[01:21] c [04:30] Good morning [04:30] Good morning [05:56] powerpc builds are lagging way behind :/ [06:01] didrocks: Good morning! [06:01] hey RAOF! [06:05] RAOF: while you are around: regarding xmir, so you are going to maintain ABI manually? like, we don't need any black soname virtual package provide? [06:05] didrocks: I'll do what the rest of X does, which is have an ABI version in the code. [06:06] RAOF: but not for mir -> xmir, right? [06:06] the ABI version in the code will be in mir? [06:07] For the XMir dependency on Mir we'll just use the shlibdeps; xmir.so links with libmirclient.so.1, so that's all good. [06:07] RAOF: ah, excellent, if you don't expect any ABI breakage without soname bump :) [06:07] didrocks: I've got a symbols file for it cooking :) [06:07] heh ;) [06:07] And none of the XMir drivers link with libmirclient; XMir hides that from it. [06:08] excellent :) [06:08] RAOF: where is the xmir ppa right now? so that I can add it for running the tests until we are in distro [06:08] xmir/mesa [06:08] ppa:mir-team/staging ? [06:08] ok :) [06:08] RAOF: and if I run on ati [06:08] I should see the fallback happening, right? [06:08] Fallback? [06:09] (unity-system-compositor not starting, regular Xorg running) [06:09] If you run on fglrx you should see the fallback to regular X, yes. [06:09] yeah, great, we'll test both cases that way :) [07:24] good morning [07:31] salut jibel! [07:34] salut didrocks [07:35] sil2100: hey, good morning! [07:43] didrocks: good morning! Time to unblocketh some stackeths! [07:43] sil2100: I think you see that most are blocked on ppc [07:43] there are only 2 builders, one which has built some private stuff apparently [07:43] (backlog of 4 hours now apparently) [07:44] sil2100: we can be bad citizen and I'll suggest to run cu2d-skip on powerpc on the indicators and other that will start building [07:44] Oooh [07:44] sil2100: the only thing is that indicator-session will be then published even if failing to build on ppc I guess [07:45] I guess we don't care that much, right? [07:45] well, indicator-session will be blocked in proposed [07:45] but I think that's fine [07:46] Ah, ok, thought powerpc is mandatory in overall [07:46] Since most of the touch bits don't have powerpc at all [07:46] sil2100: yeah, but for those that are arch: any it won't be copied [07:47] sil2100: however, we can bypass dailies to have the rest building [07:47] hey desktopers [07:47] sil2100: so just run cu2d-skip on what's waiting on ppc [07:47] we'll have 2 builds (one on the ppa, one in proposed) instead of 1 [07:47] but "meh" [07:47] salut seb128! [07:47] didrocks, lut ;-) [07:51] Hi seb128 ! [07:51] didrocks: ok, then let me look at cu2d-skip ;) [07:53] sil2100, hey ;-) [07:54] didrocks: skipped (probably) ;) [07:55] sil2100: the next build line should tell if you are right ;) [07:55] sil2100: oif as well? [07:55] WARNING Request to ignore indicator-session on powerpc. \o/ [07:55] ah, oif is on raring -> don't care :) [07:55] OIF is green on head [07:56] sil2100: hum, you just ignore indicator-session? [07:56] SDK has some trouble with the source package name, but strange [07:56] not the other indicators? [07:56] sil2100: yeah, it's the new component, I'm dealing with it [07:56] * didrocks waits for https://code.launchpad.net/~didrocks/unity-action-api/source-package-rename/+merge/174937 to get merged [07:57] didrocks: just indicator-session, since other indicators are ok, right? [07:57] sil2100: they are waiting on building as well [07:57] sil2100: and seeing the ppc backlog, I think we should skip all powerpc builds today [07:57] Can I somehow tell cu2d-skip to skip all projects from a stack? [07:57] Or manually? [07:57] sil2100: you have to list them, but you can do that in one command [07:58] see --help :) [07:58] ah no [07:58] Help says I can only comma-seperate archs ;p [07:58] it's the archs [07:58] bad memory :) [07:58] so yeah, one command per project [07:58] didrocks, did you ping infinity or the is guys to see if there is a way to get the 3rd ppc builder back (sagari)? [07:59] seb128: I think you just did (so no, I didn't) :) [07:59] didrocks, yeah, not sure that infinity is around though [07:59] let me try #is [07:59] I didn't see him this morning [07:59] oh, not going to be useful [07:59] https://launchpad.net/builders/sagari says "NOTE: En route from Boston to London" [08:00] seb128: I heard there are pipes under the ocean :) [08:00] ;-) [08:00] sucks to be down to 1 actual builder (the other one is going openjdk, that's going to take a while) [08:01] right, and it was 0 for a while [08:01] as there was a private ppa build [08:02] bah, didrocks' system is too smart [08:02] why? :p [08:02] I should stop listing bugs in my commits when I do workarounds for them [08:02] ahah ;) [08:02] hey [08:03] I tend to write "do this because of lp #nnn in the toolkit" [08:03] sagari is in transit to london from what I understand [08:03] yeah, my regexp is quite relaxed on purpose [08:03] oh you foudn that [08:03] and that ends up with a ubuntu-system-settings component to be added to the bug [08:03] and closed by the upload [08:03] seb128: see, more karma, you can even reclose it a second time then :p [08:03] Laney, hey [08:04] Laney, yeah, we found that note ;-) [08:04] didrocks, hehe, more karma for me or for the upload bot? :p [08:04] should be better [08:04] if you ever noticed that it takes ages to upload its builds [08:04] didrocks: ACK please, testing code \o/ http://10.97.0.1:8080/view/cu2d/view/Head/view/Misc./job/cu2d-misc-head-3.0publish/lastSuccessfulBuild/artifact/packaging_changes_phablet-tools_0.15+13.10.20130716-0ubuntu1.diff [08:04] sil2100: how to say no to testing code? ;) +1 [08:05] didrocks: (all indicators skipped already) [08:05] Laney, no, I never noticed, I noticed that it takes ages atm to get builds through though :p [08:05] hate ppc [08:05] sil2100: \o/ we'll have to look at other stacks that are going to get unblocked now to skip ppc on them as well I think [08:06] you could ask for some rescoring [08:06] I've nothing urgent [08:07] it's just annoying that we don't get daily landing this morning because things are waiting to be built on ppc still [08:07] that and the some hours backlog for anything we will upload [08:08] seb128: hum, are you sure you did powerpc for indicators? last check didn't catch that [08:08] does anyone here have a multitouch capable laptop and saucy? [08:08] mlankhorst, I don't [08:08] didrocks, you want sil2100? [08:08] sil2100: ^ [08:08] seb128: sorry [08:08] ;-) [08:08] np [08:09] mlankhorst: yes, if a macbook pro counts [08:09] unity-gyk-module -> that doesn't work :p [08:09] Laney: erm touchscreen, not touchpad sadly :P [08:09] oh [08:09] that's what I know as multitouch ... [08:10] oh sorry, does anyone have a laptop with a touchscreen? [08:10] sil2100: you didn't run them from the modules it's waiting for [08:10] Laney: hm maybe I was just too used to those old resistive touch screens ;) [08:10] sil2100: just those that didn't build :/ [08:10] sil2100: where did you fetch the list? [08:10] I guess maybe it's a more general term [08:10] but they called these multitouch trackpads when they came out :P [08:12] * didrocks looks for sil2100 :p [08:12] Eeek [08:12] But wait [08:13] hmm [08:13] What did I do wrong that it's still waiting for libusermetrics, bamf and such [08:14] Ah shit one typo, but what about the others [08:14] unity-gyk-module [08:14] sil2100: you are reading http://10.97.0.1:8080/view/cu2d/view/Head/view/Indicators/job/cu2d-indicators-head-2.1build/268/console, right? [08:14] you see there the components that are still pending building === didrocks1 is now known as didrocks [08:14] didrocks1: yes, I made a typo, but some of them I did a skip already [08:14] interesting to receive my own message :) [08:14] sil2100: which ones? [08:15] I see indicator-client, indicator-sync [08:15] that you tried to skip [08:15] but they are not building at all this round [08:15] (not sure why you listed them) [08:15] But I also listed all the others, but they don't skip [08:16] I even explicitly set unity-gtk-module now, since I made a typo [08:16] sil2100: ok, let's see that one [08:16] didrocks: will something happen if I re-skip again? [08:17] sil2100: normally, no [08:17] skipping is just touching a file [08:18] ok, I see an indicator-sound.powerpc.ignore [08:18] with a bunch of others (you didn't need to skip the whole things, what doesn't build :p) [08:18] but maybe there is a bug in the code [08:18] * didrocks opens vim [08:18] Maybe my gouranga skip made the script to overflow? [08:19] sil2100: well, I doubt there are that many :p [08:20] for arch in self.archs.copy(): [08:20] if os.path.isfile("{}.{}.ignore".format(self.source_name, arch)): [08:20] doesn't seem complicated… [08:20] but yeah, bamf didn't ignore it [08:21] I think all should be ok, I skipped all of them because I wrote a script that does that for a stack list [08:21] Too bad I typoed ;/ [08:26] oh oh oh [08:26] I think I understand what's happening [08:26] hm? [08:26] as I pass to the constructor a list [08:26] in fact the object self.archs is shared between components [08:26] so, if I remove from one [08:26] it will be removed from all of them [08:27] didrocks, sil2100: the good news is that indicator-sound is building on ppc as you sort it out [08:27] didrocks, sil2100: so you probably don't need to skip ppc soon :p [08:27] seb128: well, we need that for unity and other stacks I guess [08:27] https://launchpad.net/builders/ross [08:27] ;p [08:28] ok, there is one on the 3 cases where the list maybe be shared [08:28] * didrocks tries that [08:28] sil2100: I think you are the first one to try to skip more than one componet :p [08:28] component :p [08:30] ok, let's see run 270 for build [08:31] (one all the prepare for unity finishes as it's taking all the slots) [08:38] :D [08:38] I mean, you want to run the build job explicitly for indicators? [08:38] Or just re-run the stack with foo? [08:38] The phone stack finally got unblocked too [08:38] sil2100: I run only the build job for now [08:38] sil2100: see the little "cu2d-indicators-head-2.1build" waiting :p [08:38] sil2100: I'm pretty confident that will fix the issue [08:38] sil2100: see run 270 ;) [08:38] sil2100: now, it's skipping the other on powerpc as well [08:38] * didrocks removes the debug info and commit [08:38] \o/ [08:39] sil2100: sorry for the bug, I'll rerun indicators with foo to unblock the stack [08:39] didrocks: ok, no problem - should I also run my gouranga script on the unity stack to skip powerpc for all? [08:39] sil2100: yeah, let's do that [08:40] sil2100: maybe you should patch cu2d-skip to take that arg [08:42] I'll patch it up in a moment, since I'd have to see where I could get the list of projects in cu2d-skip, since right now it's a grepping hack on a stack file ;) [08:43] This time I'll modify it and only skip those that need skipping, one moment [08:43] ok [08:48] seb128: indicator-session still FTBFS on powerpc FYI [08:49] didrocks, yeah, there is a real bug in the tests, infinity gave access to charles to one of his boxes for debugging yesterday but I don't know if charles managed to find anything [08:49] didrocks, I'm going to wait for charles to be around and make the build not fail on broken ppc tests if he doesn't have a fix today [08:49] seb128: seems at least, he didn't merge any potential fix :) [08:49] didrocks, no point blocking the archs we care about on a ppc issue [08:49] ok, thanks! [08:49] agreed [08:49] yw [08:51] sil2100: for libusermetrics, it seems you did a typo though [08:51] sil2100: mind skipping it? [08:52] didrocks: done [08:52] the tests pass, so then, indicators will be either in manual publish mode or will publish :p [08:52] sil2100: thanks! [08:52] btw. what would happen if the stack name is 'invented' ;p ? [08:52] I mean, non-existent ;p ? [08:54] Since I ran my modified script and one cu2d-skip had the 'stack' and 'project' arguments switched, so I skipped the unity project in the unity-scope-zotero stack ;( [08:55] sil2100: ah, I think cu2d-skip jenkins jobs just fail [08:55] as it can't touch the file [08:55] * didrocks checks [08:55] sil2100: indicators in manual publishing mode btw :p [08:56] sil2100: yep, it failed: http://10.97.0.1:8080/job/cu2d-skip-project-archs/60/console [09:15] sil2100: hum, didn't change global status on https://code.launchpad.net/~didrocks/cupstream2distro-config/unity-action-api-daily/+merge/174816? Let me do that now [09:17] sil2100: has you done packaging review for the indicators stacks btw? (see my ping 20 minutes ago) [09:18] * sil2100 backlogs 20 minutes ago [09:20] hmm, where did you ping me? [09:20] Ah [09:20] AAh [09:21] Ok, scratch that! [09:21] * sil2100 is modifying the cu2d-skip script [09:21] I'm reviewing [09:21] Sometimes when I hear 'packaging review' I start thinking 'hmm, what new packages do I need to modify packaging?' [09:22] sil2100, how's ap testing looking? any eta on landing to s? [09:23] mhr3: hi! Let's wait for todays testing results, we'll know the final status once it finishes [09:26] sil2100, very well [09:27] didrocks: packaging changes look ok, the indicator-session/sound are big ones, since there was this switch to CMake [09:27] didrocks: for ACKing: [09:27] http://10.97.0.1:8080/view/cu2d/view/Head/view/Indicators/job/cu2d-indicators-head-3.0publish/lastSuccessfulBuild/artifact/packaging_changes_indicator-applet_12.10.2+13.10.20130716-0ubuntu1.diff [09:27] http://10.97.0.1:8080/view/cu2d/view/Head/view/Indicators/job/cu2d-indicators-head-3.0publish/lastSuccessfulBuild/artifact/packaging_changes_indicator-session_12.10.5+13.10.20130716-0ubuntu1.diff [09:27] http://10.97.0.1:8080/view/cu2d/view/Head/view/Indicators/job/cu2d-indicators-head-3.0publish/lastSuccessfulBuild/artifact/packaging_changes_indicator-sound_12.10.2+13.10.20130716-0ubuntu1.diff [09:27] http://10.97.0.1:8080/view/cu2d/view/Head/view/Indicators/job/cu2d-indicators-head-3.0publish/lastSuccessfulBuild/artifact/packaging_changes_unity-gtk-module_0.0.0+13.10.20130716-0ubuntu1.diff [09:28] sil2100: hum, on the cmake, I would like to have a trailing comma if possible (but can be a following MP) [09:28] I hate ascii order for build-deps, it makes no sense :/ [09:28] but well… [09:28] larsu: please try to not use ascii order in build-deps, more logical order is better :p [09:28] sil2100: the change in unity-gtk-module is wrong though [09:29] sil2100: tedg replace a file in /etc/X11/ by an upstart job [09:29] didrocks: well, ascii order is usually forced by wrap-and-sort ;) [09:29] sil2100: yeah, and it sucks TBH ;) [09:29] first, I think it's better that we check the upstart job is working (we don't have all the same signals on desktop) [09:29] second, the file in /etc is a conffile [09:30] it needs to be removed in post* scripts [09:30] so this one has to be fixed [09:30] hm, can we have both at once? [09:31] sil2100: what's the point? he stops shipping it [09:31] so only people upgrading will have the file, it needs to be removed === vrruiz_ is now known as rvr [09:33] didrocks: ooh, I didn't know that was a thing. Will do from now on :) [09:33] larsu: thanks! we try to have in daily release: (like debhelper), (autoconf, cmake…), , [09:34] larsu, still order alphabetically in logical groups (like lib*) [09:34] yep ;) [09:34] * sil2100 reads up [09:34] that's nitpicking though :p [09:35] yeah, but we had coherence everywhere, would be good to keep it like that [09:35] in practice build tools end up first even if you use alphabetic order [09:35] cmake, debhelper, cdbs [09:35] otherwise, it will derive and everyone is going to be lost maintaining different packages with different rules [09:42] didrocks: ok, so I'm upgrading to the daily-build u-g-m to check if the upstart job works and gracefully removing that conffile when upgrading [09:43] sil2100: I'm sure it doesn't remove the conffile [09:43] I know [09:43] I said I'll be working on that ;p [09:43] ah, thanks! [09:43] Maybe wrongly phrased, hm [09:43] you will remove the conffile and see if the upstart job work :) [09:45] Yes, and then make u-g-m remove it automatically once it all works, hopefully [10:01] Laney, jfyi https://code.launchpad.net/~unity-team/unity8/unity8.background_gsettings/+merge/174958 [10:01] seb128: yay! [10:02] they did the testImage thing too, that's always good to see [10:02] unless that was borrowed from system-settings (which would be even better :P) [10:02] Laney, I pointed Cimi to system settings as a qml gsettings users so he might have copied it from there [10:03] fair enough [10:03] I was kind of scared they might just come up with some method of their own [10:03] like avoiding GSettings entirely [10:03] so I'm happy [10:04] especially after I heard the other day that it's not in scope for normal apps [10:04] it made me worried that we didn't care for using it [10:04] (at least for 13.10) [10:05] Laney, well, they seemed to still be discussing it for normal apps, they were asking for example of usecases where we see it as needed [10:05] sure [10:07] Laney, I'm looking at using it in the phone-app and the sound panel today for the ring tone [10:08] are we going to be defining a schema? [10:09] didrocks, sil2100: do we have a package atm that seems fit to add new gsettings key on the touch image? [10:09] e.g similar to gsettings-desktop-schemas but for Unity [10:09] seb128: not that I know of [10:09] Laney, can we define keys without a schemas? [10:09] ok, jenkins has been shut down/crashed [10:09] so all jobs are stuck [10:09] :-( [10:09] sil2100: rerunning them [10:09] I mean "we" as in system-settings [10:11] ;/ [10:12] didrocks: btw. do we have anything against dpkg-maintscript-helper and the Pre-Depends: dpkg (>= 1.15.7.2) ? [10:12] sil2100: no, it's the right way to deal with it :) [10:13] sil2100: interesting, the apps stack failed because of missing libdee-qt5-3 and qtdeclarative5-friends-plugin [10:13] Installing qtdeclarative5-friends-plugin as Depends of share-app [10:14] sil2100: I'm adding that one, one sec [10:14] hm, interesting [10:16] sil2100: mind acking? https://code.launchpad.net/~didrocks/cupstream2distro-config/refresh_apps_deps/+merge/174964 [10:17] * didrocks relaunchs the stack meanwhile now that it's deployed [10:18] didrocks: hmmm, there is share-app twice in there? [10:18] didrocks: in the packages: list [10:18] sil2100: argh, wrong copy/paste [10:18] :) [10:19] thanks for spotting it [10:19] fixing :p [10:19] The clipboard is teh devil sometimes [10:19] ;) [10:20] sil2100: pushed rev 515 [10:22] didrocks: approved! [10:25] sil2100: thanks! [10:25] Laney, ah, good question ... I was still pondering the options for the ring tone key, system settings or phone app, but eventually we will need a gsettings-desktop-schemas equivalent I think [10:27] yeah probably [10:28] As long as everyone's agreed on using GSettings for stuff like that I'm happy === sil2100_ is now known as sil2100 [10:39] seb128: have you seen that the license list scrolls over the header in s-s? [10:39] s/license/package/ (in the license screen) === huats_ is now known as huats [10:44] didrocks: I need to jump out for a moment to the store, could you take a look at https://code.launchpad.net/~sil2100/unity-gtk-module/remove_conffile/+merge/174969 in the meantime? :) I hope sym-links are ok [10:45] sil2100: hum [10:45] I think you have a better way [10:45] sil2100: btw, your local test run? [10:45] removing the file manually [10:45] and the job comes up after that? [10:45] (at login) [10:46] sil2100: man dh_installdeb [10:47] package.maintscript [10:47] that will create the maintainer scripts file [10:47] without you having to deal with symlinks [10:48] didrocks: the local test-run showed that GTK_MODULES and the other variable are set correctly [10:49] Oh, ok, I think I remember that solution, will use that instead [10:49] sil2100: ah good, so just this minor change ^ [10:49] and we are good to merge/rebuild the component :) [10:49] and then, publish! [10:49] with indicator-session stuck in proposed though [10:50] Laney, yeah, I just saw that, the header also doesn't go away when scrolling where it should, Saviq just pointed that to me a bit earlier in #ubuntu-unity [10:51] seb128: It's OK I've fixed it now [10:51] will MP in a second [10:51] Laney, what was the issue? [10:51] k, will see that [10:51] is that a clipping issue? [10:51] use anchors.fill: parent [10:52] you were setting the height which I guess messed it up [10:52] hrm, the scrollbars aren't very good [10:52] I tried adding one so you could scroll that huge list faster but I can't actually grab it [10:53] Laney, savig recommend "to increase max flicking velocity in "Software licenses" :D" [10:53] haha [10:53] we went for: [10:53] maximumFlickVelocity: height * 10 [10:53] flickDeceleration: height * 2 [10:53] that's what they did in unity8 [10:53] but we don't have the Flickable there [10:53] so it'd be global [10:53] maybe that's OK? [10:54] Laney, what do you mean, don't you have a ListView there? [10:54] Laney, but anyway yeah, I'd say that's good to be global [10:54] oh, I thought you set it on the Flickable [10:54] Laney, ListView is a Flickable [10:54] ah [10:54] a subclass, that is [10:55] let me try that then [10:55] Saviq: is it known that the Scrollbar isn't very easy to interact with? [10:56] Laney, I don't think it got the desired attention [10:56] Laney, or that it in fact is supposed to be interactive? [10:56] oh, it isn't? [10:56] Laney, but we're not using it in unity8 so can't comment much [10:56] it says it provides "Scrolling functionality" [10:56] ... [10:57] oh well [10:58] those changes seem nicer, yeah [10:58] cool! [10:59] ;-) [11:01] didrocks: if anything, pushed the modified version of the branch ;) === greyback is now known as greyback|food [11:05] sil2100: hum, the version isn't right [11:05] it should be 20130716 [11:06] "prior-version should be based on the version of the package that you [11:06] are now preparing" [11:06] from the manpage [11:07] should even be 20130716.1 now :) === slomo_ is now known as slomo === MacSlow is now known as MacSlow|lunch [11:25] sil2100: to get things moving, I've published the apps stack FYI [11:25] robru: FYI ^ [11:26] sil2100: same for phones and sdk stacks (because of the 2 new packages in distro) [11:27] seb128: sil2100: mind looking at https://code.launchpad.net/~didrocks/cupstream2distro-config/remove-unity-common/+merge/174979? [11:27] (it passed, I published it and unity tests are now running) [11:40] didrocks: looking [11:40] didrocks: and fixing ;) The polish description wasn't really acurate in the manpage [11:41] sil2100: same in the french one [11:41] sil2100: I wonder if it's just not an old manpage with wrong info [11:41] when I read it, it struck me [11:41] hence LANG=C man… :p [11:41] brr ;) [11:43] didrocks: pushed! [11:44] sil2100: approved :) once merged, feel free to relaunch that component build in the indicator stack [11:45] so, in a nutshell: [11:45]