=== jono is now known as Guest11562 === jono is now known as Guest60019 [06:30] good morning desktopers [07:00] good morning [07:04] hey didrocks [07:05] swinging back on the 8am start? ;-) [07:05] salut seb128, ça va ? [07:05] yeah ;) [07:05] ouais, et toi ? [07:05] no pitti, I think I was first for once :p [07:05] hi all :) [07:05] heh [07:05] hey hikiko [07:05] seb128: well, you then leave for some minutes, so it's cheating :) [07:05] seb128: encore un ou deux trucs dans la gorge, mais ça va :) [07:06] cool [07:06] no :) I was first I just forgot to say hi!! [07:06] hehe [07:06] (hi!!) :P [07:06] hikiko, saying hi is part of the game ;-) [07:06] hey hikiko! [07:06] hey hikiko [07:07] hey seb128 hey didrocks hey pitty ! [07:25] hey seb128 didrocks [07:25] hey darkxst [07:26] webkit2gtk FTBFS with a segfault in ld bfd on arm64, ran out of ram perhaps? [07:26] that will block the transitions [07:28] right, I just retried that a minute ago [07:28] retried indicator-datetime on ppcel64 as well [07:28] but trying to read http://people.canonical.com/~ubuntu-archive/proposed-migration/update_output.txt [07:28] there seems to be more [07:28] evening darkxst [07:29] I wonder if the 2 big groups should be hinted to go as one [07:29] seb128, yeh, but I couldnt find anything else looking through the GNOME stuff [07:30] like britney is not trying to make e-d-s and gnome-desktop go together [07:30] and they need to [07:30] need Laney, I don't know how to hint britney to do that [07:30] didrocks, ^ or do you know? [07:31] seb128: I can't hint more than touch things [07:31] k [07:31] seb128: and I don't think you can either [07:31] so let's wait for Laney [07:31] yep [07:38] seb128: Are there any common devices (not made by IBM) that use the ppc archs? [07:38] And not old Macs :) [07:38] duflu, I've no idea sorry [07:38] Watson... I guess. [07:39] Oh, and PS3 is a relation of ppc [07:40] And Wii [07:46] funny that some try to automate installation with pexpect and such, and don't use --help to see that there are switches already doing automation for you :) [07:47] (and way more reliable than pexpect) [07:48] nice, a new contributor! [07:48] * didrocks really likes when new contributions can add a new framework with just few lines: https://github.com/ubuntu/ubuntu-make/pull/170/files [07:48] (and all declarative) [07:48] duflu, there are embedded ppc chips, probably not exactly common though [08:08] bah, I hate autotools [08:41] Laney, darkxst, bug #1512435 has changes to review, I would welcome your opinions on those [08:41] bug 1512435 in unity-control-center (Ubuntu) "Stop depending on gnome-control-center" [High,In progress] https://launchpad.net/bugs/1512435 [08:45] good morning! [08:45] hey larsu, how are you? [08:47] seb128: great, thanks! Sunny morning here again [08:47] seb128: how are you? [08:47] same here :-) [08:47] nice :) [08:48] good morning larsu! [08:48] sunny morning, lucky you guys [08:48] cloudy here [08:48] and not digital clouds [08:49] bonjour didrocks! [08:49] haha [08:49] seb128, I would lean towards not sharing the data, with g-c-c in universe and most normal users not having both packages installed (and its small anyway), doesnt seem a huge issue if a few static images are duplicated? [08:50] darkxst, the issue is that things look to /usr/share/pixmaps/faces not in private subdirs [08:50] this is supposed to be a cross desktop standard location [08:51] so we can't easily ship the files in both packages [08:51] they would conflict [08:52] seb128, right [08:53] we could teach u-c-c to look in a private dir first and then to the normal one and include its images in a private subdir, but then we would have icons listed twice when g-c-c is installed [08:53] or am I missing something obvious? [08:55] seb128, probably they would be duplicated [08:55] the pkla change will also work for ubuntu GNOME? [08:56] just keep the shared faces package I guess [08:56] ubuntu-gnome-desktop recommends policykit-desktop-privileges [08:56] so it should work yes [08:56] morning [08:57] I did a small experiment these days: https://www.youtube.com/watch?v=UYeFN0vOG_M and I think we could have shadows for any window (here I get the window shape for shaped windows) I don't know how to do it in compiz yet but the idea I had at the sprint (that we can just draw the shape twice in the texture and do some visual effects so that it looks like a shadow) seems to be feasible ( desrt willcooke ) [08:57] good morning willcooke [08:57] darkxst, that binary include permissions for other things like mounting disks [08:58] seb128, well except for the idiots who insist on installing with --no-recommends, and then complain stuff is broken, maybe it should be a dep instead? [08:58] your call, but it seems like it should [08:58] hi hikiko - cool stuff! [08:59] you can look to /var/lib/polkit-1/localauthority/10-vendor.d/com.ubuntu.desktop.pkla to see what is included in there [08:59] hikiko, nice! [08:59] hey willcooke, darkxst [08:59] hey willcooke [08:59] hey larsu, willcooke [09:00] darkxst, btw new gjs has failing test on powerpc, I retried 3 times it seems to fail consistently [09:00] build is failing on that [09:00] hi larsu darkxst :) [09:01] hello desktoppers [09:01] hi Sweet5hark [09:02] good morning!!!! [09:02] hey Sweet5hark [09:02] hey Laney [09:02] hey hikiko [09:02] * Sweet5hark looks at backlog, sees complains about clouds. [09:03] we have end of the world fog here since Monday morning. [09:03] same [09:04] less bad today though [09:04] seb128, probably an endiness bug, but I have no ppc hardware to debug, just revert to ignoring ppc tests? [09:04] darkxst, or talk to mbielb maybe he's interesting in debugging it? [09:04] it was that way for a long time, until someone came along and fixed them [09:05] morning Laney! [09:05] hey Laney! [09:05] morning Sweet5hark [09:05] seb128, I have a feeling debian ignore tests more or less always [09:05] darkxst, doesn't hurt to mention those issues to them in any case [09:06] hi larsu didrocks [09:06] how's it going [09:07] very good thanks [09:07] how about you? [09:08] also good! [09:08] pub quiz tonight ;-) [09:08] nice! [09:08] * seb128 feels ignored by Laney not saying hi :-( [09:08] oh no [09:09] * Laney purrs at seb128 [09:09] :-) [09:09] here for you man [09:09] * seb128 is happier now [09:10] * seb128 pets Laney [09:10] * Sweet5hark just lighted the IR/redlight/heat lamp. screw you, Hamburg weather. ¯\_(ツ)_/¯ [09:14] seb128, I will ping him' [09:20] seb128, I guess policykit-desktop-privileges recommends comes from one of the foundation seeds, and not ours [09:22] could be [09:22] thanks for the ping [09:22] * Laney eyes wk2 [09:22] oh it failed before too, interesting [09:23] oh no clicked the wrong thing [09:23] * Laney fails [09:25] Laney, can you tell britney to try to batch the world in one go? ;-) [09:25] it won't work [09:25] unity [09:25] epiphany [09:27] well, it would tell us if those are the only ones and it's going to be needed at some point since it doesn't try groups together, no? [09:28] shrug, the webkit2gtk arm64 retry failed :-/ [09:28] I've a feeling it's going to be an annoying one [09:30] the list it has seems right [09:30] k, I think I'm just clueless with britney's output then [09:30] I don't get why e-d-s and gnome-desktop3 are not in one set when they have common components [09:31] they can't success without the other one [09:32] look for this one [09:32] Trying easy from autohinter: gnome-contacts/3.18.0-1ubuntu2 folks/0.11.1-2build2 evolution-data-server/3.18.1-1ubuntu1 [09:32] compiz needs rebuild against new libmetacity-private [09:32] gnome-ssession is blocked at excuses [09:32] speaking of which, when is excuses updating? [09:33] it's over 3 hours old [09:33] is that normal? [09:33] every time proposed changes [09:33] or tests come in? [09:33] I retried tests and fixing stuff for like 2 hours [09:33] I think [09:33] look at the log/ [09:34] I don't understand why gnome-session doesn't pick the firefox i386 status [09:34] it's in progress but the test failed http://autopkgtest.ubuntu.com/packages/f/firefox/xenial/i386/ [09:34] I even did a retried that failed as well [09:35] but that didn't success to change the status on the summary [09:37] hopefully it will re-run now that I did an upload [09:37] if nothing else [09:38] it's supposed to see that there are tests in progress and try to run if there are [09:38] I did retry builds that failed before and successed this time [09:38] that's not enough to trigger a refresh either? [09:39] anyway [09:39] let's see if it does update [09:39] did anyone look at the Trevinho silo? [09:40] I didn't [09:40] I though you were handling it [09:41] yeah left some comments last night though [09:42] yeah, doesn't seem like there was any activity after that [09:42] ok [09:42] I called it a day around the time you commente [09:42] d [09:42] but there is nothing in the night irc log [09:42] need to wait for Trevinho I guess [09:42] seb128: actually I was here for an hour or so, but I forgot to open IRC -_- [09:43] sorry, good morning [09:43] hey Trevinho [09:43] hi! [09:43] Laney: so... did you see my comments? [09:43] Laney: hi! [09:43] sure [09:43] you are saying that you want it to be ~bzr? [09:43] * Trevinho didin't check th reply :P [09:43] didn't reply [09:43] sso was a barrier too far [09:44] ah... :D [09:44] and tell me about this {} thing in .install [09:44] Laney: well, I didn't release so I thought it was a way to have the ubuntu version bumped (as pre-0.5.2 releease), but not the upstream one [09:44] which one? [09:44] I read the comment but I didn't understand, or maybe didn't grep too well [09:45] (that branch is quite old) [09:45] https://ci-train.ubuntu.com/job/ubuntu-landing-011-2-publish/91/artifact/bamf_packaging_changes.diff [09:45] ~bzr means the symbols get a weird version [09:45] Laney, Trevinho, that looks like trying to handle case where the package is multiarched or not [09:46] is that dynamic/magic depending of the debhelper version/serie? [09:46] that confuses me [09:46] and I don't think dh_install really officially supports it, there used to be a lintian tag for that [09:46] maybe still is [09:46] for what ? [09:47] {,}? [09:47] yes [09:47] I saw it used in other places I think [09:49] * Laney shrugs, it's not supported [09:49] could well be, I don't have any example on disk [09:50] don't see any weird magic in the packaging either [09:50] Laney: ah, the {*,} was because I wanted to support packaging even where's there's no multi-arch, but maybev it's not needed [09:50] would better be used for dh_exec, no? [09:51] Laney: anyway any change you want, I cand do them [09:52] seb128: looks like britney just finished a run which took ages [09:52] so should start another one... [09:53] Laney, ok, thanks [09:54] already did actually [09:55] great [09:55] Trevinho: I would just remove that change if I were you [09:55] shrug, why gnome-session thinks that the firefox autopkgtest is still in progress [09:55] it probably had the results from before [09:56] it waited for like 90 minutes at the end [09:56] don't know why [09:56] same gtk+3.0 thinks that gnome-photos fails [09:56] when it doesn't, retry worked [09:56] like the previous run was [09:56] Tue, 03 Nov 2015 07:27:05 +0000 [09:56] STATS: [09:56] Tue, 03 Nov 2015 07:31:08 +0000 [09:56] but this one... [09:57] oops pasted the wrong bit [09:58] Laney: I've spotted another small regression , so I'm going to rebuild anwyay [09:58] Laney: "that change" is the .install thing= [09:58] ? [09:59] 07:26:25 -> 07:27:05 vs 08:13:08 -> 09:37:48 [09:59] Trevinho: yes [09:59] popey: popey, old friend! [09:59] Good morning Mr Shark [09:59] Trevinho: I don't understand what you're saying about it not being an upstream release though [09:59] popey: which were the ppa again that you wanted to have libreoffice-vanilla stuff into? [09:59] what is the difference between this upload and a 0.5.2 release? [09:59] stupid me didnt copy that. [10:00] Laney: to release I mean I should do the tarball, close bugs... All these annoying stuff :P [10:00] Sweet5hark, https://launchpad.net/~canonical-community/+archive/ubuntu/ppa [10:00] Sweet5hark, i use that one for building then copy to /stable or /debug [10:00] popey: ah, ok. [10:01] Trevinho: alright [10:01] lemme know when it's ready again [10:03] popey: I created/pushed lp:~libreoffice/ubuntu/+source/libreoffice-vanilla in the libreoffice packaging team alongside the repo for debian packaging: https://code.launchpad.net/~libreoffice/ubuntu/+source/libreoffice-vanilla/+git/libreoffice-vanilla [10:04] popey: we could move that to ~canonical-community, but maybe it makes more sense to keep it with the other libreoffice stuff. [10:05] It does make more sense, yes. [10:06] popey: k [10:07] will download the latest modifications from ~canonical-community then and commit/push them to git then. And then prepare a 5.0.3 build [10:10] Sweet5hark, magic, thanks! [10:10] Laney, so should we do anything for those autopkg status to be updated or just wait? [10:12] some of them are being updated [10:12] I'll just retry all the regressions [10:12] the ones that get stuck as "in progress" are more annoying [10:16] popey: here is a crazy thought: should we maybe make the libreoffice-vanilla source package build LibreOffice twice (once with debug symbols, once without) and provide two binaries, instead of having two branches? [10:17] That does sound crazy, it could be very handy to have the debug build, especially on the phone, where this stuff hasn't been run much before. [10:18] popey: advantages: no fiddling with multiple branches and multiple uploads, just one fire-and-foget source package disadvantages: takes even longer to build, not so easy to do a quick do-not-need-the-debug-stuff-build .. [10:19] popey: you would have debug symbols. In fact you would have debug symbols all the time. [10:20] Sweet5hark, I think it's worth doing, yes. [10:21] popey: you would have libreoffice-vanilla (source package) build twice and create two binary packages: libreoffice-vanilla and libreoffice-vanilla-debug (which wouldnt be installable at the same time). [10:21] Sweet5hark, that works for me, we could modify the build script to choose which build to integrate [10:22] popey: and heck thats gonna be driving with a truck over the debian policy manual, but for these builds, I dont think we care much. [10:22] Sweet5hark, well, we take the deb, pick it apart and repackage it anyway ;) [10:22] Sweet5hark, popey, hi, why aren't you stripping the dbg symbols in a different package? [10:22] popey: ^^see [10:22] or even let the ppa-builders create the dbgsym package automatically [10:23] doing two builds are quite crazy [10:23] ricotz: Im kinda reluctant to depend too hard on features of the debian packaging tools for this. [10:24] so use the dbgsym feature [10:24] I personally don't mind how it's done. [10:26] ricotz: dpgsym feature _is_ a feature of debian packaging (and it would require changes on how we get the stuff out of the package/repackage again) [10:27] Sweet5hark, are you sure you know what I mean? [10:28] Sweet5hark, just don't strip the debugs symbols and turn on the dbgsym creation in the ppa-settings [10:30] ricotz: yes, and then we have multiple packages that need to be installed alongside each other, have dependencies etc. -- exactly what we what to avoid in the snappy world ... [10:32] popey: or is that ^^ ok with you/the snap-creation tools these days? [10:37] Sweet5hark, we yank the debs from the archive, unpack them and put them in the right places. if there are two locations, or two package names we can pick from, we can do that. [10:38] Sweet5hark, to be honest I expect to rarely (hah) need the debug builds, but when I need them I don't want to have to trigger some 10 hour build, but want to just switch to the debug build quickly [10:40] seb128: just retried all regressions / in progress tests [10:40] http://autopkgtest.ubuntu.com/running.html [10:40] Laney, great [10:40] is there some magic command that does that? [10:41] popey: well, what ricotz is suggesting is to have two packages: the good old libreoffice-vanilla package (without symbols) and second one that you can install on top to get debug symbols. that the canonical old school debian packaging way. [10:41] yeah https://wiki.ubuntu.com/ProposedMigration/AutopkgtestInfrastructure#Re-running_tests [10:41] Sweet5hark, I'm fine with that. [10:41] I hacked it to consider in progress too, not just regressions [10:42] (don't really understand why tests get stuck at in progress) [10:42] Laney, great [10:42] popey: I dont know how that works with whatever you do for snappifying. Then again, you dont need the symbols snappified, I guess as you only use it on proper desktops ... [10:43] popey: ok. [10:43] Sweet5hark, we can work around whatever you do really :) I can just have two options in cmake, which can deal with whatever you do [10:44] popey: kk [10:45] popey, so you are post-processing the built deb and not using it as is? [10:46] ricotz, kinda, we unpack it and bake it into a click package for the phone [10:46] popey, so just build a debug package without stripping anything, and strip the binaries in your process as needed? [10:47] Hm, not sure about that. [10:48] (I mean build *one* binary package which still includes the symbols) [10:49] the debug build is significantly larger, and we don't need debug much of the time [10:50] popey, you can strip those in your repackging process? [10:50] and build two snappy packages then [10:50] yeah, means downloading giant package and stripping it. [10:51] which is way faster than building 10hours twice [10:51] we aren't (currently) using snappy, it's click on the phone, but I see your point [10:51] ok, make your choice then [10:58] hi [10:58] havent seen robert_ancell in a while [10:59] (especially since I try to follow natural day night cycles.. recently) [10:59] could someone else look into updatint libinput: https://bugs.launchpad.net/ubuntu/+source/libinput/+bug/1504168 [10:59] Launchpad bug 1504168 in libinput (Ubuntu) "fix handling of mx4 touch screen" [Undecided,New] [11:01] are those patches committed? [11:03] hello from xenial [11:04] Laney: the last one was reviewed and approved but relies on the two before.. [11:05] the last iteration only had a problem in an edge case and two findings in the debug tools [11:05] last one .. as in third patch in the list.. [11:11] Laney: we already carry previous versions of that patch series.. [11:11] I am also ok with first using v7 which was seen by more eyes than v8 [11:16] Trevinho: I can't build your branch :( [11:17] Trevinho: http://paste.ubuntu.com/13090807/ [11:17] looks unrelated, but I really don [11:17] *don't care for tracking down boost problems :/ [11:19] larsu: ah, yeah... on xenial... [11:19] larsu: try to merge with xenial proposed branch? [11:19] where is that? [11:19] larsu: or comment that line in launchercontroller, it's not that important [11:19] lol [11:19] lp:~ci-train-bot/unity/unity-ubuntu-xenial-landing-011 [11:20] thanks - let's try this [11:29] Trevinho: http://paste.ubuntu.com/13090856/ [11:31] larsu: ouch, yeah... you need new bamf then [11:31] larsu: it's just better to comment that line in LaucnherCOntroller [11:31] or replace it with the fixed one.... (which I can paste if youn want) [11:32] * larsu wonders how this can even happen [11:33] Trevinho: commenting line 153 is enough? [11:33] yes... [11:33] larsu: it happens because in xenial libsigc++ changed, and they removed some stuff as now they can be done with c++11 [11:34] Trevinho: they are not backwards compatible? Hm, I guess that's ok... I changed my entire OS to the next version after all [11:35] * Trevinho now has a new shiny and cheap chinese smart plug... With openwrt. My office lights will be finally in sync with my PC :) [11:35] larsu: no, they're not [11:37] larsu: they removed some classes such as sigc::group and the sigc::_N ... which were used to do computations on the fly (specifying how to handle arguments when connecting the callback) [11:41] Trevinho: bah! Now it fails because of -Werror and a signed/unsigned comparison [11:42] in a nux header no less [11:42] SIGH [11:42] larsu: fuck that compiler ! [11:42] :P [11:42] That's weird though, we didn't get that in the ppa [11:43] in NuxCore/PropertyOperators.h:191 [11:43] larsu: it's too early to use xenial :P, I just use an lxc for now :) [11:43] larsu: mh I believe that's not where the issue is, but where the property is used [11:43] that's what the error line says [11:43] let me check [11:43] larsu: look in the log, I guess it's inside switcher model maybe? [11:44] Laney: anything I can do for my PPU application? [11:45] happyaron: just wait, shouldn't be too long [11:45] ok [11:45] Trevinho: yes through the SwitcherModel, but only because it includes Nux' Property.h indirectly [11:46] Laney: silo should be ready now [11:46] larsu: I guess it's using a call somehwere, you can or change the 0 comparisons with [11:46] 0U [11:47] or... change detail_selection_index to detail_selection_index() [11:47] in the system header? [11:47] larsu: can you paste me the whole error? [11:47] cause it seems something else to what I though [11:47] yes, it is something else than what you think :) [11:47] already scrolled passed, sorry [11:48] cause yesterday I got this inside CI, but I fixed in that way [11:48] building again without -Werror [11:48] yeah, that's fine [11:48] I hope ... just removed it from CMakeLists.txt [11:54] didrocks, Heard back from Cadsoft about Eagle. They are +1 in principal, just need to get the legal stuff sorted [12:04] * desrt yawns [12:05] morning desrt [12:05] good morning [12:05] * desrt will be out for a couple of hours walking around before starting work this morning :) [12:05] adjusting to the timezone? [12:05] willcooke: nice! :) [12:05] yup [12:06] enjoy! [12:07] desrt: break some nice portals ;) [12:07] didrocks: we're heading for a part of town that is such a massive res f8 that it's ridiculous [12:07] must be 100 p8 out here [12:08] enjoy ;) [12:08] Trevinho: http://paste.ubuntu.com/13091016/ [12:08] Trevinho: I think I will stop this now. [12:08] didrocks: thanks :) [12:10] * larsu is off for lunch [12:11] larsu: yeah, that's fine... Also because hikiko just shown me something that we could use instead of this workaround [12:11] so... let's see how that goes [12:11] We'd still need to add an atom to gtk windows for being able to recognize them, but not the radius [12:11] oh ok [12:12] off to lunch now - let me know when you have asomething I can try === dpm__ is now known as dpm [12:14] Trevinho: done [12:15] Laney: ❤️ [12:16] ARGH [12:16] Trevinho: 2015-11-03 12:15:42,070 ERROR Needs review: https://code.launchpad.net/~3v1n0/unity/add-launcher-icon-to-dash/+merge/275968 [12:16] 2015-11-03 12:15:42,070 ERROR unity has merges in bad states [12:16] soz [12:20] Laney: ohuch, yeah andyrock forgot to top-approve.. [12:20] I did it myself [12:22] ok, ty [12:24] Good morning [12:29] hey pitti! [12:32] hey pitti [12:32] how's austin? [12:33] hey pitti [12:35] still dark -- woke up at 5:30 again :) [12:37] poor roommate :P [12:38] it's actually not even existant yet -- thick fog, can't see a thing :) [12:38] * pitti wanted to go running this morning, meh [12:38] Laney: nah, barry just woke up, he was fine === alf is now known as alf_ [13:00] ok, going to run for the first time after last week being sick, wish me luck! [13:01] good luck didrocks [13:01] thanks willcooke ;) [13:04]