25] robert_ancell: good morning, I thought you might be interested by Mike Gabriel's new uploads yesterday and today, particularly his Remote Login ones [23:25] https://ftp-master.debian.org/new.html [23:27] jbicha, what's that? [23:27] yeah, I figured he forked before talking to y'all [23:27] oh, is this the old remote login stuff from Ubuntu? [23:27] yes, but renamed [23:30] some of this stuff is old, like his libappindicator fork has his commits from 2015