[00:01] morning bryce, RAOF, TheMuso and robert_ancell [00:01] https://wiki.ubuntu.com/DesktopTeam/Meeting/2011-05-24 [00:01] ready? [00:01] And rearing to go. [00:02] :) [00:02] o/ [00:02] awesome... [00:02] heya [00:03] robert_ancell might still be ill, was hoping he was here. [00:03] wow, still? [00:03] poor guy [00:03] there is a western edition todo to talk about bzr usage (see wiki) [00:03] bryce: yeah...hope he is all better soon... [00:03] Hey folks. [00:04] jasoncwarner, so we're going to switch xorg packaging from git to bzr? [00:04] ;-) [00:04] No I think the bzr stuff is for gnome packaging. [00:04] pitti told me SVN to store it and hg-svn to do your local work [00:04] no? [00:04] :) [00:04] :) [00:04] I'll use the CVS-svn interface [00:05] seriously though, I do hate git [00:05] lol [00:05] Ok...for realz though. [00:05] lets get through the meeting and then we can discuss when robert feels better :) [00:06] [TOPIC] X update [00:06] I think that all our blueprints are approved and ready to go. [00:07] And bryce has re-targeted our versions-tracking infrastructure at oneiric (thanks!) [00:07] I see xserver 1.11 has _already_ broken api [00:07] You know what they say - break early, break often! [00:08] yeah I need to rewrite that script in python and have it auto-probe the ubuntu version so I don't have to manually fiddle with it every release [00:08] RAOF, (I don't think they really say that) [00:08] IMO it should be break early, break once. [00:09] there's now an x-backports ppa set up in ubuntu-x-swat [00:09] bryce: Anything more to talk about re: Mesa switching to an SRU-friendly point release policy? [00:09] timo and sarvatt will be filling it in as we go [00:10] oh yeah, I went up to Intel in hillsboro for a meeting with Keith Packard and the graphics team last week [00:11] they're thinking about trying to do point releases of mesa and keeping it stable, in hopes we could pull in point release updates to mesa post-release [00:11] I explained our sru policies wrt point release updates, so guess we'll see how they do [00:11] they were very encouraging of doing more bug upstreaming too [00:12] so this week I've been mostly hammering on my upstreamer tool stuff in arsenal to try to get it more functional === robbiew is now known as robbiew_ [00:12] Intel's development process has changed, too; they're becoming much more distro focused in terms of when they consider hardware support "available" [00:13] so sounds like Gen7 hw (Iron Lake?) will be a smoother transition than Sandybridge was [00:13] IvyBridge. [00:13] Ivy Bridge, right! [00:13] anyway, I also inquired about getting us some HW [00:13] Yeah was reading about IvyBridge and the inclusion of Thunderbolt the other day. [00:14] which turned out that Intel interfaces with pgraner [00:14] Yeah. Those chips are due out around the end of the year, and the kernel support for them has landed, as has some mesa bits. There should be time for solid testing before the distro release season :) [00:15] they said they have to ship to a biz address, so the systems get sent to Lexington and I guess parcelled out from there [00:15] RAOF, anyway I figure either you or I should bug pete about ensuring we're somewhere on the list to get one so we can do gfx testing [00:16] Yes! [00:16] It's probably significantly cheaper to ship Lexington → Portland :( [00:16] (Than Lexington → Hobart) [00:17] heh [00:17] anything else you guys want to talk about? [00:17] When can we expect oneiric to not look like arse? :) [00:18] uhm...uhm.... [00:18] uhm... [00:18] Heh. [00:18] * TheMuso is pondering upgrading to oneiric this week, but not sure how badly things will break. Might try a VM first. [00:19] Ok...do you guys want to wait for robert before talkign about bzr or do you want to go ahead and talk about it and catch him up? [00:19] So I can at least make sure accessibility won't break when I upgrade for real. [00:19] oh, there is one thing... while we're escaping nvidia/fglrx breakage due to xserver abi breakage, what about the kernel? [00:20] bug #776895 [00:20] Launchpad bug 776895 in fglrx-installer "fglrx 2:8.840-0ubuntu4 fails to build against 2.6.39 kernels, due to missing linux/smp_lock.h" [High,Triaged] https://launchpad.net/bugs/776895 [00:20] Oh, yeah! fglrx doesn't build without the big kernel lock! [00:20] Urgh. [00:20] Re bzr, I don't mind either way. The vast majority of stuff I package either doesn't have a packaging branch in bzr, or has branches maintained for Ubuntu in Debian vcs. [00:21] The latter suiting me just fine. [00:21] So whatever is decided, I can work with, if it affects me. [00:22] I guess I have a question for you X guys, somethign that doesn't affect me, but people from the community may be asking about. What about bumblebee and oneiric, what will your official position be? [00:22] desrt: hello [00:22] TheMuso, might affect us via sponsors queue work and such [00:22] TheMuso: bumblebee? (/me googles) [00:23] RAOF: A little piece of software to help users with NVIDIA optimus make use of both video chips in a way. [00:23] hybrid graphics eh? [00:23] * TheMuso doesn't plan to try it, as I am one of the fortunate ones who can turn on/off whatever video chip I need. [00:23] Something like that. [00:23] Not as automatic as in Windows, but users can manually choose to run certain apps on the intel or NVIDIA chips. [00:24] TheMuso, I'd like to see us with a more cohesive solution to hybrid graphics before we start officially supporting it [00:24] SO I am just thinking ahead about people asking for it etc. [00:24] bryce: Right, I thought as much, but I am thinking about what the community may ask for. [00:24] I'm worried there's going to be a lot of different one-off "solutions" which are going to make support complicated in the near term [00:24] Again, I don't care, doesn't affect me, I work around it by using intel only. [00:25] * TheMuso nods. [00:25] So, my position would be that the *correct* solution is Dave Airied's gpu buffer sharing + other work. [00:25] what we're hoping is that these different options will compete and the field will narrow as less good options die off [00:25] Right. [00:26] but for oneiric I think our hybrid graphics plans are [00:26] I don't think I'd object to something like bumblebee in the archive, though. [00:26] Yeah, undefined is what I thought. [00:26] it would be great to see some community involvement in pushing things like bumblebee forward [00:29] Ok cool that satisfies my curosity. [00:29] curiosity even. [00:29] Hm. Although obviously before bumblebee can be uploaded it'll need to build from source :) [00:29] am I remembering correctly RAOF that you asked Rick for hybrid HW? [00:30] bryce: Yes, I did. Well, I ended up asking Pete and then Chris; there's apparently one spare after UDS. [00:30] good deal, maybe you can arrange to take possession at the sprint ;-) [00:31] RAOF, looks like sarvatt has a fix for bug 776895, so that should resolve the kernel breakage. But I do worry about it generally. [00:31] Launchpad bug 776895 in fglrx-installer "fglrx 2:8.840-0ubuntu4 fails to build against 2.6.39 kernels, due to missing linux/smp_lock.h" [High,Triaged] https://launchpad.net/bugs/776895 [00:33] I have to step out for a bit...be back later. [00:33] I wonder whether that will introduce crazy races; dropping the BKL means that fglrx will be subjected to things possibly changing under it. [00:34] I'm not sure what we can do about the kernel breakage in general, though. [00:35] Oh, boo. Evolution no longer starts :( [00:35] hmm [00:36] well, maybe we can make jockey not offer the driver [00:36] Are there any current radeon GPUs that are not supported by the open driver yet? [00:36] Oh, right. [00:36] TheMuso: I don't believe so, no. [00:37] HD 69xx [00:38] ah ok. [00:38] * RAOF needs to remember that patches for HD 69xx support being committed is not the same as HD69xx support existing in a release :) [00:55] kinouchou: bonjour! [00:58] :) [00:58] how are you? [00:58] good, actually. just woke from a nap which i took because i had a headache. headache is gone. :) [00:59] i have many headache [01:00] you need more naps, clearly [01:00] yes [01:01] the bad news is that my girlfriend has stolen my ubuntu.fr pin [01:01] she says that it reminds her of the 80s [01:01] :( [01:02] I have 2000 ubuntu-fr pin but in France [01:02] where are you now? [01:03] at home near Paris [01:03] ah. it sounded like you were somewhere else. [01:04] i say that because I can't give you a new pin [01:04] clearly the solution is for me to come to paris. see you soon. [01:05] yep [01:06] okay. i've booked my flight. i arrive in 1 month exactly. [01:11] desrt: bonne nuit, i go to sleep [01:12] good night [01:12] see you in 1 month [01:12] ok, i wrote in my calendar [01:13] ta === bryceh is now known as bryce [01:23] anyone want to fix librsvg again? [01:26] micahg: Whats up with it? [01:26] TheMuso: same problem, one more substitution needed in the postinst [01:27] TheMuso: multiarch dir not being accessed [01:27] micahg: If you have a patch, I'm happy to sponsor. [01:27] TheMuso: k, give me a minute [01:27] Sure. [01:31] http://web.mit.edu/andersk/Public/ubuntu/librsvg_2.34.0-0ubuntu3_multiarch.debdiff [01:31] andersk: no, that's not right, the non-multiarch path shouldn't be in the second line [01:31] It is in libgdk-pixbuf2.0-0.postinst. [01:32] andersk: you'll get the same error again since the dir doesn't exist === rickspencer3_ is now known as rickspencer3 [01:32] I'm about to test the fix [01:32] It does for me, since libwmf0.2-7-gtk has /usr/lib/gdk-pixbuf-2.0/2.10.0/loaders/io-wmf.so. [01:33] If that’s going to go away, I guess libgdk-pixbuf2.0-0 needs to be fixed again too. [01:34] andersk: yes, but not everyone will have that installed and yes, it's going away with the multiarch transition [01:37] If I remove libwmf0.2-7-gtk, I get an error that /usr/lib/gdk-pixbuf-2.0/2.10.0/loaders/*.so doesn’t exist, but at least the entire command doesn’t fail. [01:38] But I assume you’re in the middle of writing something smarter, so I’ll let you go ahead with that. [01:38] yeah, just trying to test [01:40] TheMuso: http://paste.ubuntu.com/612527/, I can give you a bug if you want [01:41] But that will break libwmf0.2-7-gtk, right? [01:43] andersk: probably, but it should be fixed in any case [01:43] micahg: If there is a bug for it, it would help to close it, otherwise there is no problem. [01:43] TheMuso: I see one from the old issue, not for the ubuntu3 specific issue [01:44] andersk: well, it shouldn't actually break it, it just won't load the stuff from that library [01:44] libgdk-pixbuf2.0-0.postinst looks in both directories and uses 2>/dev/null to suppress the warning if the older path goes away. [01:44] micahg: Ok then don't worry about it, I'll just upload as is. [01:44] micahg: So you tested and it worked? [01:45] TheMuso: yeah, let me just check with slangasek about what andersk is asking [01:45] Ok. [01:52] I updated my patch to add the 2>/dev/null, if that’s the route you want to take. I think it makes sense, because this is supposed to do the same thing as libgdk-pixbuf2.0-0.postinst. [01:55] Ok so what needs changing in that patch that was patebinned? [01:55] Haven't looked much into multi-arch myself just yet. [01:58] andersk: have you tested your version? [01:59] Yeah, just did, it works. [02:00] TheMuso: ok, let's go with his solution, I'd rather 2 checks for this and not suppress errors, but that would diverge from what's currently being done, so it should be discussed [02:04] Ok. [02:05] So where do I find that patch, or do I misread, and we are leaving it for now? [02:05] http://web.mit.edu/andersk/Public/ubuntu/librsvg_2.34.0-0ubuntu3_multiarch.debdiff [02:05] Ok thanks. === braiam is now known as la === la is now known as la_tierra === la_tierra is now known as braiam [02:11] andersk: Ok thanks for your patch, the package seems to build ok, uploading now. === asac_ is now known as asac [03:23] c === smspillaz|zzz is now known as smspillaz [05:28] Good morning [05:32] Hey pitti. [05:32] c [06:44] hey pitti! [06:47] hey braiam [07:19] good morning [07:20] bonjour didrocks [07:20] guten morgen pitti! [07:24] pitti, hello [07:24] hey ricotz [07:24] the gnome-keyring build still fail at the exact same position in the ppa [07:24] locally it builds fine in a natty pbuilder [07:25] here as well, I build it three or four times on current oneiric [07:25] i have no idea what is causing this :( [07:25] pitti, btw, dont forget a dep on libarchive-dev for libtotem-plparser-dev [07:26] ricotz: yes, I added it, and bumped libquvi-dep [07:26] -dev even [07:26] * pitti diffed configure.in to see [07:26] just uploaded to sid [07:26] hmm, you did? [07:26] i dont see the dep here [07:27] and if http://svn.debian.org was working I could show you.. [07:27] i mean the -dev itself needs to depend on it [07:27] oh, to the -dev [07:27] this might cause pkg-config to fail, i have been there :( [07:28] why? [07:28] Requires: glib-2.0 gobject-2.0 gio-2.0 [07:28] Requires.private: gthread-2.0 libxml-2.0 gmime-2.4 libquvi libarchive [07:28] hey ricotz [07:28] it's only an internal dependency [07:28] it should only depend on the "Requires" -devs, no? [07:28] pitti, hmm, for the missing libquiv it fail the last time [07:29] pitti, see 2.32.4-2 [07:29] didrocks, hey [07:29] that seems like the kind of thing which would cause unnecessary library dependencies in packages which use libtotem-plparser-dev? [07:29] ricotz: anyway, I'm happy to add it if it causes problems [07:30] the last time it caused a build failure of totem [07:30] pitti: Yeah git.debian.org is not working for me, something to do with a read-only filesystem. [07:31] TheMuso: svn works, but websvn isn't back up yet (or might never be again, not sure) [07:32] ricotz: I don't see anythign libarchive-ish in the header files; could that be due to a build system bug? why would totem need libarchive? [07:33] pitti, oh, meant while the libquvi-dev reference was missing the last time it causes the totem build to fail [07:33] which i assume could be the same while missing the libarchive-dev now [07:34] ricotz: hm, if it fails now, I'd rather fix that properly, I think [07:34] if you remove libarchive-dev, what does pkg-config outputs? [07:35] assuming you already updated totem-pl-parser locally [07:35] argh, meh [07:35] $ pkg-config --cflags totem-plparser [07:35] Package libarchive was not found in the pkg-config search path. [07:36] dear pkg-config, this really sounds broken [07:36] ok, I'll add it then [07:36] thanks for pointing out [07:36] yes :\ [07:40] you're welcome [07:43] fix uploaded and committed [08:16] pitti: try http://anonscm.debian.org/ (see http://lists.debian.org/debian-devel-announce/2011/05/msg00009.html) [08:17] TheMuso: ^^ this applies for git.debian.org too [08:17] geser: ah, thanks [08:17] geser: that's a bit of a shame, though, it'll break hundreds of Vcs-Browser: tags in source packages :/ [08:17] http://anonscm.debian.org/viewvc/pkg-gnome/ -> empty :/ [08:20] pitti: not only Vcs-Browser but all Vcs-* tags as the read-only access changed too [08:44] I don't think it's finalised yet [08:45] i.e. I expect the old links to be restored [08:50] hey [08:50] bonjour seb128 [08:50] hey pitti, how are you? [08:51] pretty well, thanks! [08:51] how about you? [08:52] have banged on versions.html a little, some less bright orange there [08:52] I've fine thanks [08:52] versions> great [08:53] the "to update" almost fit on one page now ;-) [08:54] pitti, is the script which is doing this page available somewhere? [08:54] seb128, hello [08:54] ricotz: lp:~ubuntu-desktop/+junk/versions or something [08:54] pitti, thanks [08:56] ricotz, it's written at the bottom on the webpage [08:57] salut seb128! [08:57] lut didrocks, ca va ? [08:57] seb128: ça va bien, et toi ? [08:57] ca va nickel ;-) [08:57] seb128, ahh ;), should have scrolled down this far :\ [08:58] ricotz, you have a key on your keyboard sending you at the bottom :p [08:58] ricotz, do you want to use it for the ppa or something? [09:00] seb128, hehe [09:00] seb128, yes, i was hoping to get a better overview [09:02] morning [09:02] hey rodrigo_ [09:02] hey rodrigo_ [09:03] didrocks, can you get bug #779717 reviewed? [09:03] Launchpad bug 779717 in indicator-multiload "indicator-multiload causes a memory leak in compiz when run under unity" [High,Confirmed] https://launchpad.net/bugs/779717 [09:04] it has a valgrind log of a IndicatorObjectEntryProxyRemote::GetPixbuf() leak and some patches [09:04] seb128: sure [09:04] didrocks, the leak is in unity, the bot picked the wrong component [09:04] didrocks, thanks [09:05] oh nice, Michael found a fix? [09:05] (he's a friend of mine who talked to me about this recently) [09:05] oh, that explains the piware email ;-) [09:06] pitti, he got some patches on the bug at least [09:06] the issue is that dx is so behind on bugs that they probably don't see comments or patches coming [09:06] bonjour didrocks, seb128 [09:07] so, seems dns is failing a lot these last few days? [09:07] not here [09:08] hmm, not sure what's up, but I get a lot of disconnections [09:08] didrocks, seems bug #758248 is the same and jay is on it? [09:08] Launchpad bug 758248 in nux "memory leaking in compiz" [Medium,Triaged] https://launchpad.net/bugs/758248 [09:09] didrocks, see comment #32 [09:09] seb128: this is the "most extreme" from it, isn't it? [09:10] not sure but seems there is different leaks [09:10] the fix from jay is in nux, the patches from michael in unity [09:10] seb128: I think mdz's bug showed other issues in nux [09:10] so worth getting the set reviewed [09:10] yeah [09:11] njpatel, ^ could be worth reading for you as well [09:13] seb128, are all the patches on that bug? [09:14] njpatel, #779717 has 2 patches from pitti's friend [09:14] njpatel, the other bug has a comment from jay added yesterday saying he has a fix in nux 0.9 [09:14] the set seems worth reviewing in context of one of the sru rounds [09:15] njpatel: just ensure that the team look at the pending merge, it should be ok then [09:16] good morning everyone [09:18] hey chrisccoulson_! [09:19] hi didrocks, how are you? === chrisccoulson_ is now known as chrisccoulson [09:19] chrisccoulson: I'm fine, thanks, and you? [09:19] yeah, good thanks [09:19] i'm just wondering whether to do a dist-upgrade right now. i've not done any updates since i upgraded to oneiric last week [09:20] seb128, just to make sure, updating eog to 3.0.2, so the version should be 3.0.2-0ubuntu1, right? [09:21] hey chrisccoulson, isn't today an uk holiday? ;-) [09:21] rodrigo_, yes [09:23] seb128, unfortunately not today ;) [09:24] ;-) [09:26] the amount of packages on https://wiki.ubuntu.com/MozillaTeam/FirefoxUpgradeChecklist/5.0 is really depressing [09:27] i need to think of a way to kill off the extensions on lucid and maverick [09:29] hmmm, it seems that globalmenu-extension is broken again in firefox 7 === tkamppeter_ is now known as tkamppeter [09:32] firefox is at version 7 already? [09:34] geser, yes, on mozilla-central [09:34] https://edge.launchpad.net/~ubuntu-mozilla-daily/+archive/ppa/+build/2524747 [09:34] 7 is probably what oneiric will get [09:44] geser: I was trying to push via ssh, not pull. [09:57] rodrigo_, hey [09:57] ups [09:57] didrocks, hey [09:57] ubuntu.session -> [09:57] RequiredProviders=windowmanager;panel;file-manager; [09:57] hey seb128 :) [09:57] DefaultProvider-file-manager=nautilus [09:57] can I drop the DefaultProvider-file-manager line and the ;file-manager RequiredProviders? [09:58] it's what is opening a nautilus dialog on login I think [09:58] the new nautilus also has an autostart desktop for that now [09:58] seb128: sure, but we should ensure what to do for finale [09:58] well that doesn't change [09:58] hum, so it's autostarted in /etc/xdg/autostart/ ? [09:58] they moved from a component to an autostart [09:59] didrocks, yes, with a conditional start on a gsettings key [09:59] ok, perfect then :-) [09:59] doing that now [09:59] so deciding what to do is just deciding on the key value [09:59] didrocks, I'm doing it [09:59] didrocks, don't bother [09:59] ok, thanks :-) [09:59] I just wanted to check with you I was not overlooking something [09:59] seb128: you should do it as well for the ubuntu-2d.session [09:59] didrocks, thanks ;-) [09:59] ok [10:00] the other, I didn't touch to get the upstream experience [10:00] thanks seb128 :) [10:10] didrocks: hi, did 3.8.14 get postponed (again) ? [10:11] cdbs: why again? [10:11] didrocks: I didn't see a release, did it get released yesterday? [10:11] cdbs: it was planned yesterday, right, but as I wrote in the desktop team report, it will be later this week [10:11] hmm [10:11] cdbs: but it has never been planned before, hene the "again"? [10:11] hence* [10:12] didrocks: Well I thought it was, as you had said so :) [10:12] cdbs: when did I say that? I didn't AFAIK [10:12] just that it was planned yesterday [10:12] and I told yesterday that it will be later this week, in the desktop team meeting [10:13] didrocks: Okay, I really should read the meeting logs or so [10:13] cdbs: and in the report: https://wiki.ubuntu.com/DesktopTeam/Meeting/2011-05-24 [10:13] this timezone is sucky [10:13] cdbs: yeah, you should have a look, we tend to put all info there :) [10:13] at least, try! [10:13] (I won't sign for completeness there :-p) [10:14] cdbs: however, I don't know if there is a direct link to always "the latest report" [10:15] That's not a problem [10:15] atleast not for me. I can check out regularly, like, every Monday [10:16] yeah, I'm wondering still if we should have such link, nevermind anyway :-) [10:30] shankao: hey, so you had some questions on the last gnome-session update? [10:30] hi, after the last gnome-session upgrade, for some reason gnome-session-fallback was not in my system, so the ubuntu-classic option was not selectable [10:30] is maybe a dependency problem? [10:30] shankao: did you get the latest gnome-panel? [10:30] shankao: gnome-panel recommends gnome-session-fallback [10:30] so people upgrading will get it that way [10:31] new install won't as gnome-panel won't be seeded [10:31] nope, I was upgrading [10:31] I have installed the missing package by hand now [10:32] anyway I can take a deeper look at why it was left this way [10:32] shankao: gnome-panel 1:2.32.1-0ubuntu7 recommends gnome-session-fallback, so if you install recommends (which is the default), it will be installed [10:32] so yeah, more clue will be appreciated :-) [10:33] ok, my version is gnome-panel 1:2.32.1-0ubuntu6.5 [10:33] shankao: seems that explains it then :-) [10:34] yep, thanks :) [10:34] yw! [10:37] umh, seems that version of gnome-panel (ubuntu7) is Dependency wait on vernadsky (i386) [10:38] https://launchpad.net/ubuntu/+source/gnome-panel/1:2.32.1-0ubuntu7/+build/2523117 [10:40] jasoncwarner: https://blueprints.launchpad.net/ubuntu/oneiric/+specs?searchtext=desktop-o [10:45] didrocks: can you look at the depwait shankao pointed about? gnome-panel depwaits on libedataserverui1.2-dev but there is only libedataserverui-3.0-dev (and libedataserver1.2-dev) [10:46] geser: I think we rather should work on migrating gnome-panel to 3.0 [10:46] geser: not sure I can do it today, already on other stuff, but will give a look if nobody take it [10:48] didrocks: no problem, I just wanted you to know about the depwait as they can get easily unnoticed [10:48] geser: yeah, it was unnoticed TBH :-) thanks for pointing it shankao, geser! [10:49] hello! as of release 11.04, unity seems to crash(segfault) if xinerama is enabled. is this a know issue? or shall i open a bug once i got all information collected? [10:49] didrocks, we should perhaps make gnome-session recommends -fallback until the e-d-s gnome-panel thing is sorted [10:49] seb128: well, do you think we won't do it for alpha1? [10:50] for e-d-s/gnome-panel [10:50] didrocks, what, updating gnome-panel? not if that involve going to gtk3 and updating the indicator stack [10:51] seb128: I can maybe try to build it with libedataserverui-3.0-dev first and se what happens [10:51] but I really want to first do this unity-2d sortage [10:51] chrisccoulson: do you have a minute to hop into mumble, desktop 1-1? [10:52] pitti - yeah, i just need to find my headset [10:52] 1 second [10:52] didrocks, [10:52] $ ldd /usr/lib/libedataserverui-3.0.so.0.0.0 | grep gtk [10:52] libgtk-3.so.0 => /usr/lib/libgtk-3.so.0 (0x005af000) [10:53] didrocks, not likely to work with a gtk2 gnome-panel [10:53] some more details: three 1080p-displays connected via 2 HD4350 boards, everything working fine when using gnome classic [10:53] seb128: didn't you want to test gtk2/gtk3 linking to the same process? :-) maybe time to experiment! [10:53] didrocks, that's why debian renamed the e-d-s source to have both gtk2 and gtk3 version in the archive until they go to testing [10:53]