=== bjf is now known as bjf[afk] [01:38] Can I safely assume the X transition is under way with all these xorg uploads? [01:40] Not yet. [01:40] These are preparatory steps, which won't break anything. [01:40] So that when we *do* break everything it's quicker to unbreak it :) [01:50] Right. [02:05] TheMuso, a rough plan of attack is at https://wiki.ubuntu.com/X/Roadmap/Natty if you're interested [02:06] bryceh: thanks [02:06] like raof said we're just tending to preparatory stuff right now [02:07] yep gotcha [02:14] Hm. I appear to spell build as “bulid” with remarkable frequency. [02:24] I would like to call attention to bug #574187 which is a pam problem with Gnome screensaver and the nopasswdlogin group. It's really annoying to disable password logins but then get locked out by the screensaver. [02:24] Launchpad bug 574187 in gnome-screensaver "Screensaver asks password when the user is in the password-less login group" [Undecided,Incomplete] https://launchpad.net/bugs/574187 [02:27] jhansonxi: the bug is currently marked as incomplete, which means that it's awaiting action from the reporter before further consideration [02:27] if the bug still exists, you could mark it as confirmed [02:29] Confirmed. I usually don't confirm bugs as I prefer someone slightly more official to do that. === asac_ is now known as asac [06:03] Good morning [06:19] pitti: good morn [06:19] hey desrt, how are you? [06:19] great [06:19] on holidays for the first time in a while :) [06:20] desrt: oh, nice! so what are you doing on IRC then? :-) [06:21] pitti: My thoughts exactly, and, good morning. [06:21] c [06:21] hey TheMuso [06:21] pitti: meh. where else am i going to hang out? :) [06:22] ...anywhere away from the computer would be my answer to that. :) === almaisan-away is now known as al-maisan [08:11] good morning [08:12] good morning didrocks [08:13] hey mvo, how are you? [08:13] didrocks: mostly good, hurt my knee yesterday while playing hockey, but it seems not be too bad (fingers crossed). and you? [08:14] mvo: oh, hockey is really a dangerous game, hope you'll feel better soon :( [08:14] mvo: kind of cold here, but otherwise, excellent, thanks :) [08:15] it is sometimes :/ [08:15] happy to hear that [08:20] hey didrocks, mvo [08:20] mvo: oh, playing Hockey again? [08:20] hey pitti, playing retargetting bugs? ;) [08:21] didrocks: cleaning up the a2 bug list a bit [08:21] didrocks: for some reason I thought a2 was this Thursday [08:21] I'm not sure why QA is targetting for alphaxxx every single bugs they found and add a natty task [08:22] pitti: no worry, we only rely on upstream milestone in any case. I'll talk to Brian so that he stops doing that (or for beta only for very important ones) [08:22] didrocks: merci [08:49] pitti: yeah, when I get the chance I play :) not always easy timewise for me [09:04] morning [09:08] mvo: the lucid update in bug 670629 has some confusing comments -- is it intended to show the EULA on upgrades? [09:08] Launchpad bug 670629 in msttcorefonts "EULA not shown for Microsoft Fonts" [Medium,In progress] https://launchpad.net/bugs/670629 [09:08] good morning rodrigo_ [09:09] pitti: yes, that is what I understand from the legal feedback we got. that we need to show even for existing installs (once) [09:09] mvo: ok, thanks; so I'll move that to -updates, ok? [09:09] hi didrocks [09:09] pitti: sounds good to me [09:20] hey [09:22] didrocks: no need to review my merge request for the ~ubuntu-desktop evince packaging branch anymore, TheMuso already fixed the FTBFS with a similar fix as I proposed [09:23] geser: thanks for the notice [09:23] salut seb128 [09:25] bonjour seb128 [09:26] hey didrocks, pitti, how are you? [09:27] seb128: I'm great, thanks! (apart from being confused about when alpha-2 happens :) ) [09:27] pitti, did you mean to move those bugs milestones or was that part of the confusion? [09:27] just reading my bugmails from the night [09:28] seb128: I still mean it; those weren't alpha blockers [09:28] high/alpha 2 means "we can't release a2 without this fixed", which was clearly exaggerated for most of them === tkamppeter_ is now known as tkamppeter [09:29] pitti, k, need to think about that, that's how dx handle their milestones [09:29] pitti, so you just dropped target from their list [09:30] we need a way for teams to track what they want to work on without creating noise on the r-t list [09:30] seb128: if they are actually working on it by a2, we can put them back [09:30] but they seemed unrealistic to fix in one day [09:30] seb128: but they also have their milestones on their upstream tasks [09:31] right [09:34] hey seb128, I'm fine, thanks, you? [09:36] didrocks, ca va bien ici ;-) [09:36] pitti, well, "one day", they plan another bug fix rollout on monday or tuesday [09:37] pitti, so some days [09:37] ah, ok [09:41] How do i make my clock "24 hour" system wide. What text file or envirnomentvariable do I edit, or what GUI-tool do i use? I am fed up of fightinng the AM/PM-ghost :-s [09:43] Brumle: in unity or classic GNOME? [09:43] pitti: latest LTS :) [09:43] Brumle: set LC_TIME to a locale which uses 24 hours [09:49] pitti, seb128: can you please have a look at this sync request from robert -> https://bugs.launchpad.net/ubuntu/+source/clutter-gst/+bug/658851 ? [09:49] Launchpad bug 658851 in clutter-gst "update the clutter-gst package" [Wishlist,Triaged] [09:50] micahg, ping [09:51] rodrigo_: sure, doing now [09:51] pitti, thanks [09:57] pitti: Thanks. I'll look into it :) [09:58] Hello, can someone help with LP 608515 ? It is happening on my maverick system. I just tried with a natty VM, and that problem does not happen. What I fear is that the problem would still persist even when I upgrade my system to natty. [09:58] Launchpad bug 608515 in empathy "Can't establish audio or video calls" [Low,New] https://launchpad.net/bugs/608515 [10:00] chrisccoulson: good morning [10:00] chrisccoulson: say, do you know if we currently build firefox with some special debug options or so? it has felt like a tarpit for the last weeks at least [10:01] switching tabs, trying to mark text, and other actions takes 5 seconds [10:02] (very often, not always) [10:03] it spins up the CPU to 100% during that [10:08] hi pitti [10:09] no, we haven't changed anything recently [10:09] pitti - you don't use the nvidia driver too do you? [10:10] chrisccoulson: intel [10:10] pitti - does it happen on a new profile? [10:10] chrisccoulson: it's never really been snappy, but right now it makes you want to throw the thing out of the window.. [10:10] chrisccoulson: checking [10:10] pitti, I don't get that issue on intel video there [10:11] chrisccoulson: hm, after restarting it seems to be fast, and then gets worse over time [10:12] but my previous restart was about 2 h ago, so I'll check it with a new profile and let that run for a while [10:33] didrocks: is it intended that libzeitgeist and zeitgeist-extensions want to move to universe? don't we need those for unity? [10:34] pitti: I think it's because unity-place-* aren't recommended anymore by unity as they were uninstallable, will be fixed with this week release [10:34] ah, I see [10:41] pitti - do you know what keeps swt-gtk in main? i looked at it's rdepends yesterday and saw only universe packages [10:42] it's the only thing keeping xulrunner-1.9.2 in main now [10:42] -- natty/main build deps on libswt-gtk-3.5-java: [10:42] gwt [10:42] -- natty/main build deps on libswt-gtk-3.5-java-gcj: [10:42] gwt [10:42] that's it [10:43] pitti - oh, i didn't check the build-deps. but there are no run-time dependencies keeping it in? [10:43] none that I can see [10:51] didrocks: "appdevs-desktop-n-quickly Medium finish /opt work regarding the decisions above" -> is this still actually blocked? Is there still something to decide here? [10:54] pitti: there is still a pending email sent to the application review board without any answer. I need to resent one [11:01] pitti: just did verification for 666511/lucid-proposed [11:01] err, why this nick? === lan3y is now known as Laney [11:01] Laney: ah, thank you1 [11:02] np [11:02] I fixed it for squeeze ages ago, would have taken care of Lucid too but it was already uploaded by then so slipped by me [11:51] pitti - oh, i just saw your e-mails about main being frozen ;) [11:52] chrisccoulson: the followup as well, I hope? [11:52] * pitti <- brown paperbag [11:52] pitti - yeah, i saw that too :) [11:52] heh :-) [11:52] pitti: more coffee in the morning ;) [11:54] pitti: hi there! you scared me, *a lot* [11:55] * pitti hugs nessita -- was a nice wakeup call, wasn't it? [11:55] :-D [11:55] I was already thinking how will I convince you all to upload a few packages :-P [11:57] (that's pitti's technic to get more WI fixed in a small amount of time ;)) [11:58] hey nessita! [11:58] hi didrocks! [12:01] hi nessita [12:01] hi rodrigo_, how are you? [12:01] nessita, feeling cold, I just came back from some errands on my motorbike, so hands frozen :-) [12:01] nessita, what about you? [12:02] all the opposite, my hands are not cold and the weather is freaking hot. Though yesterday it rained a lot so today is a little bit fresher than yesterday [12:03] right, summer there now :) === MacSlow is now known as MacSlow|lunch === zyga is now known as zyga-coffee [12:09] pitti, hmm, "disable hibernate in the UI: DONE"? why is it disabled? [12:10] rodrigo_: was discussed at uds; apparently it's failing for a lot of people, and is slow [12:10] it's slow yes [12:10] our OEM guys disable it for every rollout they do, too [12:10] apparently the major laptop manufacturer want it that way, or so [12:13] ok === zyga-coffee is now known as zyga [12:40] pitti - how strange, i've also just noticed firefox running really slowly on my system, but that only started this morning [12:40] and nothing changed recently :/ [12:59] talking about updates, I haven't updated for a couple of weeks [12:59] is it safe to apt-get upgrade now? [13:10] pitti: I solved the C code problem I mentioned last week, so the MPs linked to https://bugs.launchpad.net/bugs/693337 are ready for review. [13:10] Launchpad bug 693337 in language-selector "Menus for choosing language should have one option per available translation" [Undecided,In progress] === MacSlow|lunch is now known as MacSlow [13:35] GunnarHj: hi! great, thanks! [13:36] GunnarHj: was on the gnome hackfest last week, so I didn't have time to review [13:36] * pitti -> back in 45 mins [13:51] pitti: Yes, I knew about the hackfest. :-) [14:13] seb128: argh, you forgot to remind me for the meeting reminder :) [14:13] * pitti sends now [14:14] is unity supposed to have a battery indicator? [14:16] bcurtiswx_ - perhaps [14:17] only if you're not connected to power, i think [14:17] do you have a battery? ;) [14:17] dpm_: I'll fix the missing files later today [14:17] chrisccoulson, im on my laptop and i just unpluged it [14:17] no battery [14:18] hmm. maybe check your settings in gnome-power-preferences [14:18] didrocks, cool, thanks ;) [14:18] bcurtiswx, there you can set when you want to show a battery indicator [14:19] ah ya know what, could be GNOME3 PPA.. i have that gnome-power-manager [14:21] chrisccoulson, cyphermox. a laugh for today http://paste.ubuntu.com/558123/ [14:22] seems GPM still has some GTK2 symbols [14:22] ah, gnome-power-preferences has gone away now [14:22] it's a g-c-c panel instead [14:22] that's why you see that :) [14:22] i'd imagine those symbols are in another library it's linking against [14:28] tedg, I have some indicator nit fixes lined up. When is next release? [14:29] mterry, Thursday. [14:29] tedg, ok, hopefully I'll have merges filed today. Nothing major yet. I'm trying to clear out the critical warnings when running the panel service [14:30] mterry, Sweet! Good stuff. [14:30] for example, I found the darn code that prints out the 'isn't an int32 variant' :) [14:31] chrisccoulson, http://imagebin.org/134333 [14:31] chrisccoulson, my battery is unplugged mind you [14:32] mterry, Cool, it's easiest if you put each fix in it's own branch. What I usually do then is have a single packaging branch that merges those in so that I can test it on my own machine, but then I can propose them all separately. [14:33] tedg, ok [14:33] tedg, they're in lots of different packages anyway [14:34] mterry, Heh, that makes it easier to keep them separate :) [14:42] hi chrisccoulson, do you think you could have a look at adding the FF translations to the package as we discussed in the Rally? [14:43] mvo: do we still need software-properties-gtk in the software-center world, or was this integrated into sc? [14:43] mvo: if so, want me to port it to pygi? [14:45] hi guys, can I (please) have a sponsorship for https://code.launchpad.net/~nataliabidart/ubuntu/natty/ubuntu-sso-client/ubuntu-sso-client-1.1.9/+merge/47399? [14:48] hey nessita [14:48] nessita, does it fix the sso dialog on session start? ;-) [14:49] nessita, I can do the sponsoring [14:50] pitti: feel free to port, but we plan a big push toward a better UI for O [14:50] pitti: its still needed and will be needed in the future [14:50] mvo: ok [14:50] but the UI and the code will change as it will (finally) be ported to policykit :) [14:52] dpm - yeah. i want to figure out what mercurial repo they use to build the xpi's really though, and use that rather than shipping xpi's [14:52] seb128: nopes, the release of u1client will. It should go out today, we already requested dobey to do it. We had to delayed it due to some issues on syncdaemon [14:52] seb128: sorry :-( [14:52] dpm - the issue with merging xpi's in to the firefox source is that we always build the firefox pre-releases, and there are no xpi's at this stage [14:53] i don't want to block updates on waiting for translation xpi's ;) [14:54] mvo: are you a big fan of SimpleGtkBuilderApp? I. e. do you want to keep it, or can I kill it while porting? [14:54] chrisccoulson, I thought XPIs were always at http://ftp.mozilla.org/pub/mozilla.org/firefox/releases/ - that's where we get them when we upload them manually [14:55] dpm_ - that is for release builds. we prepare our source tarball and uploads before they are released [14:55] i'm just doing b10 now [14:56] nessita, no worry, I was just checking because jasoncwarner asked me if that was going to be fixed for alpha2, it happens on the livecd still [14:57] seb128: yeah, it will [15:02] chrisccoulson, ok, so if we cannot import them like that, what about other options of automating the import? What about what we talked about using the same method to generate the en-US template to build the xpis? We've uploaded the translations manually this time, but I do not follow the FF releases that closely, and we cannot guarantee that they are manually imported before every upload [15:03] dpm_ - building the xpi's using the code from mercurial seems like the right way. i just need to figure out their release process for translations, and also whether the build system supports building multiple languages too [15:03] nessita, happy birthday ;-) [15:04] seb128: thanks! am I getting a cake too??? [15:04] happy birthday nessita! [15:04] ooh, happy birthday nessita! *hug* [15:04] seb128 - it's not your birthday today too? [15:04] ;) [15:04] lol [15:04] nessita, next time we see each others maybe you get one ;-) [15:05] chrisccoulson, I don't think it's nessita's one either ;-) [15:05] awesome! [15:05] I just sponsored her update [15:05] lol [15:05] which is like a birthday present [15:05] * seb128 hugs nessita [15:05] * nessita hugs seb128 [15:07] * desrt /join #hugfest [15:08] chrisccoulson, sounds good. It would be awesome if we could automate this. Is there anything I can help with? [15:08] (that doesn't require becoming an expert in the Mozilla build system? ;) ) [15:08] felicidades nessita! [15:09] :-) [15:19] rodrigo_: I may not have time today to talk about CouchdbDocument - I'll be on call for at least an hour or so, and I know you finish the job around 6pm or so. If I don't get back in an hour or, can we talk tomorrow? [15:20] kklimonda, sure :-) [15:20] or so [15:20] ;) [15:27] KenEdwards: you lied. [15:27] err [15:27] kenvandine: ^^ you lied rather. [15:27] KenEdwards: sorry :) [15:27] ? [15:28] kenvandine: midori doesn't only do the yellow bar by strcmp on 'https' in the url [15:28] dobey, i just read that in a bug report :) [15:28] dobey, np ;-) [15:29] dobey, have they implemented a UI for managing them? [15:29] kenvandine: it's actually tri-state with newer libsoup/webkit. it checks the SoupFlags on the message for _TRUSTED_CERT [15:29] kenvandine: no, there's no way to manage them yet, but untrusted certs it makes the urlbar background red instead of yellow [15:29] ah... better than nothing [15:29] someone should do the same in epiphany [15:30] although i think it might still be doing that wrong, but haven't looked into it much further yet [15:30] from python i don't seem to have access to the SoupRequest [15:30] and midori crashes more than windows with a virus [15:31] and it's not really the ui i want, so still going to bring back encompass [15:36] kenvandine: you can probably do the same thing midori does, from python, if you use introspection instead of bindings? [15:36] dobey, yeah, should be able too [15:36] not sure which is more work though [15:36] doing that, or opening the default browser for oauth [15:38] does it matter? [15:38] the user experience should be more important than what is more work [15:43] hi, i'd like to replace my nvidia video card to a better one. How do i check which card is currently actuall installed? [15:45] xw4400: #ubuntu is the help channel, you will probably get better response there. [15:46] ah, ok. This is a dev channel? [15:49] xw4400: some like this [15:55] nessita: still need the sponsoring? (was in meeting until now) [15:56] pitti: not right now, seb128 did it [15:56] ah, great [15:56] I will need another one for u1cp in a few hours [16:08] wow! didrocks is an archive admin! [16:08] promotion? [16:09] more like "dumped on him while he wasn't on guard" [16:14] :) === bjf[afk] is now known as bjf [16:18] * bcurtiswx_ bows down to didrocks ;) [16:19] * seb128 dumps taks on didrocks, welcome! [16:20] tasks [16:20] seb128: nooooooooooooooooooooooooo :-) [16:20] I still don't have my shell access, so I can't do anything :) [16:20] you can NEW from the webui [16:21] oh right, I should have a look to that NEW thing :) [16:22] session restart, brb [16:22] chrisccoulson, is there a known problem where firefox becomes "undecorated" [16:23] kenvandine, what problem? not that i'm aware of ;) [16:23] for some reason firefox just lost it window decorations... but none of my other windows did [16:23] it is still working fine though [16:23] gotta be a compiz bug :) [16:23] hmmm, that sounds odd ;) [16:23] but weird... [16:24] does it come back if you switch to full screen and back again? (ie, hit F11 twice) [16:24] oh, i've just noticed that full screen isn't really full screen anymore [16:25] unity bug though ;) [16:27] chrisccoulson, oh... so is full screen like 2/3 of the screen with no decorations? [16:28] kenvandine, the panel is still displayed with full screen now [16:30] not for me [16:30] but i don't have decorations when not full screen :) [16:31] DING DING DING meeting o'clock [16:31] hello! [16:32] :) [16:32] * kenvandine waves [16:32] no Jason today? [16:32] no [16:32] LCA [16:32] heya [16:32] hey [16:33] bryceh, chrisccoulson, didrocks, tremolux, Riddell, cyphermox, mterry, rodrigo_, seb128, tkamppeter: meeting time [16:33] hi [16:33] hi! [16:33] hi [16:34] hey [16:34] apparently no actions from previous meeting [16:34] https://wiki.ubuntu.com/DesktopTeam/Meeting/2011-01-25 for the current reports about Unity etc. [16:35] kenvandine: do you have a parnter update? [16:35] yup [16:35] i should put that on the wiki too :) [16:35] partner, too [16:35] anyway [16:35] of U1 [16:35] for U1 [16:35] syncdaemon performance improvements, should land end of next week [16:35] unity integration has begun, should land in time for FF assuming all goes well with the libunity api they need [16:36] shotwell integration work will start next week, minimal work required, also should land in time for FF [16:36] nice [16:36] DX: [16:36] like, publishing photos to the cloud? [16:36] libunity landing this week, hopefully enough to get the U1 team comfortable with the API [16:36] pitti, basic [16:37] making sure photos can be synced easily [16:37] like i think things like revision info and metadata about photos [16:37] hi [16:38] the other big thing from DX this week is work on moving menu parsing out of appmenu-gtk and libappindicator and into dbusmenu-gtk [16:38] which should help with bug triaging and hopefully fixing [16:38] that is all i have, questions? comments? [16:38] kenvandine, libunity landed last week no? [16:38] kenvandine: (libunity already landed btw) [16:38] kenvandine, well at least didrocks uploaded and I reviewed it [16:38] yeah... but not with the new PAPIs [16:38] APIs [16:38] ok [16:38] i think the beginning of the new APIs are landing this week [16:38] so new update you mean [16:39] yeah, not new package :) [16:39] kenvandine, is it worth trying to fix appmenu issues until the parsing update lands? [16:39] like lpi menu items not working [16:39] don't think so [16:39] ok [16:40] if there are appmenu issues to debug btw, i could take a look at them whilst this stuff is all fresh in memory :) [16:40] i think a few people mentioned at the rally there were some long standing bugs [16:40] there is indeed, feel free to pick on the bug list [16:40] there are, i think tedg hopes to get that code in one place so it is easier to triage and fix [16:40] cool, i'll take a look [16:40] not sure how much that will be useful with the refactoring though [16:41] so maybe check with ted to get the updated code before starting on those [16:41] sure, no problem [16:42] oh, one more thing [16:42] kenvandine: so you'll add to the wiki as well? thanks [16:42] libappindicator gtk3 port is up for review now [16:42] so should land this week [16:42] \o/ [16:42] happy to test [16:42] great [16:42] :) [16:43] kenvandine, up to upstream review, i.e ted? or packaging one? [16:43] pitti, will do [16:43] I can help to review [16:43] upstream [16:43] ok [16:43] didrocks: do you want to paste/say something extra/discuss unity updates? [16:43] well, people can read on the wiki I guess [16:43] but I have a BREAKING news! [16:43] we need to read now? how boring! [16:43] (which is more a fix in fact) :) [16:44] ok, copy/paste then, just for seb128 ;) [16:44] New unity landed this week [16:44] * a bunch of bug fixes and the dash has its first release (really early prototype though)! [16:44] a lot of accessibility improvments as well. [16:44] we got an intermediate upload to fix an annoying "invisible unclickable window" bug. [16:44] the places have been rebuilt to drop some packages from the archives, the big come back is for this week release. [16:44] still some stacking issues (menu disappearing, windows not clickable) in compiz. Apparently, there is a fix proposed, so fortunately, this will be fixed this week. [16:44] also, some specific issue with nvidia card (black screen) still there, as well as launchers weird behavior in some cases. [16:44] libunity is now a new source to make the foundation API more unrelated to the view. [16:44] -> Next week, on your almost natty alpha2 computer: [16:44] full release upload planned on Thursday with working places (\o/), miscellaneous improvements and still bug fixes. [16:44] didrocks: do we need to discuss the detection? right now it seems to be painfully expensive [16:44] another "final" upload on Monday evening/Tuesday morning (european time) with crusty fixes and small incremental changes for getting ready for alpha2. [16:44] a compiz upload which should speed the startup time and fixes some annoying bugs (no ABI break planned). [16:44] -> misc: [16:44] we know we have a lot of false positives of "your computer can't run unity" because the detection tools take more than 3s to answer on a cold cache (and the timeout was 0.5s, then 1s). For now, I'll upload a gnome-session package with a timeout of 5s for the release. The detection tool loads 15 MiB of libraries which can explain that. Some work on that planned for alpha3. [16:45] support of multiple fallbacks added to gnome-session. So, if you install unity-2d, the fallback for the unity session is unity-2d instead of gnome-classic without a warning message in that case. [16:45] didrocks, sorry I was just joking since IRC or wiki we need to read anyway but thanks ;-) [16:45] lot of nice things coming this week [16:45] first the news ;) the stacking issue (some menu disappearing in the background or window taht you can't click) is fixed! [16:45] nice! [16:45] just a few minutes ago :) [16:45] yay [16:45] * chrisccoulson hugs didrocks [16:45] \o/ [16:45] chrisccoulson: hugs smspillaz rather :) [16:45] he put a lot of effort in it [16:46] congrats to him! [16:46] * chrisccoulson hugs smspillaz [16:46] yay smspillaz! [16:46] wasn't trivial ;) [16:46] nice smspillaz! [16:46] pitti: re: detection [16:46] smspillaz: rock! [16:46] pitti: one of the issue is that we are linking against 15Mb (the full nux stack) [16:46] didrocks: what does it currently do? [16:47] just for basically checking opengl capability + lspci [16:47] it's not enough to check the graphics driver and available X.org modules? === nessita1 is now known as nessita [16:47] apparently not, they want to ensure the full nux is woking [16:47] working* [16:47] and make some tests [16:47] again, the tests are not expensive [16:47] initializing nux is [16:47] pitti: :) [16:47] didrocks: so in sum they are [16:48] yeah :) [16:48] so, I've talked about it with jay [16:48] we'll figure out for alpha3 [16:48] ok, thanks [16:48] if it's still expensive, the plan is to run it in gdm [16:48] but even 3s is a lot… [16:49] such a thing should require .1 s.. [16:49] and to cache it (I can patch gnome-session for picking it pretty easily) [16:49] well, the gnome one is taking .3s [16:49] (on cold cache) [16:49] ok, let's discuss off-meeting [16:49] right, in a nutshell, we have a workaround for alpha2 [16:50] looking forward to places [16:50] yeah :) [16:50] tremolux: do you have something extra to say/discuss wrt. s-c? [16:50] nice work on startup speed! Is http://people.canonical.com/~mvo/software-center/mini10-startup/startup-times.png created automagically now? [16:50] nope, unless anybody has questions ? [16:50] < 3 s is great [16:51] pitti: we still run it by hand :) [16:51] pitti: but yes, thanks! it feels really fast now [16:51] pitti: also, thanks for your cold-cache script, I tried it and it seems to do the trick [16:51] tremolux, I thought ratings and reviews was supposed to drop last week? [16:51] tremolux: nice [16:52] tremolux, but it's next week? [16:52] mterry: yes, mvo was at a sprint all week so we held off [16:52] mterry: in other words, I mis-promised ;) [16:53] mterry: but we will have it for alpha 2 [16:53] np, just curious [16:53] sure, np [16:54] Riddell: anything to announce/discuss in Kubuntu land? [16:54] hi [16:54] Riddell: still trouble with gstreamer/gvfs and the like? [16:54] * 4.6.0 being tested now, due for upload tomorrow morning [16:54] * Digikam tested now, due for upload tomorrow morning [16:54] * KDevelop being packaged, due for upload tomorrow morning [16:54] * 17 milestoned bugs for alpha 2, they won't all make it but we can try, http://goo.gl/yGhJd [16:54] * slow but steady progress being made on todo items https://wiki.kubuntu.org/Kubuntu/Todo [16:55] wohoo, release [16:55] slomo said he'd remove the depends on gvfs, and I blacklisted it for now anyway [16:55] s/depends/recommends/ [16:55] I got libproxy to build after fixing the uninstallability, so tomorrow's CDs should be a bit smaller [16:55] we removed geoio-database which should save a wee bit too [16:56] blacklisted rather [16:56] Riddell: cool; as already said, feel free to upload that right away; we can sync the next time anyway then [16:57] thanks [16:57] any other business? [16:58] so, thanks everyone! [16:59] thanks all [17:00] thanks [17:01] thanks === asac_ is now known as asac [17:18] seb128: do we have -shell PPA instructions for people yet? [17:29] jcastro, don't ask me? [17:30] jcastro, ricotz might know [17:30] jcastro, the GNOME3 ppa doesn't have gnome-shell yet so not sure how we could have instructions [17:30] * jcastro nods [17:34] seb128, jcastro, there are no official introstructions besides some forum-guides using my ppa [17:34] ricotz, do you plan to get it in the gnome3 ppa as well? [17:36] seb128, i could look into it, now that clutter 1.5 is available [17:36] ricotz: can you ping me when it's in there? i'd like to sort out all the people trying random blog entries with older PPAs and all that [17:39] jcastro, yeah i can [17:39] ricotz, is there any depends you still need? [17:39] how do you do daily builds if the versions required are not there? [17:39] seb128, gjs and mutter [17:40] seb128, i also updated clutter [17:40] why can't you updated those in gnome3? [17:40] seb128, you can see which packages are needed looking in the ppa [17:40] ricotz, couldn't you update those in the team ppa rather than in yours? [17:41] so it would benefit others... [17:41] seb128, maintaining the git snapshots is easier than document all changes of every release :P [17:41] seb128, so you want the git snapshots in the gnome3 ppa? [17:42] heh [17:42] not really no [17:42] there is a gjs tarball every 5 weeks or so [17:42] seems it should be doable to package tarballs in the ppa no? [17:43] do you know the reason for the different handling of xulrunner between debian and ubuntu [17:43] you mean libmozjs? [17:43] it's not a public library and mozilla doesn't garanty stability [17:44] it's broken to ship it as a normal library [17:44] seb128, yeah that is possible, it is quite near the debian packaging (gjs, mutter, clutter) [17:44] ricotz, yes. debian carry patches to version libmozjs and install it in /usr/lib. we don't to maintain ABI compatibility with upstream [17:44] chrisccoulson, ok. didnt know that they maintain the abi stability [17:45] ricotz, well, there is no ABI stability really. debian just add a patch to add a version number to libmozjs.so, and they bump the version when ABI breaks [17:46] chrisccoulson, ok, and ubuntu didnt wanted to go that way? [17:46] ricotz, we used to, but we try and stay as close as possible to upstream [17:46] alright [17:46] by default, libmozjs isn't even built as a separate library anymore [17:47] it's all bundled statically in libxul [17:47] Of course, the way Debian is doing it is the way it should be done upstream as well :) [17:48] Amaranth, not really. upstream don't ship spidermonkey as a separate product. it's part of the toolkit instead [17:48] if they shipped it as a library for people to link against, then perhaps ;) [17:49] that might happen in the future, but i wouldn't really want to maintain yet another package [17:49] Huh, I guess the webkit stuff is the same way [17:49] You have to link to all of it to get JSC [17:49] seb128, anyway, i will try to look into uploaded the releases [17:49] Amaranth, oh, i didn't know that [17:50] ricotz, thanks [17:50] well, upstream seems to be leaning in the direction of properly versioning spidermonkey [17:50] that sounds like the same for mozilla now (having to link against libxul), but we actually changed the build config to still build libmozjs [17:50] gnome-shell linking against libxul would be hideous for startup time ;) [17:50] chrisccoulson: Yeah, there is just a /usr/lib/libwebkitgtk-1.0.so.0 [17:50] that sucks! :) [17:51] seb128, could you ping robert to update clutter to 1.5.14? [17:51] micahg, yeah, as a separate product though [17:51] i don't want to maintain another copy of spidermonkey ;) [17:51] I still can't believe GNOME is going forward with depending on two browser engines, seems weird [17:51] chrisccoulson: I thought it would be in tree and we can just check the appropriate files on build [17:53] ricotz, he's at a conference this week (lca) [17:53] seb128, ah, there is one thing with clutter which breaks the build of gnome-shell, it is the rdepends patch which removes x11 from clutter-1.0.pc [17:53] ricotz, I will try to do it [17:54] (the backend/frontend patch) [17:57] ricotz, is that coming from debian? [17:58] seb128, no [17:58] bbl [18:00] ok, will check that [18:13] mvo: got software-properties-gtk mostly working with pygi; missing is the drag&drop (that only works with gtk3, but no biggie IMHO), and getting the XID of a window (not available; but I tested the apt cache updating, and the main window is disabled anyway, just like in the current version) [18:17] kenvandine, mterry: could one of you try to figure why launchpad-integration failed to build? [18:18] it's a gir issue [18:18] didn't happen on my box but I'm not uptodate with changes from this week [18:19] seb128, sure [18:21] seb128 - bug 707543 - quality report! [18:21] Launchpad bug 707543 in gnome-terminal "no work" [Undecided,New] https://launchpad.net/bugs/707543 [18:21] it looks like a firefox bug ;) [18:23] lol [18:23] go users [18:24] chrisccoulson, you should stop reading bug reports now ;-) [18:24] mterry, thanks [18:24] seb128 - i almost have already ;) [18:24] i only noticed that because it's my most recent e-mail [18:24] good night everyone! [18:26] 'night pitti [18:26] haha [18:27] good night pitti [18:29] seems the launchpad guys fixed the bug watch code [18:29] it's spamming me now [18:31] kenvandine: ping. do you know why rhythmbox-ubuntuone-music-store has rhythmbox-plugins as a Depends? [19:01] ok, time for dinner, see you tomorrow! [19:01] dobey, i don't know off hand [19:08] kenvandine: hrmm, ok [19:11] dobey, rhythmbox-ubuntuone-music-store had been added before rhythmbox-plugins was split out of rhythmbox [19:11] ok, weird [19:17] seb128, are you seeing bug #707478 ? [19:17] Launchpad bug 707478 in launchpad "Freedesktop bug watches are (incorrectly) updated with "importance unknown"" [High,Triaged] https://launchpad.net/bugs/707478 [19:17] pitti: hey, you still around to sponsor an upload? [19:17] nessita, he's not [19:17] seb128: oh, ok. Would you be able to? [19:18] if not, I'll ask again tomorrow [19:18] nessita, yes I can do that [19:18] bryceh, I will need to check I just got spammed by bug watches updates today and deleted those [19:18] so I will need to wait for extra ones [19:19] seb128: https://code.launchpad.net/~nataliabidart/ubuntu/natty/ubuntuone-control-panel/ubuntuone-control-panel-0.8.0/+merge/47447 thanks! [19:19] seb128, probably you won't get more [19:19] bryceh, I still have those concerning the team assigned bugs though [19:19] which seems to confirm your bug [19:19] nessita, you're welcome [19:19] seb128, I also got spammed, but all the mails seemed to be from freedesktop.org's bugzilla, so it may be localized to just that tracker. If you're seeing it on any GNOME bugs, that would be an important data point [19:21] bryceh, no, it seems the GNOME just got a normal update [19:21] it's likely that the update didn't run for a while has they got it working today [19:21] those have a correct status it seems [19:22] bryceh, but I can confirm your fdo issue [19:25] seb128, thanks [19:27] pitti: nice, thanks [19:35] nessita, ok, uploaded, ignore the email you will get about ubuntu-sso-client_1.1.8, I ran dput on the wrong file in my ubuntuone sponsoring dir [19:39] seb128: thanks! I'll be testing on my prisitine natty box when the binary is built [19:39] ok [19:58] seb128, btw anything more we need to do on bug #675622 so it can get a sync? [19:58] Launchpad bug 675622 in glew "Please sync glew 1.5.7-1 from Debian experimental (main)" [Undecided,New] https://launchpad.net/bugs/675622 [19:59] seb128, https://code.launchpad.net/~mterry/launchpad-integration/fix-packages/+merge/47453 [20:37] bryceh, subcribe ubuntu-archive or ping Riddell it's his archive day [20:37] mterry, why can't you upload that? [20:37] seb128, it's not in ubuntu-desktop set [20:37] mterry, email cjwatson ;-) === charlie-tca__ is now known as charlie-tca [20:38] seb128, OK :) [20:38] mterry, did you check it's not in the desktop set? [20:38] it's surprising [20:39] seb128, I did, let me double-confirm [20:39] seb128, ahem, my mistake. [20:39] ;-) [20:41] bigon, there is no need to manually sync your updates yourself, sync request are reviewed daily nowadays [20:42] mterry, is that change a recent one? is that likely to be requiring updating other sources? [20:42] seb128: ah ok [20:45] seb128, I'm not sure. gobject-introspection got a new upstream version updated on the 13th, it could have happened then [20:46] seb128, I'm not sure exactly how it used to behave. I'm a little confused it used to work as it was -- not sure how it could have known to link against gtk3 when building the typelib [20:47] or maybe it didn't check for missing symbols. That's probably more likely. And it was just broken the way we did it, but didn't stop the build [20:47] ok, the fix seems correct I was just wondering if we should review other sources for similar issues [21:00] seb128, right, and perhaps it would be wise, but we don't necessarily have reason to believe other people screwed up like I did when setting that _PACKAGES line originally. We'll hear about it when they FTBFS, eh? :) [21:00] right, doko was speaking about doing an archive rebuild test again so let's see what's coming from it [21:01] kenvandine, gtk+2 still fail to build btw [21:01] humm [21:01] built in pbuilder for me... i'll look [21:01] kenvandine, thanks [21:10] empathy natty 2.32.2 looks for gcr-0 , has that been renamed? [21:10] libgcr-0 , libgcr-3-0, libgcr-dev [21:10] all tried [21:12] bcurtiswx_, check gnome-keyring in the ppa but could be that the version is not recent enough [21:13] seb128, this has to be without the gnome3 ppa [21:13] im backporting some patches to the version that will be in natty [21:13] oh 2.32 [21:13] seb128, yup [21:14] it's libgcr-dev [21:14] seb128, yea it's installed and configure doesn't see it :-\ [21:14] bcurtiswx_, what error do you get? [21:14] seb128, i wonder if since i have the PPA as a apt-source if there's something with that.. lemme get the config.log [21:15] it's likely [21:18] seb128, http://pastebin.com/YucxwAA6 [21:19] bcurtiswx_, what libgcr-dev version is installed? [21:20] seb128, ah ha Installed: 2.92.92.is.2.91.4-0ubuntu1~build2 [21:20] ok, that's the issue [21:20] disable the ppa source for it [21:22] ok, I need to go, see you later [21:22] seb128, OK have a good night [21:31] kenvandine, hey dude [21:31] so, I notice evince is not using the app menu? [21:32] should I file a bug? [21:32] hey rickspencer3 [21:32] one sec [21:32] it is for me... [21:39] rickspencer3, launched from firefox? [21:39] chrisccoulson, correct [21:42] rickspencer3, http://bazaar.launchpad.net/~mozillateam/firefox/firefox-4.0.head/revision/740 [21:42] you just reminded me ;) === ivanka is now known as ivanka-train [21:42] oops, the changelog entry doesn't make sense ;) [21:50] chrisccoulson, haha... as always, a firefox bug :) [21:51] lol [21:52] kenvandine, the "unset UBUNTU_MENUPROXY" but was inherited from the nightly builds, which need it because appmenu-gtk makes them crash on startup [21:52] so..... [21:52] appmenu-gtk bug! [21:52] ;) [21:55] rickspencer3, ok, just uploading a fix for that now [22:00] where is the gnome certificates file located? [22:01] bcurtiswx_, what do you mean? [22:01] i can never remember, cacertificates.ca or something like that. [22:01] bcurtiswx_: i don't think gnome maintains a certificate keychain of its own separate from the system. the system's certificates are maintained by the ca-certificates package [22:02] bcurtiswx_ - it depends, there is more than one source of root certs [22:02] ca-certificates being one, and NSS being another [22:02] it ca-certificates [22:02] it's* [22:02] well [22:03] evo uses NSS [22:03] are we doing the Eastern Desktop Meeting today? I know it's an AU holiday so pretty much everyone's gone :-) [22:04] chrisccoulson, /etc/ssl/certs/ca-certificates.crt [22:05] bcurtiswx_ - what are you trying to do? [22:06] chrisccoulson, empathy is looking for the ca certificate file it has(d) a with-ca-certs='' with 2.32.2 (not with 2.91.x) [22:06] ah, ok [22:06] chrisccoulson, what took so long? [22:07] I ask a quetsion at 13:31, and you are uploading a fix 13:55 [22:07] a 24 minute lag :) [22:07] rickspencer3, network ;) [22:07] ffox is nearly 60MB [22:07] it takes a while to upload on my connection ;) [22:07] haha [22:08] thanks chrisccoulson [22:08] you da' man [22:08] :) [22:09] bryceh, is it safe for me to dist-upgrade right now? [22:09] or did your new xorg goo start kicking in? [22:13] rickspencer3, new libdrm and mesa are in [22:13] hmmm [22:13] maybe I'll wait until Friday ;) [22:13] we have no reason to think that there are problems, but you know how mesa and libdrm updates are ;-) [22:13] get the new xorg and the next unity stuff [22:13] oth, maybe I should take the plunge :) [22:14] I can always fall back to my netbook [22:14] bryceh, btw, the xorg team is absolutely rocking ... [22:14] edgers seems like a really fun and engaged group of contributors too [22:15] rickspencer3, yep it's going good this cycle [22:17] rickspencer3, which is fortunate since Unity and HWE need X solid [22:17] for sure [22:20] how is app-install-data-ubuntu created? [22:20] woah, i only just uploaded ffox 4 b10, and mozilla are talking about starting b11 builds on friday! [22:21] they're serious about a feb release [22:21] if I want to add something there who should I poke about it? mvo? [22:21] kklimonda, it's in bzr somewhere [22:21] just grab it, add what you need and then ping mvo :) [22:22] i can't remember where it is though ;) [22:22] no proble, I'll find it [22:22] thanks [22:24] hey guys [22:24] hi lamalex [22:25] rickspencer3, i am current and it seems fine for me [22:26] coolio [22:34] so I have to make a dbus service file to integrate with the sound menu? [22:39] jaytaoko, welcome to desktop [22:40] hello! [22:40] jaytaoko, I guess due to the AU holiday the meeting didn't get started, but you and I can chat about unity bugs now if you want [22:40] jaytaoko, you probably saw my email that libdrm and mesa are uploaded now [22:41] bryceh: actually I didn't [22:41] those are pretty much the same versions that have been in xorg-edgers (but rebased against actual releases) [22:41] bryceh: when did you sent it? [22:41] bryceh, I'm working on the autopilot rewrite now. It will be about 100 times more awesome in a couple of days [22:41] RAOF, ^ [22:41] jaytaoko, aha then good thing we're meeting! [22:41] i'm trying to track down an issue with a 1x1 InputOnly window that seems to follow the mouse around the desktop (except when the mouse is grabbed or something) [22:41] jaytaoko, it was sent to ubuntu-x@ (which you should subscribe to) and to ubuntu-devel@ [22:41] anyone know what app might be managing that? [22:42] jaytaoko, please forward it to any unity/compiz lists where you think others may have missed it [22:42] bryceh: sure [22:42] it is the child of the window with the pointer [22:43] jaytaoko, hopefully you guys have been running xorg-edgers so already have gotten testing done on it, but if not then you can now :-) [22:43] bryceh: I upgraded my intel system, it has the new xorg now [22:43] jaytaoko, we will be following up with xserver 1.10 upgrade in a couple days probably [22:44] bryceh: I am compiling Unity on it [22:44] that will break driver ABIs so we'll be rebuilding all of those too [22:44] bryceh: ok [22:44] and we have new -intel, -ati, and -nouveau we've been holding off on since they need new mesa and so on [22:44] so... lot of change coming this week [22:44] bryceh: I want to find out if the glsl compiler is working as expected for intel [22:45] bryceh: about fglrx, have you had some update from Felix? [22:45] jaytaoko, *nod* I've heard there were some last minute changes there (due to licensing or something) [22:45] bryceh: licensing about the glsl compiler? [22:46] jaytaoko, yes he mentioned wanting to put you into direct contact with one of their engineers in China [22:46] jaytaoko, but seems this is going to be heavily NDA'd so best for you to work through Henry Hall on that [22:46] bryceh: alright! [22:47] bryceh: I wonder if on their own they were able to discover something about the fglrx driver. [22:47] bryceh: there is this unknown regarding the texture_from_pixmap extension [22:48] bryceh: according to sam, on the fglrx driver we cannot have direct rendering + texture_from_pixmap extension [22:48] *nod* [22:49] jaytaoko, will you be able to work around that via blacklisting the driver from doing direct rendering for the time being? [22:49] bryceh: ok [22:49] bryceh: we have to detect that a system is using the fglrx driver [22:49] jaytaoko, fwiw when we update to xserver 1.10 that will permanently break fglrx until AMD releases an updated driver to us [22:50] bryceh: :-( [22:50] in the past they've sent us an update right prior to release, so there won't be much time left for testing after that [22:51] so you may want to hang onto the current daily iso for testing [22:51] also AMD may be able to provide NDA'd access to the testing driver; talk with Henry if you need that [22:51] bryceh: ok. I will [22:52] jaytaoko, Alberto and Sarvatt also have access to that driver so you could alternatively ask one of them for access. But may be better to work through Henry even so. [22:52] on to bug reports... things have been going really well so far [22:52] bryceh: since AMD support the open source driver will it be ok if we recommend the open source driver to users instead of the fglrx [22:53] jaytaoko, that is fine by me, however be aware for the latest hardware the open source -ati driver may not give full 3D support, so users on that hardware would need -fglrx for that [22:54] bryceh: ouch! [22:54] jaytaoko, however I should think 2D unity should work ok on the open driver in such cases, which might be more practical to recommend [22:54] jaytaoko, http://wiki.x.org/wiki/RadeonFeature gives a table showing which hardware has what level of 3D support currently [22:55] we will be doing an update to a current git snapshot of the radeon driver, so what we have in natty should match what's shown in that table pretty closely [22:56] regarding bugs, here's the natty bugs on our todo list currently: http://blumonc/Arsenal/Reports/ubuntu-x-swat/workqueue-natty.html [22:56] bryceh: alright [22:56] or I should say, those are all the actionable X bugs [22:56] which is an amazingly short list [22:57] bryceh: link does not work [22:57] I charted the # X bugs in natty against the equivalent point in development for maverick: [22:57] sorry, should be http://www.bryceharrington.org/Arsenal/Reports/ubuntu-x-swat/workqueue-natty.html [22:57] http://www.bryceharrington.org/X/Reports/ubuntu-x-swat/totals-natty-workqueue.svg [22:58] bryceh: we're going to pull in an ati git checkout soon aren't we? we will have 3D support for everything but the latest 69xx GPU's then [22:58] the dark line in that chart is # bug reports in maverick, so you can see things are really stable [22:58] Sarvatt, yes that's correct [22:58] it will be interesting to see how that chart changes once all the new X stack is in place :-P [22:58] but hopefully it won't be too bad [22:59] jaytaoko, are there further X / driver bugs on your end that need our attention? [23:00] bryceh: what about suspend/resume? [23:01] bryceh: I wonder what we could do on the software side to have this smooth on Nvidia GPU [23:01] jaytaoko, most often resume issues have been kernel-side troubles. Currently I'm not aware of any X-side suspend/resume issues that we're tracking, but I can take a look if you have bug #'s [23:02] there are several blank-screen and gpu lockup bugs (also probably kernel bugs) which we're tracking in X (see above report), some of which may relate to suspend/resume in some fashion. [23:03] bryceh: let me check... [23:03] jaytaoko, one thing that's been a constraint on the kernel side for dealing especially with the nvidia-specific hangs, is that they don't have the hardware to test with [23:04] kenvandine, xchat-gnome to indicator-appmenu , do you know if thats coming or in plans? [23:04] jaytaoko, which I think could be remedied pretty easily via shipping a few of them some nvidia gfx cards... [23:04] bcurtiswx_, working fine here... [23:04] * bcurtiswx_ kicks his laptop [23:05] although I'm not certain suspend issues are so easily reproduced on desktops, but at least it wouldn't cost as much as a laptop to provision [23:05] bryceh: I have to find that bug: it is aboout some garbage on the desktop after a resume on Nvidia [23:05] kenvandine, resetting unity may help? [23:05] jaytaoko, ok. shoot it to me via email or irc when you find it and I can take a look [23:05] bryceh: the only system that has caused me troubles with suspend /resume is nvidia [23:05] jaytaoko, other issues? [23:06] jaytaoko, yeah fwiw nvidia hw has always been problematic with resume [23:06] bryceh: resolution change and multi monitor [23:06] bryceh: how do we approach that? [23:07] bryceh: Are we rendering into one big surface and X take care of the rest for multi monitors? [23:07] bryceh: or is ther something we hace to do in code? [23:07] jaytaoko, as I understand it, a lot of logic and heuristics were built into gnome (metacity?) to handle this, while compiz has a different codebase for handling it [23:07] so a thought there is to look more closely at the way it was done in gnome, or even port/reuse the logic. [23:08] bryceh: is there some X documentation about how to handle it? [23:08] jaytaoko, there is `man 3 Xrandr` but that's pretty low level [23:08] unclutter was the app creating funny windows and sending bogus events; anyone know whether this is often installed on ubuntu? [23:09] jaytaoko, maybe my best suggestion is to look through the gnome-rr.c code in libgnome-desktop [23:10] bryceh: I have to talk to sam about the support in compiz [23:10] it provides a wrapper around the lower level Xrandr library calls [23:10] bryceh: writting this down... [23:10] bryceh, jaytaoko: there are also GdkScreen methods for querying that sort of information [23:11] jaytaoko, also RAOF has looked at this stuff more recently than I so would be a better reference for it [23:11] and the GdkScreen methods work for nvidia-current et al. which use Xinerama and not Xrandr [23:12] broder: is Xinerama another api similar to XRand? [23:12] broder, good point [23:12] jaytaoko, yes [23:12] jaytaoko: it's basically the precursor to Xrandr, but NVIDIA has never been good at keeping up with the times [23:12] there is also TwinView, which is nvidia-specific [23:13] right, and i don't think any of the gdk/gtk/gnome code knows about those extensions [23:13] xserver 1.10 will have some randr updates that apparently make it possible for nvidia to finally join the crowd at supporting xrandr [23:13] however that'll depend on them prioritizing the feature development work in their driver and getting it to us on time [23:13] (I don't know if there is a roadmap for this feature, so am not counting chickens yet) [23:13] bryceh: so are we looking at several implementation depending on the vendor of the gpu? [23:14] jaytaoko, yes. I would suggest prioritizing Xrandr 1.3 at the top. [23:14] byceh: ok [23:14] as far as i know the GdkScreen monitors will work against all drivers [23:14] bryceh: so what is gnome doing when running on an nvidia system? [23:15] but you can only query information through those, not modify it [23:15] and like broder says, you might be able to use the Gdk api and get it all for free [23:15] jaytaoko, like broder points out, they have an abstracted api in gdk that covers it [23:15] bryceh, jaytaoko: i've definitely verified that gdk_screen_get_n_monitors and gdk_screen_get_monitor_geometry work against fglrx, nvidia-current, -intel, -ati, and -nouveau. i don't know about the rest of the api though [23:16] (i'm going to go test gdk_screen_get_primary_monitor against nvidia-current right now, because if that works it would make my life a lot easier) [23:16] broder, iirc some of the resolution switching stuff isn't quite as nicely abstracted. But it's been a long time since I dug into it and things may have changed. [23:16] bryceh: changing resolution is handled through the GnomeRR* stuff, not the GdkScreen stuff [23:17] broder: thanks for the info [23:17] and yeah, the GnomeRR* stuff is exclusively Xrandr [23:17] jaytaoko, maybe once alpha-2 is out we should get ahold of or write some sort of multi-head test case thingee [23:17] bryceh: absolutely... [23:18] broder, yeah for Xinerama and TwinView you can't change resolutions dynamically; gotta edit xorg.conf. So that's probably not something unity needs to worry about quite so much. [23:18] jaytaoko, ok other issues/topics? [23:18] bryceh: you can change dynamically with TwinView, it just requires using an X protocol extension that nvidia invented (that's what nvidia-settings uses) [23:19] broder, ah, yeah then that would need tested [23:19] bryceh: yes, I am preparing a desktop system and I would like to know how to easily change the GPU without upsetting the system to much [23:20] bryceh: i would like to switch from an nvidia to a ATI gpu and vise versa [23:20] jaytaoko, first uninstall any proprietary drivers and remove your xorg.conf if any [23:20] jaytaoko, then you can swap cards back and forth pretty much at will [23:20] bryceh: ok [23:20] once the card is in, and system booted, you can reinstall the proprietary driver [23:21] bryceh: I will write a script to do that, [23:21] jaytaoko, if you get into trouble, do an apt-get purge on the proprietary bits to be absolutely sure [23:21] but I think that's a lot less necessary these days [23:22] bryceh: alright. do I have to remove the proprietary drivers when X is down? I guess yes [23:22] jaytaoko, bryceh: you're better off letting jockey handle the install, uninstall. you can do jockey-text -e xorg:nvidia_current to enable, jockey-text -d xorg:nvidia_current to disable [23:22] jaytaoko, no you can uninstall them with X booted [23:22] same with xorg:fglrx [23:22] jaytaoko, just use the Hardware Drivers applet [23:22] yep, just what broder said [23:23] use jockey for it, and it should take care of everything, including doing the xorg.conf changes [23:23] bryceh broder: thanks. I will be doing that a lot [23:23] by the way, since we're talking multihead, i'd appreciate feedback on the proposal i sent to http://mail.gnome.org/archives/gnomecc-list/2011-January/msg00007.html for automatically configuring hotplugged outputs, especially if you were at the multihead session at UDS [23:23] jaytaoko, fwiw this is where having several computers can save some time... ;-) [23:24] jaytaoko: by the way, it looks like gdk_screen_get_primary_monitor() seems to work with the nvidia driver, so i think that means it'll basically work everywhere === bjf is now known as bjf[afk] [23:24] bryceh: indeed! Actually I do have a Pentium 4 AGP system and I use it to debug the ATI issues [23:25] bryceh: but it is old and slow to compile... [23:25] jaytaoko: oh, no. take that back. it just picks the "first" monitor rather arbitrarily. lame [23:25] broder, I'm +1 to improvements to g-s-d. I'm advocating the idea of getting the expectations documented, since different people sometimes have differing ideas on how things should work, so the functionality sometimes tends to churn [23:26] bryceh: agreed, though the current g-s-d code kind of reads to me like it was an initial implementation just to get the functionality there that nobody ever went back and fleshed out [23:26] jaytaoko, what I've done is slap different cards into cheap hw, then use nfs mounts to my fast main box where I build the debs (using pbuilder) [23:26] i'll try to make sure that anything i write up is well documented. ideally with a test suite [23:26] broder, excellent! [23:27] broder, also touch base with RAOF as he had some ideas/plans in improving the configuration tool too, so maybe there's some potential for collaboration [23:27] bryceh: ok, will do [23:27] jaytaoko, alright anything else? [23:28] bryceh: no I think that is it for me... [23:28] ok great, was nice chatting with you [23:29] bryceh: What I gather from our conversation is that we have to follow on with ATI and fglrx [23:29] bryceh: and start prototyping with multi monitors [23:29] jaytaoko, yes [23:30] bryceh: alright! thanks to you and broder!