[03:36] Good morning [03:37] desrt: could be that suspend was blocked once, and then our systemd patch tried it anyway? I haven't uploaded the dropping of the patch yet [03:41] Morning pitti [03:41] bratsche: hey Cody, long time no see! How are you? [03:41] Pretty good, pretty good.. how are you doing? [03:42] overcoming ubuflu from last week's sprint, quite fine otherwise [03:42] still happy in my upstream qa role [03:42] Very cool [04:54] g'morning [08:04] morning! [08:04] hey Laney! [08:04] hey didrocks, how's it going? [08:04] Laney: I'm good, thanks! just beeing awake from 1am to 5am, but well… :) [08:04] yourself? [08:04] still jetlagged? [08:04] seems so :) [08:05] I had something similar on Sunday night but slept for over 12 hours on Saturday, so ... [08:05] went climbing yesterday outside in the amazing sunshine which seems to have tired me out enough to fix it [08:05] :> [08:06] Laney: I came back on Sunday (well, leaving on Saturday), so hopefully, it will get fixed tomorrow :) [08:06] I'm sad to have missed you all on client sprint, but I did some nice horseback riding on sunday, trip was supposed to be 2 hours but it became 3, weather was great [08:07] nice! [08:07] most important thing I learned was that a 'calm canter' isn't :) [08:11] or worse, it probably is... [08:13] good for developing the thigh muscles eh [08:14] definitely, I did some biking yesterday too, there's some definite improvement compared to a few months ago [08:15] I didn't even get slowed down that much by a strong headwind [08:21] rockin' [08:21] * Laney dist-upgrades the desktop to saucy [08:21] rarin' [08:22] * mlankhorst is installing raring desktop on panda, just to see if there's hw acceleration available or not, and then to check if that will still work on 1.14 with -ignoreabi [08:23] thought we considered panda desktop to be dead [08:23] unfortunately it is blocking upgrade to 1.14 xserver :/ [08:23] "Please restart the computer to begin using your updated software [Restart] [No other option] [No close button]" [08:23] * Laney grr [08:24] right click -> quit works [08:25] oh, but then refuses to run update-manager -d [08:25] yep, quite fun! [08:26] * mlankhorst was having some fun with plymouth + mir today, turns out that there was no input support yet in mir, so no way to hook it up to plymouth === vrruiz_ is now known as rvr [08:53] ogra_: so, about what was discussed yesterday about panda et al. there has never been a blob for it on the main archive, so installing it has always resulted in a sub-optimal experience, especially on quantal & raring which don't have unity2d anymore [08:53] ogra_: so I don't see how it could block any xorg update [08:54] tjaalton, the blob is preinstalled since precise [08:54] and is in the archive as well in resctricted [08:54] ogra_: on what? [08:54] huh? [08:54] on the panda desktop images [08:54] I've always needed the ppa [08:54] anyway, raring is broken already, right mlankhorst? [08:55] http://paste.ubuntu.com/5640860/ [08:55] tjaalton: not 100% sure, it installs with swrast, so if we only care about sw acceleration there's no blocker in moving to 1.14 [08:55] raring is fine and was released [08:55] i have not a single bug about any graphics related regressions [08:56] and the drivers for pandaboard were EGL or GLES already, no OpenGL support [08:56] yes [08:56] still sounds incredible that it's blocking the x86 world from moving on [08:56] and compiz works fine on it [08:56] not different to nexus7 [08:56] tjaalton, i was told we will not get any updates on the x86 binary side either [08:57] erm doesn't nvidia drivers already work for 1.14? [08:57] at least nobody is assigned to work on it [08:57] ogra_: we have drivers for 1.14 already since february [08:57] or march for fgrlx [08:58] where can I find the omap image? [08:58] but anyway pandaboard never had libgl drivers, so the only way compiz works is with swrast or a modified version to use the EGL libraries or something [09:00] tjaalton, on cdimage [09:00] omap image for raring that is. I only see precise on cdimages.u.c/release [09:00] tjaalton: there is a raring usb image [09:00] that's the one you need :/ [09:01] oh found it [09:01] http://cdimage.ubuntu.com/releases/raring/release/ [09:01] its an SD card image .... needs an USB stick/disk as target media [09:02] note that i would have preferretd to keep the nexus7 image (we even have updated tegra drivers for 1.14) ... [09:02] hey desktopers [09:03] salut seb128! [09:03] seb128: team meeting report reminder (think about modifying it btw ;)) [09:04] ogra_: yeah I was blind, downloading atm.. [09:04] mlankhorst, note that SWRAST wont be usable (unless it sped up a lot it is more like a slideshow than a desktop) [09:04] ogra_: and normal compiz IS usable? it uses opengl which panda doesn't support [09:04] it uses the gles implementation which linaro worked on for two years [09:05] didrocks, hey, yeah I was about to do that, I just added sil2100 and Mirv to the team first [09:05] * mlankhorst checks [09:05] seb128: excellent! [09:07] ogra_: hm seems you're right, still I'll try to find out if omap works on 1.14 if I specify -ignoreABI, it probably should unless it does some deep hooking into xserver [09:08] omap4 or omap ? [09:08] omap4 [09:08] i dont think the binary blobs work [09:09] it's just the xserver part that matters [09:09] they likelly just need a recompilation gainst the new abi [09:09] but there is nobody working with us who could do that [09:09] and it might need kernel changes which we will not do [09:09] (iirc the latter was the blocker) [09:10] it's not using the blob now? [09:10] it is [09:11] a quantal blob with a quantal kernel [09:11] on raring? [09:11] and there is no intention to update either afiak [09:11] yes [09:11] so much for backport stacks then :/ [09:11] kernel and blob are supposed to stay the same until we can drop the image [09:12] when is that? [09:12] well, feel free to backport/forward port the 1700 patches the TI kernel ships [09:12] no, I'd drop it asap :) [09:12] tjaalton, i wish i knew ... likely once we dont need any testing of xorg apps on arm anymore [09:13] who runs those now? [09:13] or needs [09:13] on saucy anyway [09:13] convergence [09:14] ogra_: so why not simply do the same as we are doing for precise, and offer a quantal xserver for pandaboard too [09:14] if we already do that for the kernel anyway [09:14] you should probably talk to slangasek about that [09:15] assuming -ignoreABI doesn't work [09:15] but I'm upgrading to 1.14 now to find out :-) [09:15] :) [09:24] seb128: thanks! [09:26] Mirv, hey, yw ;-) [09:29] bonjour seb128, ça va ? [09:30] pitti, salut, oui, et toi ? [09:31] seb128: mieux qu'hier, merci [09:31] no fever any more, and jetleg by and large gone [09:31] argh, I enabled the gnome3-team (and staging) ppa, now gnome-terminal's menu bar is back to the window instead of in unity's panel [09:32] probably because they have both an app menu and a window menubar [09:32] which makes no sense... :( [09:32] xclaesse: yeah, you shouldn't use this ppa if you want a working unity session [09:32] xclaesse: we didn't include all the latest components for some reasons… :) [09:33] so far it is the only slightly annoying bug I see :) [09:33] pitti, great [09:40] ogra_: seems to start and composite when I recompiled xf86-video-omap and added -ignoreABI to load the blob part [09:41] ah, sounds good [09:41] see if it crashes after a while though :) [09:41] ideally we could get a recompile though, nothing majorly was reworked but some internal structs did get shuffled around [09:44] hm first crash when I was changing backgrounds, I'll try unmodified since it wasn't that stable before I upgraded to begin with [09:58] seb128: hey, how's it going? do we have a BP for the settings work? and a current one for user session stuff? [09:59] Laney, hey, I'm good, thanks. How are you? [09:59] had a good trip back? [10:00] aching! [10:00] the trip was good, then I slept lots and then went climbing yesterday :-) [10:00] out at a quarry nearby [10:00] hehe, some exercice is good to kick the remaining bits of jetlag out :p [10:01] Laney, blueprints: [10:01] https://blueprints.launchpad.net/ubuntu/+spec/client-s-use-upstart-user-sessions [10:01] https://blueprints.launchpad.net/ubuntu/+spec/client-touch-system-settings [10:01] [10:01] they are pretty empty atm [10:01] sure, just wanted to make sure they were there [10:01] merci [10:01] I'm not sure yet how to really populate the settings one, we are sort of blocked on having the container app usable to be able to test things [10:02] but that doesn't stop us to play with qml meanwhile [10:02] well, currently we need to get up to speed on qml/sdk anyway [10:02] right [10:30] Hey seb128, Laney [10:32] hi darkxst [10:32] how's it going? [10:32] ah good, enjoying the cold weather! not so good for climbing though ;) [10:33] cold? :( [10:34] well Melbourne cold about 13 today [10:35] ah, yeah, suppose it's getting on into autumn there [10:35] should be up to 21 here today which is pretty good for the time of year [10:35] yeh I think we are back to mid 20's for the rest of the week [10:36] anyway, pitti wasn't too keen on the idea of adding another api to packagekit for l-s [10:36] did you speak to upstream? [10:36] pk upstream? [10:36] well, "get available locales" isn't really package specific [10:36] yeah [10:36] or gcc [10:36] it might fit into accountsservice slightly better [10:37] but I think it should just go into g-c-c's region panel [10:37] instead of depending on a new API which hasn't even been discussed/accepted upstream yet [10:37] well I was expecting it to be discussed upstream first [10:38] don't really care where it lives - just the initial idea of an l-s API seemed the wrong level [10:38] not upstreamable at all [10:39] Laney, I seriously doubt g-c-c would take that language stuff either way [10:39] yeah? [10:40] they wanted to add a button "install languages" [10:40] they have had designs for that for quite a while [10:41] I don't know if those designs were actually approved? [10:43] I guess the packagekit stuff (what-provides etc) would be ok, but I don't think they will take any of the ubuntu specific langugae changes [10:45] but that isn't an ubuntu specific problem [10:45] pitti, who else uses language packs? [10:45] (neither figuring out the set of allowed locales, nor installing additional packages for a language) [10:45] darkxst: all distros have extra packages that are language specific; this isn't limited to langpacks [10:46] but anyway, "which packages do I need for language XX" is already provided by aptdaemon and PK [10:46] (what-provides LANGUAGE_SUPPORT) [10:46] yes I am using that bit [10:46] what we discussed yesterday was "which locales are available" [10:46] which is essentially parsing and de-duping /usr/share/i18n/SUPPORTED [10:47] or "locale -a" if you only want already installed locales [10:51] pitti, ok will do it that way [10:52] neat [10:52] l-s already does that in Python, so the logic exists already [10:56] I didnt really want to duplicate l-s logic, but if thats the best option then I will [10:57] the l-s logic will go away at some point [10:57] (i. e. when the g-c-c region panel is functional enough) [10:57] ok === greyback is now known as greyback|lunch === greyback|lunch is now known as greyback [13:29] seb128, didrocks: dunno who packages empathy in ubuntu, but you guys probably wants to pick https://git.gnome.org/browse/empathy/commit/?id=add79d54e3c0149a2c336077d1e37c60d54c5ace [13:30] xclaesse, thanks [13:30] xclaesse, is that bug #1177285 ? [13:30] Launchpad bug 1177285 in gnome-online-accounts (Ubuntu) "gnome-control-center crashes on fill_contact_info_grid: assertion failed: (spec != NULL) after few seconds after opening online accounts " [Undecided,Confirmed] https://launchpad.net/bugs/1177285 [13:30] the commit is in master & 3.8 because we don't support older versions [13:31] but all versions are affected [13:31] 3.6 crash as well [13:31] seb128, exactly that crash [13:31] xclaesse, good, it was on my list of things to look at since we received some reports about it [13:32] xclaesse, thanks for the ping ;-) [13:32] seb128, the reason is that gabble hardcode vcard fields supported by Google (fn, n and photo) but recently google decided to add url [13:32] so another fix will arrive in gabble as well, but empathy shouldn't crash even if gabble does something wrong :) [13:32] right [13:33] xclaesse, does it apply to 3.4 as well? (that's the current lts version) [13:33] seb128, probably, I tested only with 3.8 and Guillaume tested with 3.6 [13:33] ok, I will have a look [13:33] thanks [13:34] seb128: forcing pandaboard with -ignoreABI on 1.14 seems to be unstable, but it works if I run with the franken xserver we were using for testing tegra against 1.14, so we could do a custom xserver-xorg-core package to keep pandaboard working [13:34] hm [13:34] 3.4 does not have libempathy-gtk/empathy-user-info.c [13:34] code was moved, let me check [13:35] mlankhorst, nice, well I guess it's up to you (xorg team) to figure out what solution you prefer, hold on the xorg update or find a workaround like that [13:35] well, talk to security too [13:36] they will have to care for it for 9 months [13:36] the xserver package will be identical, it will just have an extra patch to revert to the old ABI. [13:37] it would be similar to what we were doing with -lts-quantal rename [13:38] seb128, not tested, but that should fix the bug in 3.4: http://cgit.collabora.com/git/user/xclaesse/empathy.git/commit/?h=backport&id=d015077538b66554ed75194a47446b84d8081769 [13:38] xclaesse, thanks [13:38] seb128: but if upstream essentially doesn't maintain the omap drivers any more, does it really matter whether xserver receives any security updates? [13:40] mlankhorst, well, if x86 recieves a security update, arm shoould at least not ftbfs [13:40] it won't [13:41] mlankhorst, I'm not too concerned about the security issues for that specific image [13:41] it will be an extra package, like xorg-server-lts-quantal was [13:41] given we use a binary copy of a quantal kernel anyway, i doubt secutiry matters, it only matters if your change adds work for the security team when applying fixes in general [13:42] extra paackage ? [13:43] doing the same thing for tegra would cause it to break, so I think it's better just to add a different xserver that is omap specific only [13:45] for tegra there are updated drivers [13:45] yeah that's what I mean, and there is already a script to do just that renaming in an automated fashion :) [13:45] (we just have to update the package) [13:46] so if the omap install image just adds xorg-omap, and xserver-xorg-core-omap, things would work [13:47] if you can handle that on a package level that would be fine [13:47] special casing seeds for a single subarch isnt really an option [13:47] yeah I never intended that [13:48] Laney, pitti: since the logind updates, my screensaver doesn't unlock from lightdm [13:48] is that a known issue? [13:48] like if I start a guest session, logout from it, I'm back to unity-greeter, I enter my user password, that sends me to my session but screen is still locked [13:49] before that was sending me back to an unlocked session [13:49] seb128: two-factor auth, sort of ;-) [13:49] making sure you remember your password :-P [13:50] double single factor != double factor :p [13:50] it's an anti-aging device. [13:50] heh, I know. I've got quadruple auth in most places, as it is :-) [13:51] ogra_: is it also possible to do an image with a PPA pinned to a higher priority so it will always grab the xserver from there instead of main archive? [13:51] mlankhorst, PPAs are no option for official images [13:51] fingerprint, password, yubikey server and google auth :-) [13:51] aw too bad [13:51] (might change but until now that was a TB policy) [13:52] seb128: yeah, needs an upload of lightdm; it works for me (running lightdm from the PPA) [13:52] pitti, ok, as long as it's known, thanks [13:52] yes, it is [13:52] ogra_: in that case rename is the only option then :/ [13:52] seb128: actually, it's in the u-desktop PPA: https://launchpad.net/~ubuntu-desktop/+archive/ppa/+build/4535913 === m_conley_away is now known as m_conley [13:55] pitti, nice, will try that, danke! [14:17] pitti: actually it's in proposed [14:17] I didn't notice it hadn't migrated though - let me check why that is [14:17] seb128: ^ [14:18] Laney, thanks [14:19] hrm [14:23] Laney, looking at the changelog I guess lightdm-kde-greeter needs to be rebuilt with the new soname? [14:23] yeah, on it [14:23] cool [15:06] bah BAH [15:06] lightdm has a versioned dev package [15:21] ...and the new pcfile is bugged [15:30] Sweetshark, qengho, Laney, mlankhorst, tkamppeter, attente, desrt: hey, it's meeting time [15:30] g'day mate [15:30] https://wiki.ubuntu.com/DesktopTeam/Meeting/2013-05-07 [15:30] hey [15:30] I hope all those who were in Oakland had an ok trip back [15:31] slight change of format this week, we start by the "desktop platform/apps" subteam and didrocks will handle the second part of the meeting of the ps integration part [15:31] let's get started [15:31] Sweetshark, hey, there? [15:31] so I'm excused? :p [15:31] seb128: we're seriously having a meeting today? [15:31] mlankhorst, no, xorg is part of the desktop platform :p [15:31] so we can all update on what we did during our swap day? :) [15:32] mlankhorst: nice try btw ;) [15:32] well was worth a shot! [15:32] desrt, well, you can do a public update of what you got done during the sprint ;-) [15:32] fair enough. [15:32] community was not there [15:32] I wasn't either! [15:33] ok, no Sweetshark, he's working from California this week so I'm not sure he's online yet [15:33] qengho, hey [15:33] speaking of those working from cali [15:33] larsu: ping [15:33] desrt: yep?! [15:33] qengho, you missed some fun ;-) [15:33] seb128: so, I don't know what I'm supposed to say this time. New format, eh? [15:33] larsu: go wake up Sweetshark :) [15:34] oh -- and desktop team meeting [15:34] same content, just a different split I think [15:34] qengho, the status update format didn't change, we just do split subteam for the order [15:34] didrocks' also wanted to change a bit the format for ps integration to go by topic rather than people [15:34] I worked on chromium-browser. Trying to get more system libraries into the build so that it links on 32-bit hardware. Testing the various combinations. [15:34] but that apply less to us since most of our topic have only 1 person working on them [15:35] e.g chromium, libreoffice, ... [15:35] * Laney goes [15:35] qengho, thanks [15:36] • logind transitions, a lot of them. The archive is almost complete (everything ported or with a consolekit dep added) [15:36] Oh, I packaged libv8 too. It should be useful to others ere. [15:36] here. [15:36] • gstreamer 1.0.7 uploaded [15:36] • Clearing my outstanding merges (3 left) [15:36] • Just right now discovered that lightdm's qt pcfile is broken, so switched to fixing that. It's needed to fix lightdm-kde, which blocks lightdm from transitioning currently. [15:36] • Next up: think about blueprints for UDS, if any, and start looking at QML. [15:36] \0 [15:37] Laney, thanks [15:37] mlankhorst, hey [15:37] I've been putting some work in plymouth-mir, works but no input support yet, further work is postponed until mir api becomes more complete. Investigation into upgrading to 1.14, more testing for touch bug 56578. Preparing upload of lts-raring to precise-proposed. [15:37] Launchpad bug 56578 in mlview (Ubuntu) "Please sync mlview (universe) from unstable" [Undecided,Fix released] https://launchpad.net/bugs/56578 [15:37] https://bugs.freedesktop.org/show_bug.cgi?id=56578 [15:37] is panda back from the dead now? [15:38] I'm going to do a lts-style rename for panda xserver + xf86-vido-omap [15:38] Freedesktop bug 56578 in Server/Input/Core "race condition with active/passive grabs when opening menus with touch" [Normal,Assigned] [15:38] it seems to be stable with the abi reverts I was using for testing tegra touch [15:40] great [15:40] mlankhorst, thanks [15:40] np, off to the store, bbs [15:40] tkamppeter, hey [15:41] no tkamppeter? [15:41] attente, hey [15:42] seb128, hi [15:42] more gnome-control-center region panel patching, getting the keyboard indicator to work in the panel, debugging autopilot test cases in unity-gtk-module [15:42] this week, wrap up two items related to indicator-keyboard, then start to port it to Ubuntu touch [15:42] attente: did you get the module landed finally? [15:42] desrt, no not yet [15:43] one of these days... :) [15:43] attente, did you get the tests to run? [15:43] ha.. yeah.. [15:43] seb128, yeah [15:43] it turned out they were failing because of the keyboard indicator :S [15:43] what is blocking landing then? only waiting on the #ps integration team? [15:44] we still have patched versions of gtk [15:44] in archive [15:44] right, well that's a chicken egg problem :p [15:44] this is true.. [15:44] who should i talk to about this? [15:44] the unity-gtk-module doesn't need to an unpatched gtk to land, does it? [15:45] well, if you think things are ready from your side, let's just ask didrocks/cyphermox in the second part of the meeting [15:45] or sil2100 [15:45] ok [15:45] I think it's up to them to land the source [15:45] your indicator-appmenu mr got approved [15:45] sil2100 would know [15:45] then we just need to drop the gtk patches [15:45] he wrote some notes that he had to revert it [15:45] attente, thanks [15:45] so better to check once he's back :) [15:45] didrocks, right, he said it was due to autopilot 1.3 breaking compat or something [15:46] yep [15:46] let's discuss it later [15:46] desrt, your turn ;-) [15:46] so i spent a lot of last week recovering from a sunburn [15:46] but not as bad as seb128 [15:46] lol, I'm good again no worry ;-) [15:47] while that was going on, i mostly finished the GtkMenuTracker which will be used as the basis of all consumer implementations of GMenuModel (gtk, gtk-on-mac, unity-gtk, unity-qml and gnome-shell) [15:47] also, at kenvandine's request i started working on a qmlscene-type program based on GApplication.... that was a bit frustrating, but i got some pretty good ideas from it [15:48] which led me towards a conversation with thomas and thomas about how actions will work in unity apps -- and bumped GActionDescription way up my priority list: going to start working on that soon [15:48] also had more conversations about our C APIs..... [15:48] finally, landed the suspend support in systemd-shim, removing the logind depend on pm-utils and breaking everyone's laptop's suspend in the process [15:48] gonna try to fix that up today :) [15:48] (fin) [15:49] desrt, did you manage to reproduce the second suspend issue? [15:49] it does it in a reliable way here [15:50] second suspend -> suspend not working the second time [15:50] no [15:50] it's always working for me [15:50] hum, k [15:50] from the session indicator, right? [15:50] let me know if you need details [15:50] yes [15:50] i'm pretty sure it's a race [15:50] on a 32bit install (if that makes a difference) [15:50] like, a signal getting emitted before it is being watched for on the client side [15:50] and then it gets missed [15:51] i'm reading code and pinging lennart -- the dbus protocol is slightly underspecified in this area [15:51] seb128: do you have indicator-session from trunk? (the logind-ified one)? [15:51] one of those 'specification by sole implementation' (systemd) things [15:51] pitti, no, I've stock saucy [15:51] actually, it hit saucy 18 minutes ago, thanks cyphermox [15:51] oh, nice, will try that [15:51] pitti: should this fix it, then? [15:52] well the interface it's using works the first time [15:52] I don't know -- suspend is always working here [15:52] so it's not totally incompatible with logind [15:52] tre interesa [15:52] it talks to upower, so I guess that "already pending" error might actually come from upower [15:52] (we had a case like that last week) [15:52] I just tried double (triple) suspending, worked every time [15:53] so yeah, try the new i-session and see what happens [15:53] you guys all run indicator-session from trunk? [15:53] ok [15:53] seb128: update to the new indicator and let me know. could be there is actually no issue at all here. [15:53] will do after the meeting [15:53] desrt, thanks [15:53] ok, did I forget anyone? [15:53] otherwise my summary: [15:53] * quite some discussions at the sprint, mostly focussed on trying to figure out the details for indicators and system settings [15:54] * looked at cleaning some of the deprecated libs from our current images, also reviewed what is still using python2 [15:54] * started merges on debian [15:54] * some blueprint work [15:54] * updated gtk 3.8.1 in the ppa, trying to get that ready for saucy [15:55] cyphermox fixed the overlay-scrollbar issue so we are getting close [15:55] oh wow, he did? [15:55] good job [15:55] yes [15:55] wow, thanks cyphermox! [15:55] https://code.launchpad.net/~mathieu-tl/overlay-scrollbar/fix-second-use/+merge/162507 [15:55] needs a review/approval still [15:55] yup [15:56] but I verified locally, it works for me (and gtk 3.6 still works fine) [15:56] [15:56] ok [15:56] otherwise, small not before ending the first part of the meeting [15:56] note [15:57] Sweetshark, qengho, Laney, mlankhorst, tkamppeter, attente, desrt, didrocks, kenvandine, cyphermox, Mirv, others: vUDS is next week, if you have any session you want to have (or that you think others should have), please register a blueprint and get it targetted for the uds sprint [15:58] ack [15:58] ack [15:58] would be good to also chase other teams so they register blueprints if needed [15:58] * desrt will probably be away this week [15:58] ackie [15:58] o/ [15:58] desrt, you were not going to register a session anyway I guess? ;-) [15:58] sil2100, hey [15:58] seems unlikely :) [15:59] * desrt could go hang out with cyphermox for mini-ubz [15:59] on the note of people being away, please try to deal with blueprints for vUDS today if you can [15:59] quite some people are off starting tonight for the end of week [15:59] national holidays wed&thursday and swap days from last week [15:59] seb128: I don't have anything really important to discuss at vUDS anyway [16:00] so rather waiting for people inviting :) [16:00] didrocks, ok [16:00] wed AND thurs! [16:00] (if needed) [16:00] you french! [16:00] on that note, we are just on time for second half of the meeting [16:00] ok, maybe let's start the second part of the meeting :) [16:00] seb128: ok, will deal with my blueprint-related things later today [16:00] if nobody has comments or questions [16:00] comments/questions? [16:00] sil2100, thanks [16:01] ok, seems not [16:01] cyphermox, Mirv, sil2100, kenvandine, robru: hey guys! hope that you trip back went well :) [16:01] your* [16:01] didrocks, your turn [16:01] let's go over the important topics from https://docs.google.com/a/canonical.com/spreadsheet/ccc?key=0AuDk72Lpx8U5dHFtUmlPOUtCRk8zR2dtaEpIbUVhMmc#gid=0 [16:01] I thought I would get crazy during my trip back though, never again! [16:01] do not forget to update it btw as it's our communication guide :) [16:01] Ok [16:02] yeah, never again, as always when crossing the Atlantic. but alive, so "well" :) [16:02] heh [16:02] as a reminder, I won't be there until next Tuesday [16:02] so I will surely put some of the tasks assigned to me on you guys :) [16:02] ok, let's start [16:02] No problem ;) [16:02] robru: "Splitting webapps and have that on daily release" [16:02] with kenvandine's help [16:02] I saw some good progress [16:03] seems we still have some issue to get fixed, right? [16:04] …no robru? ok, we pushed back the meeting for him. I hope that next week he'll be there on time at least… (the meeting is quite late for some people like Mirv, starting early) [16:04] kenvandine: around? [16:04] hey, just woke up, sorry [16:04] yup [16:04] robru: just in time! :-) [16:04] robru: hope you have some coffee ;) [16:04] robru: so, what's up on webapps? [16:05] almost ;-) [16:05] didrocks, I have branches pushed to fix most things. [16:05] 12h since I started my work day :) [16:05] didrocks, the only thing I can't fix is twitter, need alex abreu to write a manifest.json fro it [16:05] robru: ok, let's see how it goes ;) are you confident to get it in with autopilot tests for next week? [16:06] didrocks, there's an issue with jenkins not having the build-dep in thePPA so it complains of rules.mk again [16:06] yeah, I commented on the MP, ping vrruiz for it I guess [16:07] not too confident with autopilot, not sure the status of it. seems incomplete/bitrotten. but the linter script I wrote can detect basic issues ;-) [16:07] and it can be expanded to detect new issues as we find them [16:07] robru: ok, let's try at least without autopilot for next week :) [16:08] next topic is disabling indicators, this is done cyphermox, isn't it? [16:08] yup, done [16:08] do you mind updating the spreadsheet? [16:08] next one is finishing the doc when putting S in distro, cyphermox, can you please send me an email with your notes so that I can update it? [16:09] I was thinking we should finish all of it before [16:09] but sure, I'll send what I have now [16:09] cyphermox: ok, if this week, some are needed, like unity and so on, do you mind giving a hand to the person doing it? [16:09] as I won't be around :) [16:09] cyphermox: if so, writing the doc next week is fine to me [16:09] well I can send what I alreday have now, then we'll add whatever else [16:10] sure :) [16:10] thanks cyphermox [16:10] I finally found time to do the Vcs-Bzr ignoring + collecting commits for all mainline commits \o/ [16:10] (some part was on the plane, not at the bar :p) [16:10] awesome. [16:10] I'll archive it [16:10] (see second tab) [16:11] sil2100: hey! touch building in next ppa, what's the status? [16:11] hmm [16:12] Hard to say, since there wasn't much things I knew since I le [16:12] *I left the sprint [16:12] I wonder if I should have the Qt/SDK stuff as a separate item [16:12] Mirv: would make sense, please, feel free to edit [16:12] Mirv: ^ [16:12] Mirv: kenvandine: did you look at that since Friday? [16:12] platform seems fine, but still hitting utah with sdk [16:12] for my part [16:13] yeah [16:13] all the web cred bits are there for touch [16:13] according to the spread sheet, it's basically all done [16:13] I would say all what is left is the e-mail probably, but will have to double check the stacks to be sure [16:13] And then just the e-mail [16:13] sil2100: Mirv: kenvandine: so for next week, do you mind finishing the autopilot 1.3 transition (with robru I guess) in all stacks and having everything delivered in next? UTAH should feel a little better from what I know [16:13] I can do that later as well, if didrocks wouldn't be around [16:14] ACK [16:14] didrocks, it wasn't better this morning :/ [16:14] sil2100: yeah, just take the item and discuss between you guys :) [16:14] kenvandine: hum, let me check we don't dist-upgrade anymore [16:14] didrocks, hey, is it true that one failing webapp stops the whole webapps stack from being published? [16:15] kenvandine: we don't dist-upgrade anymore, mind checking with the QA guys? [16:15] robru: right, it's "per stack" [16:15] robru: the stack is validated or rejected [16:15] didrocks, who should i talk to? [16:15] that's why the stack should be own by the same team [16:15] om26er has many autopilot 1.3 branches at https://code.launchpad.net/~om26er [16:15] kenvandine: ping on #qa gema or anyone in the QA team [16:15] didrocks, any way to change that? there are 40 webapps and any one of them is likely to be broken at some point... we really need to be able to release them independently. that was the whole point of splitting them, so we can release only the ones that need fixes... [16:16] robru: well, then, the team need to fix them :) it will only rebuild those having changes, if they are broken, they need to be fixed [16:16] robru: we are validating the whole stack in integration tests [16:16] didrocks, ok, but it doesn't really make sense to hold back the whole stack just because one minor/obscure one is broken. [16:16] Mirv: ah, nice! so it's just a question of tracking if they are merged [16:17] robru: well, that's working quite fine for most of the stacks, as they are owned by the same team, we can divides stack, but as it worked for the past 6 months, I'm not keen on that [16:17] didrocks, it's not like these are critical libs where one broken one breaks the whole experience; users are only likely to care about a few of the 40 apps, better to get them "all the working ones" and only hold back the broken ones [16:17] robru: maybe you want to discuss that during vUDS? would be a good topic, but all the original design is based on that. I'm surprised you didn't get it before :) [16:18] robru: if we don't support or care on what we deliver, we should remove the app [16:18] robru, the issue is if one app fails the tests, it could be because of a regression in libunity-webapps that the other webapps don't use [16:18] didrocks, well the situation before was that all webapps in one branch meant you had to re-release all of them every time you wanted to fix just one of them. I thought the *whole point* of what i was working on all last week was to split them up into *independent* branches so they could be released *independently*. [16:18] like maybe webapp-foo is the only webapp that uses feature A from libunity-webapps [16:19] robru: yeah, if you have one fix somewhere, it will release only this fix [16:19] we shouldn't let libunity-webapps get publshed [16:19] robru: instead of pushing 40 binary packages to distro, only one being changed [16:19] for instance [16:19] robru: it's just the case of integration tests failing - if integration tests are failing, it means the failure will be visible to the user [16:19] So that's why it has to be done so I think [16:19] in practice, this is working way better than what you may think, and forcing us to deliver all working components [16:20] sil2100, no integration tests on webapps yet ;-) [16:20] robru: boo... ;( [16:20] robru: well, they are, but not activated yet :) [16:20] robru: we can discuss that at vUDS if you want to open a session on it [16:21] let's go to next topic [16:21] didrocks, ok, but it doesn't make sense to me that AngryBirds app (which is basically a stub, it has no features) can be broken, and this holds back important ones like GMail or amazon [16:21] robru: well, we can fix AngryBirds then [16:21] that would make more sense to me, as we know exactly the day it broke [16:22] didrocks, well it broke from the beginning :-P [16:22] robru: it should have never been broken :p [16:22] or never published if broken [16:23] didrocks, alright, then I might push a branch disabling some apps then. [16:23] robru: sounds good to me [16:24] I guess the rest of the list is under control and linked to the previous discussion (basically having autopilot 1.3 transition done, touch published to next ppa, get everything green, check with qa on UTAH) [16:24] isn't it? [16:24] sil2100: kenvandine: Mirv: you are mainly the ones holding those actions ^ [16:24] the utah errors are cobbler failures [16:24] yeah [16:24] about so [16:24] Yes [16:24] kenvandine: ok, so UTAH is using cobbler wrongly again, need to get them fixing it… :/ [16:25] so, now, some new incoming work for fun :) [16:25] unity raring SRU [16:25] i just asked them [16:25] Mirv: do you have time for that one? Basically delivering a SRU for raring with current unity raring [16:25] Mirv: weren't you working on that? ^ [16:25] Mirv: it's about checking that the bugs are fine [16:25] Ah, raring! [16:25] Duuuh [16:25] * sil2100 keeps confusing raring and precise [16:25] and make a manual publishing if everything all right :) [16:26] yeah, I've also the precise SRU waiting for a patch pilot :) [16:26] as it's builed daily in the daily-build ppa [16:26] didrocks: I don't yet have a touch on the practicalities of the delivery [16:26] otherwise I'm fine with checking [16:26] Mirv: yeah, no need for a patch pilot for this one :) just having kenvandine or cyphermox doing the manual publication for you [16:26] Mirv: so look at the ~ubuntu-unity/daily-build ppa for the indicators and unity stacks [16:26] Mirv, let me know when it's ready [16:26] look at the generated changelogs [16:26] ok, will do [16:26] bugs associated [16:27] and hop! upload if things are following SRU rules :) [16:27] i'm on pilot duty tomorrow :) [16:27] (upload being manual publishing) [16:27] so should be easy ;) [16:27] but i'll be out tomorrow morning, so just email me [16:27] sounds like a good list [16:27] Mirv: with luck utah would run and you will get test results even! :) [16:27] Mirv: mind adding it yourself to the spreadsheet? [16:28] didrocks, can you tell a little bit about how/when we plan to transition from PPA to start releasing directly into saucy? [16:28] We just remove the dest PPA from the configs [16:28] IIRC [16:28] robru: I guess we have enough for this week with the touch to have it built in next [16:28] robru: so I thought about next week, wdyt? [16:28] didrocks: ok, done [16:28] wfm [16:28] thanks Mirv :) [16:29] didrocks, well, I don't know what the status of the other stacks is like, but Friends as released for raring has a really horrible crasher bug in saucy due to some library API breakage, so I have a fix that I want released into saucy quite urgently ;-) [16:29] kenvandine: cyphermox: sil2100: Mirv: robru: a small note, it seems the cu2d-* command doesn't work since they restarted jenkins, cyphermox will check with qa about it, meanwhile, use the web UI. [16:30] ok [16:30] didrocks: I don't have the access rights still there [16:30] ACK :) [16:30] yeah, just wiating to hear back about it [16:30] robru: ok, for that one, you can get in touch with cyphermox, he knows how to transition (quite simple, but better to pair) :) [16:30] robru: and let's transition friends to be saucy! :) [16:30] Mirv: yeah, the request is pending (but people on holidays) [16:30] hmm,, sauce. [16:30] didrocks, cyphermox so it's possible to just do friends direct to saucy without affecting other stacks? [16:30] didrocks: ok, thanks [16:31] robru: it is :) [16:31] robru: assuming it doesn't need other pieces of the stack, yes it should [16:31] as cyphermox did with indicators [16:31] ok, last but not least… (sorry, have a lot this week) [16:31] sil2100: it's for you! [16:31] ?! [16:31] sil2100: so, mhr3 wants to get 100scopes merged back to saucy unity trunk [16:32] !! [16:32] sil2100: so just sync with him, you will have to disable the experimental stack once done and add to head/unity.cfg the scopes [16:32] sil2100, piece of cake, right? :) [16:32] Is it *ready*? Since I didn't look at it for a while, and Paweł told me during the sprint about some regression they were trying to fix in unity [16:32] didrocks: ok, will do ;) [16:32] just target "next" for this week, we'll go over the sauce next week :) [16:32] cyphermox, I guess this is what I need: https://code.launchpad.net/~robru/cupstream2distro-config/friends-saucy/+merge/162712 [16:32] sil2100: yeah, it is from what I know of :) [16:33] robru: you need to change ppa to ubuntu-unity/daily-build and remove the "dest" line [16:33] robru: yeah, just wiating for the updated diff [16:33] didrocks: ok, I'll add it to the spreadsheet [16:33] didrocks: ^ cf. diff [16:33] cyphermox: ah nice! :) [16:33] sil2100: thanks a bunch (and good luck) :) [16:34] actually [16:34] robru: ppa: ubuntu-unity/daily-build-next should remove -next [16:34] cyphermox: good catch! [16:34] ok, I don't think we need everyone to stay to transition friends to saucy, any question before wrapping up? [16:35] not from me [16:35] robru: let's get this finished, I'm starving :) [16:35] cyphermox, just pushed it [16:35] thanks Mirv, sil2100, robru, cyphermox, kenvandine :) see you next week (and don't break everything meanwhile :p) [16:35] ok [16:35] didrocks, thanks [16:35] didrocks, enjoy! [16:35] didrocks: ok! Have a nice holiday! [16:36] thanks ;) good luck guys! [16:36] thank you :) [16:36] didrocks: and see you next week as well, we'll be informing you about the progress probably during that time anyway ;) [16:36] sil2100: think about updating the spreadsheet, best way to communicate! [16:36] heh [16:37] didrocks, sil2100: what's the status of the gtk unity menu landing? [16:37] yeah, now that the meeting is over, let's go on that one :) [16:37] sil2100: you wrote "We're reverting the addition now since it would basically break unity" [16:37] sil2100: what did you mean? [16:38] cyphermox, wow, that diff just doesn't want to update on launchpad. I promise I fixed it though [16:41] yup, looks good [16:41] robru: do you have all the same commits in raring as in trunk for friends? [16:42] yeah, looking good [16:43] I think as soon as didrocks wants to stamp it, it's good :) [16:43] didrocks: you're in the review list, wanna just approve it or whatever? [16:43] and then can somebody redeploy it and everything? I need to get this fix out asap ;-) [16:44] cyphermox: I trust you, do you think I should look at it? (the discussion is enough for me to say "ok") [16:44] no [16:44] cyphermox: so go ahead ;) [16:44] it's just that you're in reviewers somehow ;) [16:44] robru: tired of getting crash reports? :p [16:44] robru: yeah. just waiting for it to get merged [16:44] didrocks, we had three separate bugs reported just for this one thing. I wish people would stop reporting duplicate bugs. [16:45] robru: well.. [16:45] didrocks, best was that the guy who reported the third such bug went back and marked the two *older* ones as duplicates of *his*. [16:45] we kind of encourage them to report duplicate bugs and have us dedup, rather than mixing things up [16:45] heh [16:45] robru: ahah, because "mine" is more important :) [16:46] launchpad should know to prefer the oldest bugs by default when duplicates are being marked ;-) [16:46] Mirv: also, I'm patch pilot today, so I can help you ;) [16:46] kenvandine, https://code.launchpad.net/~robru/friends/fix-threads-init/+merge/162692 can you approve this? trivial fix for a critical crasher in saucy [16:47] robru, done [16:47] kenvandine, thanks [16:48] alright, so now it's time for breakfast. back in a few ;-) [16:48] FYI, utah just passed on the 100scopes thingy [16:48] so there is hope for the touch stack! :) [16:48] robru: enjoy ;) [16:56] sil2100, hey === m_conley is now known as m_conley_away [17:03] robru: no rest API makes it a little difficult to even update the jobs... [17:04] as soon as retoaded fixes that I'll update the jobs and we'll be able to release [17:04] cyphermox: I've quickly rewritten https://wiki.ubuntu.com/DailyRelease/MovingNewRelease#Step_two:_once_release.2B-1_is_opened, feel free to edit it [17:06] cyphermox: I've archived the disabling indicators as well as it's done [17:08] ok [17:12] * didrocks waves good evening, see you next week! [17:13] robru: ok, stuff is updated! [17:14] robru: so everything is ready for friends to be landing? [17:21] attente: hi! Today I'm off, but let's talk tomorrow if it's not a problem [17:21] attente: in the meantime - https://code.launchpad.net/~sil2100/unity-gtk-module/minor_ap_fixes/+merge/162851 [17:21] attente: in what timezone are you? [17:22] sil2100, ah, sorry, wasn't aware [17:22] i'm UTC-0500 [17:23] er. i guess UTC-0400 [17:24] attente: ok, so let's talk tomorrow :) Since I need to go now sadly [17:24] See you then! [17:24] sil2100, this looks good, thanks! [17:28] robru: friends started. [17:35] cyphermox, yeah, everything is ready! thanks === mitya57_ is now known as mitya57 [18:07] tedg, larsu: https://bugs.freedesktop.org/show_bug.cgi?id=64115 <- tricky one, do you know of any relevant fixes as candidates to be SRUed for precise? [18:07] Freedesktop bug 64115 in UI ": UI: menus are not dynamically changing under unity" [Normal,Needinfo] [18:22] cyphermox, still around? [18:23] yeah [18:23] robru: watching friends build [18:24] cyphermox, https://code.launchpad.net/~robru/cupstream2distro-config/disable-broken-apps/+merge/162858 these two apps fail the stack and shouldn't, so I'm disabling them until one of the webapps developers can fix them. [18:25] ack [18:26] now to fix twitter... ;-) [18:26] ;) [19:03] Sweetshark, Interesting, I guess we need to know if it's only on Precise. [19:04] Sweetshark, We didn't change that much in indicator-appmenu for raring except to pull out HUD to it's own package. [19:05] Sweetshark, Added myself to CC [19:21] tedg: thx [22:45] jasoncwarner, hi, jam has a problem with his unity after upgrade; who can help him out since both seb128 and didrocks are asleep? [22:45] mterry: hey. is there any way to get deja-dup to run a hook script pre-backup? [22:48] desrt, I bet the new upstart user sessions could be configured to run something right when deja-dup starts [22:49] * desrt wants it to run just before the backup starts, and for the backup to wait until it's done [22:50] desrt, in that case, why not port deja-dup to just use upstart for its backup scheduling? then it would be really easy ;-) [22:50] that's actually a pretty good idea [22:51] I bet mike would appreciate it ;-) [23:03] * desrt has devised a way to shrink his ~/code/ directory from ~6GB down to ~15M [23:04] desrt: Let me guess. Delete it. :p [23:04] basically, it's a mix of the following: use jhbuild dvcs_mirror with a directory in ~/.cache/ [23:04] and make use of alternates when cloning [23:05] ah ok. [23:05] then, write a command (i call 'git tuck') that does this: [23:05] git stash -u [23:05] git prune; git repack -a -d -l [23:05] then nukes the working directory, keeping only the .git [23:05] and erases the index too [23:05] wow. then what, you just check out what you need? [23:06] basically this will remove any object that's in the upstream repository from the local git object db [23:06] (ie: standard git-clone --reference type stuff) [23:06] but the next biggest thing in a git repo is all of the unpacked files (99% of which are unmodified) [23:06] oh, I see, so you only keep the stuff you're working on, discarding whatever stuff has been accepted upstream [23:07] so this does a stash -u to store any changes you might have and then throws everything away [23:07] any local commits/branches you have are also saved [23:07] that's the beauty of git repack -l [23:07] it keeps only objects not found in the upstream repo [23:08] desrt, what if you want to check out an old commit for testing purposes (say, bisecting a bug?) do you have to redownload all the objects? [23:08] well [23:08] if you still have the stuff in ~/.cache then you're OK [23:09] otherwise, yes, you have to download it again [23:09] the purpose of this is to get everything into the minimum possible size for backups [23:09] desrt, I haven't paid much attention to disk usage in a long time, but my ~/src/ is only 1.1GB, so that's not bad. your trick won't work because mine's almost exclusively bzr. [23:09] :( [23:10] desrt, good point. I personally don't see much value in backing up stuff that's already online, though. even my "local changes" are pushed to a branch somewhere. [23:10] i often get into a situation where i'll build up a fair amount of local changes (due to hacking on a plane or something) that don't get pushed [23:11] yeah, gotta remember to push once you land ;-) [23:11] and for projects that i hack seriously on i usually have reams of local branches