[00:00] cjwatson, nothing is needed. everything is built and good and ready to publish. [00:00] or, wait, is it this set that were built wrongly and we need to rebuild [00:00] cjwatson, citrain just crapped itself and refuses to publish the last 8 or so packages, out of 117 total packages [00:00] why don't we just manually (i.e. by core-dev) copy the source to trusty-proposed without binaries [00:00] let it rebuild in the primary archive [00:00] cjwatson, ok, be my guest. i'm not a core dev [00:01] cjwatson, http://162.213.34.102/job/landing-006-2-publish/39/console grep this page for 'Checking', those are the packages that need to be copied. [00:01] everything is blocked in -proposed anyway until didrocks returns, so it's not like it's going to hit the archive without anyone looking at it [00:01] we may have to do another AP run with those packages shoved in manually from -proposed, or something [00:03] that list isn't totally accurate [00:03] only five of those six failed for that reason [00:03] qml-box2d failed because trusty-proposed already has version 0.1~git20140412 [00:04] that was uploaded three hours ago by Ken, by the looks of it [00:04] ugh [00:05] however, the ubuntu1 upload was apparently just a badly-versioned no-change rebuild [00:05] (should've been build1) [00:06] right [00:06] in theory the other packages were rebuilt already [00:06] I don't know whether Ken's upload covers that [00:06] Depends: libc6 (>= 2.4), libgcc1 (>= 1:4.4.0), libqt5core5a (>= 5.0.2), libqt5gui5 (>= 5.0.2), libqt5qml5 (>= 5.0.2), libqt5quick5 (>= 5.2.0), libstdc++6 (>= 4.4.0) [00:06] so it should be safe to copy the binary packages as well [00:06] that looks like it's at least partly 5.2 [00:06] rsalveti: oh, so it's just citrain being too clueless to find the right version to copy, hmm [00:07] cjwatson: exactly [00:07] well that's easy, but let me check [00:07] we were at the last step [00:07] this versioning is appalling, people need to learn to use build1 for no-source-change reuploads of stuff previously in sync with Debian [00:08] screws up our autosync processes otherwise [00:09] indeed [00:09] pokerth copied, let's see if that works [00:09] yep [00:13] ok, all those five (peg-e, pokerth, qtcreator, qtcreator-plugin-ubuntu, qtwebkit-opensource-src) copied [00:13] cjwatson: great, thanks [00:13] for qml-box2d ... my suggestion would be leave it and see if it's actually a problem, personally [00:13] right [00:13] maybe Mirv will show up and advise [00:14] robru: u1db-qt is good as well, we can land that [00:14] getting some dinner for real now [00:14] bbl [00:14] but the dependencies of the armhf build in landing-006 are identical to the dependencies of the armhf build of Ken's upload to trusty-proposed [00:14] so I think it'll actually be fine [00:16] cjwatson, can you also copy u1db-qt from the silo? we rebuilt that [00:16] cjwatson, mirv not expected for at least 4 hours [00:17] cjwatson, "ubuntu1" shouldn't interfere with autosync because citrain is basically only for stuff that we're upstream of (eg, none of this stuff came from debian anyway) [00:17] mkay, copied [00:17] robru: false [00:18] pokerth was previously synced from Debian [00:18] ok [00:18] so was peg-e [00:18] I wouldn't have made the autosync comment if I hadn't checked :) [00:18] (I operate the autosync ...) [00:19] cjwatson, ok fair, didn't know that. thanks for copying u1db [00:33] https://code.launchpad.net/~cjwatson/cupstream2distro/fix-publication-sorting/+merge/210948 should fix this; will of course want Didier's review [00:35] rsalveti: I'm back now what's up? [00:36] cyphermox, oh, I think we just needed a core dev to copy some packages, but cjwatson took care of it. [00:36] very well [00:36] thanks cjwatson [00:36] I had dinner, then had to run some errands [00:37] cyphermox, no worries [00:37] well. analyse a ton of stuff and then copy, yes. :) [00:37] ahah [00:37] cjwatson: for my edification, version checking and stuff? [00:37] cyphermox: yes, effectively manually cleaning up after https://code.launchpad.net/~cjwatson/cupstream2distro/fix-publication-sorting/+merge/210948 [00:37] cyphermox, i'm writing an email summary right now [00:37] which was my bug in the first place so it is right and proper that I get to clean up after it [00:38] oh, ah [00:38] I'll give the mp a look for review [00:38] citrain got confused by there having been a newer version of qtwebkit-opensource-src in landing-006 at one point which was later removed [00:39] and inappropriately picked the latest version rather than the most recent one [00:39] oh, cool [00:39] yep [00:40] the mp isn't urgent at this point [00:41] well, I can save didrocks the trouble and give it a look too [00:41] also accepted the binaries from Ken's qml-box2d upload from NEW [00:41] ok [00:42] hey, I already have to override my "implicitly trust cjwatson's code" rule to look at the MP. ;) [00:43] do NOT implicit trust my code written after midnight. :) [00:43] *implicitly (spelling too) [00:44] cjwatson, but could you not describe any time of day as "*after* midnight"? [00:44] * cjwatson disregards the pedantry :) [00:45] hehe [00:45] true, it's probably very very late for you [00:45] * cjwatson starts looking through -proposed [00:45] address-book-app is blocked on (a) address-book-service building on powerpc (unalaysed) (b) https://code.launchpad.net/~cjwatson/ubuntu-ui-toolkit/arch-any/+merge/209988 [00:46] robru: so it's past my EOD, I'll watch IRC if you ping loudly [00:46] *unanalysed. yeesh [00:46] we could disable that build on powerpc temporarily, or force it past p-m [00:46] robru: I got some school python game code to write, apparently [00:46] cjwatson: yeah [00:47] dialer-app has unsatisfiable deps on arm64/powerpc/ppc64el which I think basically trace back to the same ubuntu-ui-toolkit merge [00:48] ditto friends-app, gallery-app [00:48] indicator-network/powerpc unsatisfiable Depends: unity8 (>= 7.82) [00:48] cyphermox, heh, I EOD in 15 but I got some personal stuff to work on so i'll be around [00:48] robru: ah? [00:48] then maybe I'll spend some more cycles watching the stuff [00:48] messaging-app, notes-app -> ubuntu-ui-toolkit/arch-any [00:48] ugh [00:49] pyqt5 autopkgtest fails [00:49] cyphermox, you mean other landings? because as far as I can tell with qt, there's nothing to be done until didier wakes up and fixes citrain [00:49] https://jenkins.qa.ubuntu.com/view/Trusty/view/AutoPkgTest/job/trusty-adt-pyqt5/71/ [00:49] robru: well, all this stuff in proposed... [00:50] it shouldn't have made it so far, I'd say, except maybe for the parts that are wiating on ui-toolkit [00:50] cyphermox, yeah, exactly. didier blocked it in proposed on purpose, so no matter what it's gonna stay stuck there until didier wakes up [00:50] https://jenkins.qa.ubuntu.com/view/Trusty/view/AutoPkgTest/job/trusty-adt-ubuntu-purchase-service/ fails [00:50] blocked it in proposed on purpose? [00:50] cyphermox, yep. didier wanted all the qt stuff in -proposed, and wanted it to stay there until tomorrow. [00:50] I don't like the look of the pyqt5 autopkgtest failure; that looks real. can somebody look? [00:51] ImportError: /usr/lib/python3/dist-packages/PyQt5/QtWebKit.cpython-33m-x86_64-linux-gnu.so: undefined symbol: _ZN18QWebSecurityOrigin26removeAccessWhitelistEntryERK7QStringS2_NS_16SubdomainSettingE [00:51] robru: define qt stuff, as qt proper or as anything that builds against qt from citrain? [00:51] cyphermox: he blocked the whole landing [00:51] ok [00:51] cyphermox, the entire qt5.2 silo, every single package, didier blocked them all [00:52] cjwatson, got a link for that pyqt failure? [00:52] ubuntu-purchase-service failure should go away on retry, though not sure whether that'll happen automatically [00:52] robru: https://jenkins.qa.ubuntu.com/view/Trusty/view/AutoPkgTest/job/trusty-adt-pyqt5/71/ [00:53] * cjwatson goes to poke u-p-s [00:53] cjwatson, ubuntu-purchase-service is not actually used anywhere right now [00:53] thought we had removed it [00:53] the failure is noise anyway [00:53] I've retried it in which case it's moot [00:54] telepathy-ofono/ppc64el unsatisfiable Depends: ofono [00:54] cjwatson, oh, well, that failure is a failure to import qtwebkit, which was one of the ones that didn't make it to -proposed in the first pass. I assume a rebuild would fix that now [00:55] oh, could be [00:55] cjwatson, things were looking pretty good in the silo, so pretty much I would just assume to rebuild all failures in -proposed... ;-) [00:55] it might have raced with the qtwebkit update for some reason [00:56] cjwatson, "raced" in the sense that it got uploaded to -proposed hours before qtwebkit was [00:57] robru: no, this latest run took the latest qtwebkit [00:57] Get:106 http://ftpmaster.internal/ubuntu/ trusty-proposed/main libqt5webkit5 amd64 5.1.1-1ubuntu6 [8,116 kB] [00:57] hummm, then i have no idea [00:57] needs investigation [00:58] ubuntu-purchase-service passed on a retry, good [00:59] this pyqt5 was built against qtwebkit 5.2 [00:59] that's why it's failing [01:00] I am strongly inclined to just upload a no-change rebuild to trusty-proposed [01:00] any objections? [01:00] it's not one we're upstream for [01:00] nope [01:00] no objections from me [01:00] as long as things remain blocked if they're supposed to be [01:01] yeah, blocks are unversioned [01:01] ok, it's in the hints [01:03] (somebody should tell me at some point why we tried qtwebkit 5.2 and then backed out, for my edification) [01:04] pyqt5 rebuild uploaded [01:04] cjwatson, rsalveti might know that. [01:04] wonder if I can manage a quick build log scan for other instances of qtwebkit 5.2 [01:05] slightly concerned that that went unnoticed [01:05] cjwatson, the 5.2 version was showing issues with many webapps, opening a second webgl window or something, so we went to the prior working verson [01:05] ok [01:06] its symbols files mustn't have been quite right or something [01:06] cjwatson, we plan to restore it once we depend on oxide instead [01:11] robru, landing-007 is now ready for publish, when you get a free moment [01:17] bregma, there is a guy on u-t with a touchscreen question [01:24] bregma, sure! [01:25] we may have some other problems like this [01:25] bregma, published! [01:25] qml-friends qtcreator-plugin-cmake qtquick1-opensource-src qtwebkit-examples-opensource-src ubuntu-purchase-service were all built against qtwebkit 5.2, I think [01:26] robru, you are my hero [01:27] bregma, you're welcome! [01:28] confirmed [01:28] permission to no-change upload the lot of those to trusty-proposed? it freaks me out that they were built against a version of qtwebkit an upstream minor release newer than what we copied, and it may well cause problems [01:29] and given that our testing didn't catch the fact that the pyqt5 webkit module was unimportable until it got to autopkgtest, I'm not confident that the silo testing was perfect [01:29] the above is a complete list from a build log scan [01:30] oh god, and pyqt5 failed to build when I did a zero-change upload of it [01:30] presumably the patch applied to it only worked with 5.2, not 5.1 [01:31] same failure? [01:31] no, https://launchpadlibrarian.net/169428715/buildlog_ubuntu-trusty-i386.pyqt5_5.2%2Bdfsg-4ubuntu2_FAILEDTOBUILD.txt.gz [01:31] I really got to make new lp shortcuts [01:32] (I assume the same on other arches, I got failure mails for amd64/armhf/i386) [01:32] ah, but that looks very close to the same thing as before [01:32] I guess I don't know what you mean by "as before" [01:32] basically that QWebSecurity thingy probably needs to be backported too [01:33] as the earlier missing symbol [01:33] ah, the patch in ubuntu1 doesn't go near this [01:33] well, that missing symbol was indicating that pyqt5 had been built against qtwebkit 5.2, surely [01:33] yes [01:34] oh, we never built -4 against 5.0, I guess [01:35] well, I have no idea, I'll have to leave it up to you guys [01:36] cyphermox: do you think you could sort out qml-friends qtcreator-plugin-cmake qtquick1-opensource-src qtwebkit-examples-opensource-src ubuntu-purchase-service, all built against qtwebkit 5.2 and (IMO) invalidly copied to trusty-proposed as a result? hopefully they just need no-change uploads in -proposed [01:36] you're further west than I am and probably know this stuff better [01:36] sure, I'll look [01:37] but I know nothing of qt [01:37] ah, your comments above implied you did :) [01:37] ok [01:37] thanks, I'll go and crash [01:37] that said, I think I understand the pyqt5/qtwebkit failure now [01:37] cjwatson: thanks for the help [01:37] cjwatson, thanks for watching over this [01:38] np. I'm off tomorrow, hopefully others can help [01:38] pyqt5 is looking for stuff in QWebSecurityOrigin that was introduced in 5.2 and probably needs to be backported too with what Mirv took from qtwebkit 5.2 [01:38] * rsalveti back [01:38] infinity should be able to assist with archive issues (and has been fixing up override bugs associated with the Qt 5.2 landing) [01:39] cool [01:39] by my clock, infinity should be EOD 2 hours ago, so it's up to him if he's gracious enough to work after hours like us ;-) [01:40] your puny human conception of time means nothing to infinity [01:40] IME [01:40] robru: infinity sleeps weird hours, I understand [01:40] I mean, he may be EOD for all I know, but I don't try to guess :) [01:40] Generally, this would be true, except I'm running out for a date in 10m. [01:40] aha [01:40] But I did fix up a bunch of broken overrides just now, at least. [01:40] that would trump it [01:41] infinity: cjwatson: I'll do all I can to fix the things I can break. [01:41] well. I'll look in tomorrow morning before I go and start spring-cleaning [01:41] cjwatson, typically I don't memorize people's timezones, but infinity happens to be a fellow canadian and I just happen to remember he lives one timezone over from me, so it's easy for me to know that if I'm EOD 1 hour ago, his EOD was 2 hour ago ;-) [01:41] Yeah, I'll be back late tonight (or early morning, or something...) [01:41] infinity, have fun! [01:41] robru: Yeah, except that I'm around 22h per day. [01:42] (Except for today) [01:42] lol [01:42] Toodles. [01:42] of course I'm simply referring to the official 9-6 stated in your contract, not the hours you *actually* keep [01:47] cjwatson: we decided to go with qtwebkit 5.1.1 instead because of the many regressions we had with 5.2, and Mirv indeed said we could still have some packages depending on the older one [01:48] contracted hours ... what are those [01:49] my contract says 35 hours per week (hoho), it doesn't say when they have to happen :) [01:49] haha [01:49] hahah [01:49] rsalveti: ok, do you agree that we should rebuild anything that was copied to distro having been built against 5.2? [01:50] cjwatson: yes, sure [01:50] ok, cool. hopefully you can assist cyphermox with that [01:50] I've started those in sbuild already [01:50] or whatever [01:50] yup [01:50] great [01:51] how did anything land that was built against 5.2? [01:51] a fine question [01:51] * pmcgowan very confused [01:51] the runtime dependencies of the packages in question don't appear to specify 5.2 [01:51] that would be the usual guard [01:52] so either it's genuinely compatible (seems unlikely given that we saw pyqt5 autopkgtests failing), or its symbols files are broken/absent [01:52] but how did they build if not in the silo? [01:52] they DID build in the silo [01:52] the silo had 5.2 at one point [01:52] seems like a bugger discussion maybe [01:52] examples do look for 5.2 [01:52] not everything that was built against it was rebuilt after it was reverted to 5.1 [01:52] but how did they promote [01:52] because their runtime deps didn't say it was a problem ... [01:53] yup, Mirv only rebuilt a few of them [01:53] I had to figure it out by scanning build logs, which I wouldn't expect citrain to do [01:53] so they just went now, ok [01:53] ok [01:53] it definitely feels like a bug in qtwebkit's metadata [01:53] but I haven't looked into that [01:53] cyphermox: which of the qml-friends qtcreator-plugin-cmake qtquick1-opensource-src qtwebkit-examples-opensource-src ubuntu-purchase-service pyqt5 are you checking now? [01:54] started with an easy one, hopefully, qtwebkit-examples [01:54] thanks gents [01:54] I was trying to start with lower in the stack [01:54] ok, let me check qml-friends [01:54] robru: is the request for testing still valid? if so, where do I get the package from? [01:55] (partly because my mirror's running at the moment so bandwidth is awful, and partly 'cos I'm just parachuting into this) [01:55] kalikiana, oh, nevermind. rsalveti said it was good to publish. I assume he tested it. [01:55] Okay, all my override mangling should be published now, if people have builds they want to retry due to mysterious apt failures. And now really running out the door. [01:55] ah, okay [01:55] kalikiana, for future reference, if i say "silo 6", it means this PPA: https://launchpad.net/~ci-train-ppa-service/+archive/landing-006/ [01:56] kalikiana, but u1db-qt is already in -proposed ;-) [01:56] oh, cool [01:57] somebody tell didrocks in the morning that I'm fine with us forcing the packages that are uninstallable on arm64/powerpc/ppc64el due to unported deps, in this case (mainly because I've already looked into most of the porting involved, and because this landing is complicated enough already) [01:57] it'll require a bit of care in proposed-migration though [01:57] probably best to find a time I'm at a computer [01:57] ok [02:00] actually, maybe I'd better send mail [02:03] cjwatson: go sleep [02:03] robru will mail [02:03] or I will [02:04] I'm already partway through [02:04] ah [02:04] hehe [02:04] follow up if I miss anything :) [02:04] I already sent a big mail with a summary of what i did today. [02:04] ok [02:08] checking qtcreator-plugin-cmake [02:13] checking qtquick1-opensource-src too [02:20] cyphermox: did you check qtwebkit-examples-opensource-src already? [02:20] yes [02:20] it's building in sbuild as we speak [02:20] looks good so far [02:21] I just don't want to upload and see it explode [02:21] cool [02:21] checking ubuntu-purchase-service [02:21] yeah, I'm just uploading after I'm able to build locally as well [02:22] yeah [02:22] I just picked the two slowest to build I think ;) [02:22] :-) [02:23] crunching pngs as we speak [02:24] so nearly done [02:35] uploading ubuntu-purchase-service [02:35] checking qtquick1-opensource-src [02:35] ops [02:35] you got that [02:36] cyphermox: so guess we're just missing pyqt5 now [02:36] yeah, that one is still building [02:36] yup, pyqt5 is what's left [02:36] let me take a look at that one [02:36] k [02:37] pyqt5 I think is really something that needs further work in qtwebkit though [02:37] rsalveti: ^ [02:37] yeah, guess it's depending on the newer api [02:37] which we don't have [02:38] we'll land qtwebkit 5.2 once oxide is in [02:38] https://launchpadlibrarian.net/169429033/buildlog_ubuntu-trusty-amd64.pyqt5_5.2%2Bdfsg-4ubuntu2_FAILEDTOBUILD.txt.gz [02:38] I know which part is missing, it looks like some bits from QWebSecurityOrigin [02:38] but I'm not sure if there is more [02:39] right [02:39] do you know of an ETA for oxide? [02:39] maybe next week [02:40] ok [02:40] should I try to backport the qwebsecurity bits? [02:40] but let me see if this is something easy to backport [02:40] I can start that in a ppa [02:40] yeah, but guess it's harder to change webkit [02:41] so we might just disable that piece from pyqt5 [02:41] temporarily [02:41] heh [02:42] the patch that syncs webkit 5.2 into qtwebkit 5.2 is *huge* [02:42] just one patch [02:46] yeah [02:54] hm, my flo is in a really weird state, can't get to the bootloader [02:54] Mar 13 23:53:58 evasys kernel: [540140.439077] usb 3-2: Product: QHSUSB_DLOAD [02:54] Mar 13 23:53:58 evasys kernel: [540140.439081] usb 3-2: Manufacturer: Qualcomm CDMA Technologies MSM [02:54] Mar 13 23:53:58 evasys kernel: [540140.440204] usb 3-2: Qualcomm USB modem converter now attached to ttyUSB0 [02:54] and this is what I get when I connect the usb cable [02:58] just holding power for a bunch of seconds made it boot again [02:58] wonder what can be done in this mode [03:02] isn't this kind of like the old odin mode for the galaxy nexus? [03:02] something like flashing some bootloader bits [03:03] doing this it's waiting for you to flash some radio firmware or something over the tty [03:12] yeah, probably [03:31] robru: where did you send your status report? [03:31] rsalveti, oh, I sent it privately just to a few people (didier, timo, colin, and a couple others. do you want me to forward it to you? [03:31] robru: that's fine, just wanted to know if it was public so I could reply it as well [03:32] rsalveti, nah, it was too strongly worded to be public. just complaining about infra issues ;-) [03:33] haha [03:46] cyphermox: pyqt5 builds fine with http://paste.ubuntu.com/7088185/ [03:47] cyphermox: the qtwebkit 5.2 specific methods are protected by Qt_5_2_0, but that comes from qtbase [03:47] so we can't use it in this case as qtwebkit is not 5.2 [03:47] ah, cool [03:48] just need to drop this patch once qtwebkit5.2 gets published [03:48] you uploading? [03:48] cyphermox: yup [03:48] cool [04:06] robru: it s past midnight here, I'll go sleep [04:06] running out of battery by now anyway [04:06] rsalveti: thanks for the help [04:06] yeah, gone as well [04:06] np, thanks! [04:42] cyphermox, goodnight, and thanks. [05:15] yep, I just patched the Qt 5.2 away in the pyqt5's webkit part in the qt5-beta-proper PPA I mentioned to sil2100 and ricardo [05:16] let's see... [05:26] I guess my qtwebkit related rebuilds got forgotten, but it seems good now [05:28] Colin is quite a superman to jump in to patch cu2d too [05:29] reading to the end of backlog it looks pretty good, even though robru's first e-mail started with explosions and disasters :) [07:24] Mirv: hey, are you looking at proposed-migration and -excuses to see what remains as blockers for it? [07:24] (the whole transition) [07:27] didrocks: morning! looking at various things. the backlog may be worth looking, cjwatson at least had a message for you http://irclogs.ubuntu.com/2014/03/14/%23ubuntu-ci-eng.html#t01:57 [07:29] well, I guess he included most of the relevant parts in the e-mail though [07:29] qmlbox2d looks good, ken uploaded it after qtbase was already there [07:30] Mirv: yeah, seems like in the email [07:30] excellent [07:30] so, in term of crashes [07:30] I synced the qtcreator-plugin-cmake branch with ricardo's new upload, lp:qtcreator-plugin-cmake still would use the postNEW review, and another upload since I added the headers at least [07:30] we only have the unity8 crash on the list? [07:30] yeah [07:30] ah, you need another upload of qtcreator-plugin-cmake? [07:30] (blocking for the transition?) [07:31] didrocks: not blocking, just adding copyright headers to a couple of files plus if there's anything else you spot in need of updates. it can be done later too. [07:31] Mirv: yeah, let's do a silo with the sdk fix + that one [07:31] didrocks: I found in the backlog that ricardo mentioned that the only problem now is bug #1291602 that tsdgeos will look at [07:31] bug 1291602 in qtdeclarative-opensource-src (Ubuntu) "com.ubuntu.developer.andrewsomething.stackbrowser_stackbrowser_0.2.2 causes qmlscene to segfault under Qt5.2" [Undecided,Confirmed] https://launchpad.net/bugs/1291602 [07:31] once we publish that [07:31] ok, this isn't the unity8 issue though? [07:31] that Saviq mentionned yesterday [07:32] the unity8 crasher has a patch that's not included yet in qtdeclarative, saviq mentioned it's not uber critical [07:32] Mirv: it's still crashing in the tests though [07:32] yeah, I just mean whether we do it now in -proposed or after the current ones have migrated [07:32] Mirv: I would prefer that we get that fixed if possible, even temporarly, as discussed yesterday [07:32] ok [07:33] Mirv: just upload it directly to -proposed [07:33] right, doing that but doing a amd64 test build first also [07:35] ok [07:35] Mirv: I'm afraid when people start poking at buttons without understanding though :/ [07:36] Mirv: I'm going to "fix" to not have to use the ignore button btw in case of republication if I can match the same source and version [07:40] sounds good. and yes the ignore buttons have both a good and a bad side. [07:48] * didrocks will use a big red blinking button [07:55] ok approaching uploading of qtdeclarative 5.2.1-3ubuntu9, unless your big red blinking button press is something that I should wait to finish first [07:55] Friday releases \o/ [07:57] Ouch [07:57] morning sil2100 :) [07:57] Friday is the best day [07:58] Mirv: just reading robru's status e-mail and... holy shit [08:02] sil2100: I also started my morning with that and I thought "ok we probably have problems". reading further in the thread and this channel's backlogs, it turned out the brave battles fought by robru, cyphermox, rsalveti, cjwatson & co throughout the night brought a lot of good stuff done [08:05] sil2100: so while didier is pushing a big red blinking button, I'm preparing a new qtdeclarative upload to fix the unity8 crasher, and then at some point we should get a silo to land colin's UI Toolkit arch fix + new qtcreator-plugin-cmake with some copyright headers added [08:05] Mirv: indeed, phew! I guess in the end everything got done - but if I understand correctly, the u1db-qt additional merge from rsalveti got released in the end, yes? [08:05] sil2100: yes, it seems even the u1db-qt got fixed [08:06] Mirv, didrocks: hah, and I see the block still works ;) [08:07] Mirv: ok, right, a separate landing is a good idea I guess [08:07] Mirv: no no, please upload :) [08:08] ok :) [08:15] ok, that should be fixed [08:15] mind giving a quick look, Mirv, sil2100? [08:15] cihelp could you take a look why https://code.launchpad.net/~mitya57/kubuntu-packaging/qt3d-fix-ftbfs/+merge/210963 is not getting CI tested, while it should be included at http://91.189.93.70:8080/job/qtpim-opensource-src-trusty-amd64-coverage-kubuntu-packaging/search/?q=opensource-src-trusty-amd64-coverage-kubuntu-packaging ? [08:15] http://bazaar.launchpad.net/~cupstream2distro-maintainers/cupstream2distro/trunk/revision/542 [08:15] and http://bazaar.launchpad.net/~cupstream2distro-maintainers/cupstream2distro/trunk/revision/543 [08:16] now, if you publish and need to republish, citrain is smarter and won't ask you to use the ignore flag for the delta [08:16] ok, looking [08:16] (also, there is the build thing that I hot-patched yesterday which was the same source) [08:16] Looking as well [08:20] didrocks: seems to look ok this way [08:22] great :) [08:22] see, less hackish than yesterday :p [08:22] I'm a bit slower than sil2100 when it comes to reading citrain code but it looks sane :) [08:23] now with less hacks! [08:23] yeah ;) [08:23] and so, if you need to republish an already published silo [08:23] Indeed ;) But /me as a C++ user feels a bit sad to see an exception being used for 'normal' operations ;p [08:23] But besides that it's all awesome [08:23] ;D [08:24] The projects that you already published, if not available in a newer version will be ignored [08:24] sil2100: tsss :p [08:24] ok, let's keep an eye [08:25] Mirv: tell us where you are with this btw [08:25] and let's wait on tsdgeos's feedback [08:25] for the remaining issue [08:25] didrocks: the com stack bla bla one? [08:25] so qtdeclarative is building now https://launchpad.net/ubuntu/trusty/+source/qtdeclarative-opensource-src/5.2.1-3ubuntu9 [08:26] tsdgeos: that one [08:26] yeah ;) [08:26] Mirv: normally, we should be able to "merge and clean" btw [08:26] even if some packages have been uploaded directly [08:27] let me check that I'm right first :p [08:27] i've reduced the code that makes it crash *a lot* [08:27] so if i get the Qt guys to look at it [08:27] should not be "that" hard to fix [08:27] maybe even i can find the fix [08:27] hum, seems robru launch silo 001 with the wrong parameters :/ [08:27] sil2100: mind looking? ^ [08:28] didrocks: silo 001? Ok [08:28] tsdgeos: how often is that crash btw? [08:29] all the time [08:29] ok, so can't publish 5.2 without that [08:31] I think so far stackbrowser app has been the only one showing the crash, but it surely can't be the unique user of Qt in the universe [08:32] tsdgeos, happy to help, got the condensed code for me? [08:32] didrocks: ok, so it seems I have to reconfigure the silo with the ignoreconflicts option [08:32] oh man [08:32] this is silly [08:32] sil2100: hum [08:32] Oh, no, wait! [08:32] Scratch that! [08:32] sil2100: I just think that he tried to build with the wrong parameter [08:32] ;p [08:32] tvoss: https://bugreports.qt-project.org/browse/QTBUG-35891 [08:32] that's even more condensed than the one i had [08:32] ewaow [08:33] sil2100: sad that he didn't "fix" the other one :/ [08:34] * Mirv prepares also qtpim and qt3d updates in all quietness [08:34] or the latter just a multi-arch build check for mitya57 while automatic CI is somehow missing [08:35] Mirv: hum, we can do that in a followup silo, right? [08:35] with the other fixes, right? [08:35] Mirv: you can probably already configure one :) [08:35] didrocks: yep, we can add qt3d and qtpim to join ubuntu-ui-toolkit and qtcreator-plugin-cmake [08:35] didrocks: \o/ [08:35] yep ;) [08:36] Mirv: before you push qtcreator-plugin-cmake, let me do a second check [08:36] mitya57: so in case interested, the CI Train silo assigning and links to the build jobs in the CI Train Jenkins can be found here https://docs.google.com/a/canonical.com/spreadsheet/ccc?key=0AuDk72Lpx8U5dFlCc1VzeVZzWmdBZS11WERjdVc3dmc#gid=0 [08:36] sil2100: do you remember what was the second one off hand? [08:36] (in UDS it was talked not many people know about those) [08:36] https://wiki.ubuntu.com/citrain works as well [08:36] they're mainly used for lp:project:s that are upstreams, but manual uploads can be added too [08:37] so I'm (slowly) adding line 45 for these now [08:38] didrocks: trying to remember, but hmm... my memory is weary ;/ [08:38] didrocks, Mirv, what crashes in tests? [08:38] Saviq: this one is fixed apparently, there is just the one remaining on tsdgeos's plate AFAIK now [08:40] sil2100: mind looking back? :p [08:40] didrocks, /me will verify, just add proposed to device? [08:40] Saviq: let's ensure Mirv uploaded qtdeclarative first and it's built, but normally yeah [08:40] didrocks, k [08:41] Mirv: qtcreator-plugin-cmake should be multiarch: same [08:41] missing trailing comma as a binus :p [08:41] Saviq: the qtdeclarative has the unity8 patch we talked about, https://launchpad.net/ubuntu/+source/qtdeclarative-opensource-src/5.2.1-3ubuntu9 [08:41] * didrocks likes when there is "new" in a package description [08:41] hey [08:41] Mirv: saved to my gdrive, thanks. Also thanks for the "multi-arch build check". [08:42] Mirv, k [08:42] i removed my landing asks for some webapp fixes (were desktop/phone) [08:42] Saviq: but ubuntu8 is already there without it [08:42] dbarth: hum, please don't remove lines directly, just ping us [08:42] but i'd like to move forward with line 40 which is only for desktop [08:42] dbarth: if it was assigned [08:42] well, it was blocked and was not assigned (no silo, nothing) [08:42] ah ok, then ;) [08:43] could i have a silo to prep. things a bit at least? [08:43] line 40 is desktop only [08:43] and it's /fixes/ ie does not require an ffe [08:43] dbarth: yeah, but the components are used by the Qt 5.2 landing right now [08:43] dbarth: that we are having [08:43] so after that being one, it will be fine [08:44] ugh, ok then; i've tried ;) [08:44] dbarth: you can remove webbrowser-app from the ask [08:44] dbarth: I think the rest isn't locked [08:44] ah [08:46] didrocks: thanks, lp:qtcreator-plugin-cmake updated [08:47] Mirv: the flags thing can use a rework [08:47] with DEB_CFLAGS_MAINT_APPEND and so on… [08:47] didrocks: i cleared the webbrowser-app bits, let me know if that can work [08:48] sil2100: dealing with this? ^ (line 40) [08:49] didrocks: ok, just finished e-mailing - on it! [08:49] Mirv: seems as well that upstream install is hackish, but would great to have that done, but no worry for now [08:49] sil2100: and the NEW package I have to check? :p [08:49] Mirv: bonus point if you bump standards-version [08:50] Mirv: and you need to fix E: qtcreator-plugin-cmake source: not-binnmuable-all-depends-any qtcreator-plugin-cmake-dev -> qtcreator-plugin-cmake [08:50] Mirv: also, no shlib:Depends in qtcreator-plugin-cmake which is needed [08:50] Mirv: that's it I guess :) [08:51] Mirv: please, look at the lintian errors, it's helping a lot to detect those ;) [08:52] sil2100: appmenu-qt5 was the second one [08:52] sil2100: also, will need the same for the signon cmakerie, seems robru only fixed one on the 2, mind looking? [08:52] didrocks: I don't see it in the NEW queue though [08:52] didrocks: will look [08:53] sil2100: it was in the NEW queue, I asked you yesterday to remind me to give them a look again [08:53] dbarth: it seems unity-webapps-qml is part of landing 006 as well - you want me to configure this silo only for testing for you (if you would want to publish, you'd need to rebuild after the qt5.2 landing is done) [08:54] I guess better for now to ask -qml to be removed as well [08:54] dbarth: you can remove the -qml package as didrocks mentions, then there's no problem in publishing [08:54] dbarth: not sure if it's required for this fix? [08:55] sil2100: copyright is not you, but canonical [08:56] please fix it [08:56] sil2100: also, better to not differentiate the copyright for the packaging [08:56] and in addition, it's wrong :/ [08:56] License: GPL-3 [08:56] On Debian systems, the complete text of the GNU General [08:56] Public License version 2 can be found in "/usr/share/common-licenses/GPL-2". [08:56] the whole stenza is GPL2 [08:56] didrocks: crap, I wonder how that appeared there, fixing that now [08:58] didrocks: so, I made the -dev package now Arch: any too because it depends on an any package [08:58] Mirv: yeah [08:59] didrocks: the missing dependency on libc isn't avoided still, though [08:59] Mirv: it's with shlib-deps? [08:59] didrocks: yes, but it's similar to qtcreator-plugin-ubuntu, and the same reason why override_dh_shlibdeps is needed [09:00] sil2100: can you make your package future-proof and add --fail-missing [09:00] qtcreator-dev is a quite hacky for a dev package [09:00] Mirv: the -dev shouldn't dep on libc [09:00] the other should [09:00] right? [09:01] didrocks: done, can I direct-push to trunk? [09:01] sil2100: can't you have a landing with the fixes? [09:01] sil2100: as we do for the rest [09:01] Ok [09:01] we'll have a followup landing anyway [09:01] so better to have everything done [09:02] sil2100: the rest is good [09:03] didrocks: thanks! I'm fixing libaccounts-qt for the cmake thing in the meantime [09:03] ok [09:03] didrocks: I added the shlibs:Depends in the non-dev package indeed, but it doesn't bring the libc dependency with it to the plugin packages that are compiled against QtC [09:04] I think earlier it was thought that because of the plugin nature the error is "ok" [09:04] Mirv: why do you have to strip libCMakeProjectManager.so from dh_shlibdeps? [09:04] I don't remember the rationale and there is no comment :p [09:06] didrocks: without it it gives dpkg-shlibdeps: error: no dependency information found for /usr/lib/x86_64-linux-gnu/qtcreator/libUtils.so.1 (used by debian/qtcreator-plugin-cmake/usr/lib/x86_64-linux-gnu/qtcreator/plugins/QtProject/libCMakeProjectManager.so) [09:07] Mirv: ok, and why it can't find those infos? [09:07] because qtcreator doesn't have a shlibs I guess? [09:07] I remember only that upstream doesn't really support plugins from external sources, and this qtcreator-dev is just devised by us [09:07] so they are not really "plugins" [09:08] I'm not sure how to properly format a comment to debian/rules, though [09:09] Mirv: ok, try to put a comment and add the deps manually in debian/control on libc [09:09] I would say [09:10] ok [09:11] * Mirv tried to prepare a silo but merge & clean not yet run on landing-006! :) [09:11] mardy: hi! [09:11] Mirv: I now see that qt3d built in your ppa, but it has only 3 archs where it already worked. [09:12] mardy: regarding your comment on the lp:signon -common-dev package merge, does the same thing go for libaccounts-qt ? [09:12] Mirv: yeah, just use ignore conflict for it [09:12] sil2100: hi! Yes indeed [09:13] sil2100: please put your stuff in the same silo than Mirv ^ [09:13] Mirv: Can I go ahead and upload it to archive? [09:13] didrocks: ACK [09:13] didrocks: the signon and libaccounts-qt fixes as well you mean? Or can those wait? [09:13] * sil2100 adds appmenu there for now [09:14] Mirv: I sneaked my appmenu packaging merge to your landing :> [09:15] Mirv: did you include it already when you were assigning the silo, or is a reconfigure needed? [09:15] mitya57: I now got a silo https://launchpad.net/~ci-train-ppa-service/+archive/landing-005/+packages assigned, let's build it there and we get arm64 + powerpc + ppc64el builds too [09:15] I copy the sources there now [09:16] ok, now they're starting to build for all archs [09:16] Thanks! [09:17] mitya57: so then when they've built, I can hit the "Publish" button in landing-005 to get them copied to archives. alternatively, if you want to go ahead, feel free to tell me you uploaded it manually (after the builds have finished) and I can remove it from that landing silo [09:17] sil2100: ah, ok! reconfigure needed, doing that [09:18] Mirv: Please hit the publish button yourself :) [09:18] sil2100: those can wait [09:18] Mirv: thank you! :) [09:18] mitya57: ok, thanks for the fix btw, it'll enable a lot of arm64/ppc stuff :) [09:19] didrocks: ok, I'd upload qtcreator-plugin-cmake to the silo too now after adding a comment and a direct dep to libc6 [09:22] Mirv: oh, um, actually, we never got round to enabling the other silos for arm64 [09:22] I should do that ... [09:23] Mirv: did you fix the flags? [09:23] cjwatson: hey, thanks for the fix! [09:24] cjwatson: actually, the integration tests needs root (I should fix that), if you just nosetests tests/unit, they don't [09:24] cjwatson: ah, ok, well powerpc and ppc64el at least. if you do enable, I can do the copy-package. [09:24] cjwatson: and you have a good memory to remember about the [0] trick you advise me (with all you have to deal everything, I thought you would forget) ;) [09:25] didrocks: yeah, I decided I didn't want to run any tests that required root :) thanks anyway [09:25] didrocks: I reconfigured the silo with ignore conflicts and added sil2100's branch [09:25] didrocks: I had actually forgotten about it and only remembered due to the comment you left ... [09:25] Mirv: powerpc and ppc64el were enabled for all the landing PPAs, just not arm64 [09:25] yep [09:26] Mirv: awesome o/ [09:26] cjwatson: ahah, so the comment was useful :) [09:26] anyway, nice that it's explicit now [09:26] Mirv: great, got anything with a sane flag handling? [09:27] mitya57: doh, qt3d fails tests now on powerpc, I wonder if I should do the same I did for qtbase, ie. enable tests only on amd64+i386+armhf. https://launchpadlibrarian.net/169469342/buildlog_ubuntu-trusty-powerpc.qt3d-opensource-src_5.0~git20130731-0ubuntu3_FAILEDTOBUILD.txt.gz [09:27] * sil2100 needs to dress up [09:27] I'm just seeing if I can find a webop to switch on arm64 everywhere [09:27] Mirv: our Qt friends to the rescue https://codereview.qt-project.org/#change,80934 [09:28] didrocks: Saviq: tvoss: ↑↑↑ [09:28] 76 upgraded, 81 newly installed, 108 to remove and 0 not upgraded. [09:28] not good [09:28] didrocks: sorry, I'm a bit lost with what you mean with flags (the changes made to build/publish?), but prepare-silo went good http://162.213.34.102/job/prepare-silo/497/console while build complains about the version mismatch http://162.213.34.102/job/landing-005-1-build/62/console [09:28] tsdgeos: do you smell any bad imapct to the fact it's not acessible? [09:28] didrocks: it wasn't anyway [09:29] didrocks: so basically the app we have is not accessing them [09:29] 09:47:17 didrocks | Mirv: the flags thing can use a rework [09:29] tsdgeos: ok, sounds great! [09:29] 09:47:44 didrocks | with DEB_CFLAGS_MAINT_APPEND and so on… [09:29] tsdgeos, that looks good [09:29] tsdgeos: great then, Mirv distro-patch it? ^ [09:30] didrocks: I clearly haven't had enough coffee, it seems I skipped to the "upstream install is hackish ... but no worry for now". I don't see grep -r APPEND * [09:30] oh, meeting [09:31] Mirv: yeah ;) [09:31] ogra_: around? [09:31] popey: ^ [09:32] there's no vanguard in #webops (Canonical-internal), and I have to go; if somebody who happens to be there could follow up on any questions they might ask me, that'd be great [09:33] didrocks, still fighting G+ ... [09:33] tsdgeos: "yes!" to the distro patch [09:34] :) [09:34] * ogra_ cant connect :( [09:35] sigh ... with either browser [09:36] i see myself and a "please wait..." at the bottom ... not even controls [09:36] cjwatson: will try to do that, thanks again Colin! [09:36] and enjoy your week-end! [09:36] hmm, now i have controls ... but that still doesnt feel right ... [09:37] Mirv: how about just disabling particular failing test(s)? [09:39] didrocks: thanks. fwiw (if I can't be found), my plan if it's needed to override any uninstallables on arm64/powerpc/ppc64el was to add entries to FauxPackages in the britney1-ubuntu branch [09:39] (that's not complete instructions but should be enough for somebody to track things down if need be) [09:41] dbarth: tell me once you're around [09:43] Mirv: in any case, please feel free to upload anything. [09:46] cjwatson: hum, ok, I'll try to find someone whos know how to do that [09:49] mitya57: ok. yes, disabling individual tests sounds good if it's not a huge effort (in case of qtbase it was https://code.launchpad.net/~chris.gagnon/kubuntu-packaging/enable-unit-tests-qtbase-opensource-src/+merge/200535) [09:50] interestingly, ppc64el succeded [09:50] I think it was the same with qtmultimedia [09:52] What the heck [09:52] My PC simply turned off [09:53] I think it overheated but hm... === sil2100_ is now known as sil2100 [09:54] sil2100: can you check http://162.213.34.102/job/landing-005-1-build/62/console too? [09:55] Mirv: looking [09:55] ok, we need to wait for the merge&lean [09:55] sil2100: so no prob, we'll build it a bit later [09:56] Mirv: it seems that it's detecting the version that's in -proposed, so indeed, it seems to be all transient [09:59] sil2100: well, it's to protect you :p [10:01] didrocks: you mean, my thinkpad heard the word 'qt5.2 transition' and decided 'duuude, bail! *shuts down*' ;) [10:01] sil2100: no, I mean, on the error on landing-005 [10:02] Ahaha [10:02] Indeed! [10:04] psivaa: do you think our smoketesting mako devices are free now for some additional tests of that unity8 AP failure? [10:04] sil2100: i could find one of them [10:06] Ran 41 tests in 1037.561s → OK [10:06] land it! [10:06] boring ... [10:08] psivaa: could you do the same thing as yesterday? i.e. I'll send you a modified test_url_dispatcher.py and I would ask you to replace it and run unity8 test suite - would be awesome [10:08] sil2100: ack, will do [10:08] Morning all [10:10] psivaa: here's the test_url_dispatcher.py : http://paste.ubuntu.com/7089282/ [10:10] sil2100: ack, thanks [10:11] psivaa: thank you! Give me a sign once there are some results [10:11] sil2100: sure, i need to do some prep's to only make this run on a single device from a pool. so will be in a bit :) [10:17] seb128: hey, can you shed some light on this maybe? [10:17] seb128: http://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses.html [10:17] look for qml-box2d [10:18] it says "out of date on i386: qtdeclarative5-box2d1.0 (from 0.1~git20131115ubuntu1) [10:18] " [10:18] for instance [10:18] but the new version built on all arch [10:18] * didrocks is a little bit lost [10:18] didrocks, looking [10:20] didrocks, oh, easy [10:20] ah? [10:21] there was a binary rename [10:21] qtdeclarative5-box2d1.0 is the old binary [10:21] qtdeclarative5-box2d1.1 the new one [10:21] ok, so this won't block britney, transitioned are handled, right? [10:22] right, the error is still weird [10:22] (it seems to be in the "block" part :p) [10:22] yep [10:22] seems there is no rdepends anyway [10:24] didrocks, I think it might be because https://launchpad.net/ubuntu/+source/qml-box2d/0.1~git20131115ubuntu1 never moved out of proposed [10:24] and then they went away in a new upload [10:24] I'm unsure proposed handle superseeded binaries [10:24] you might need to delete qtdeclarative5-box2d1.0 binaries from trusty-proposed [10:25] yeah [10:27] thanks seb128 for looking at that :) [10:27] dbarth: I still see unity-webapps-qml in that landing - I will not be able to *normally* assign a silo for that while this merge is still there, you think we can have that landing without that component? [10:27] didrocks, yw! [10:28] so, we have the list of what to remove safely [10:29] cjwatson: FYI, we built the list of what we need to ignore. I'll put it in a bug report. There is no release team or people knowing about britney1-ubuntu around. So with that list (and having checked all reverse dependencies to not make uninstallable packages), I'll go ahead and remove binaries. Then, we'll land your fix and rebuild what we can [10:29] didrocks, what's the plan? moving it out of proposed today? ;-) [10:31] seb128: yeah, even ASAP now [10:32] great! [10:32] cjwatson: hi! is this part of some CI Train silo? https://code.launchpad.net/~cjwatson/indicator-network/porting/+merge/210394 [10:32] or can I just include it to my set of MP's? [10:33] Wellark, the CI train table is public, you can check yourself, and cjwatson is on vac today [10:34] seb128: just wanted to politely ask him as the original submitter before I make my moves :) [10:35] Wellark, I'm sure he's going to be fine if you land it [10:38] ok, next britney runs should try to migrate everything [10:38] (phew) [10:38] 117 blocks -> removed [10:38] Wellark: it seems to be part of landing on line 42 [10:38] didrocks: \o/ [10:38] sil2100: yeah, I noticed [10:39] I can figure it out with thostr_ [10:39] it's just a oneliner [10:45] address-book-app/powerpc unsatisfiable Depends: qtcontact5-galera [10:45] Mirv: I still need to remove them (but in proposed) [10:45] as they built [10:45] ok, so from proposed [10:49] ok, removed [10:49] let's wait for next britney's run [10:50] didrocks: http://pastebin.ubuntu.com/7089448/ is this what you meant? [10:50] Mirv: yeah, I guess you need to do something for CFLAGS and CXXFLAGS as well [10:50] (if those hacks are needed) [10:51] didrocks: well, it seems they are not and I guess that magic does the same as using dpkg-buildflags --get to get CPPFLAGS + CXXFLAGS [10:52] Mirv: yep [10:52] proceeding [10:52] it really took some while to me to stop staring at the upstream install mangling lines and find these ones you meant [10:54] didrocks: +1 for 237 unless popey finds something horrific [10:55] davmor2: didrocks nearly done. [10:55] davmor2: popey \o-- (just one hand until popey is done :)) [10:56] didrocks, davmor2: I saw Selene send an e-mail with some issues with 237 though? [10:56] i cant reproduce any of selene's issues [10:57] sil2100: nor could I [10:57] phew [10:58] didrocks: ok, i think it's better to wait with the silo as-is, and land once the deck is cleared [10:59] didrocks: ok [10:59] I mean [10:59] dbarth: ok [11:00] didrocks: wow sil2100 must of reading your mind when you were thinking sil2100 buys all the drinks at the next sprint ;) sil2100 that is why you said ok to didrocks right? [11:01] I... I... I didn't know I have mind reading skills! [11:04] thanks sil2100, that's very kind [11:05] +1 :) [11:05] * sil2100 sighs [11:05] Oh well, I can't help my awesome skills! [11:05] * didrocks can't wait for the publisher run [11:06] sil2100: http://q-jenkins.ubuntu-ci:8080/job/psivaa-trusty-touch-mako-smoke-unity8/29/ has not seen the url dispatcher failure [11:06] YAY [11:06] So now the big question is... WTF! [11:06] psivaa: thanks! [11:06] sil2100: yw. :) [11:19] are we there, yet? are we there, yet? are we there, yet? [11:19] * sil2100 starts sweating [11:20] didrocks: address-book-app-dbg/powerpc [11:22] Mirv: I'm doing signon-plugin-oauth2/arm64 for now [11:22] let me look at the other [11:22] right [11:22] and telepathy-ofono/ppc64el unsatisfiable [11:25] Mirv: telepathy-ofono -> done [11:26] didrocks: ya, #237 seems okay [11:26] popey: thanks! ogra_: promoting? ^ [11:26] Mirv: address-book-app-dbg/powerpc -> flushed [11:27] didrocks, yup, will do [11:27] Mirv: ok, seems that's all for our transition [11:28] so, we "just" wait on qtdeclarative-opensource-src to finish to have the autopkg test run + the armhf build [11:28] didrocks: it's all in \o/ ? [11:28] sil2100: hum? [11:28] sil2100: look at proposed-migration [11:29] the "just" is what is blocking [11:29] in theory [11:29] Mirv: I think I have to remove tdeclarative5-box2d1.0 from -proposed [11:29] due to ken's upload [11:29] as it's 1.1 now [11:30] armhf qtdeclarative is still compressing png files. someone should kill optipng with fire (it's even quite poor compared to competitors even though it's slow) [11:30] ok to qtbox1.0 [11:31] It seems to be close to finishing soon anyway [11:31] right, right, there's 1.0 in proposed without source as well as 1.1 [11:32] * didrocks flushes [11:32] so no more "out of date" apart from qtdeclarative-opensource-src, no jenkins autopkgtest running apart from qtdeclarative-opensource-src and not unsatisfiable Depends [11:33] davmor2: can you update the nexus status sheet pls? [11:33] \o/ [11:33] let's wait for qtdeclarative to finish + a publisher tick [11:33] then, click-store-side needs to be dealt with [11:34] all autopkgtests passed for qtdeclarative-opensource-src [11:35] The armhf build is still ongoing though... [11:35] yep [11:35] it's the theorical last blocker [11:36] what if optipng fails! no, really, the build has succeeded already, the optipng just drags it on [11:37] I'm so familiar at staring those png file logs [11:37] no, optipgn passed! :) [11:37] building packages… [11:37] Mirv: so, are you taking care of qt3d or should I upload it with tests disabled? [11:37] there would be export NO_PNG_PKG_MANGLE=1 to use [11:37] Phew [11:37] mitya57: I'm taking care of it, we just let this migration first pass and then I'll publish the silo [11:38] mitya57: like, of course, after disabling or fixing the tests, right [11:39] yeah, please, don't delay it more with another upload :) [11:39] === Image 237 Promoted === [11:40] mitya57: I just kicked the arm64 builds going in the PPA. if it turns out it's only powerpc failing, I'll probably disable the tests on powerpc only [11:40] (and so it actually seems) [11:41] OK [11:42] ogra_: yaay [11:45] ok, this publisher cycle should have it (starting now) [11:50] popey: will do [11:50] http://i.imgur.com/d2kslI4.jpg [11:54] hahaha [11:56] hmm.. thostr_ is on holiday or something [11:56] is there anyone willing to lend me a silo? [11:57] you would get it back on Monday [12:00] sil2100, silo for 45 pls? [12:04] didrocks: is signon-plugin-oauth2-dev/arm64 unsatisfiable Depends: signon-plugin-oauth2 + signon-plugin-oauth2-tests/arm64 still something in need? [12:04] Mirv: yeah, need to be removed as well [12:06] Mirv: done [12:07] didrocks, do you need help debugging the remaining issues? [12:07] like phonon seems to be an issue on arm64/powerpc/ppc64el and there are some issues on the main archs as well [12:08] seb128: yeah, I'll appreciate === josepht changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: josepht | CITrain support - US: robru, cyphermox, rsalveti - EU: sil2100, Mirv, didrocks | CITrain support no answer: use mup bot after 30 minutes, but choose right timezone [12:10] seb128: where do you see an issue with phonon? [12:10] didrocks, bottom of http://people.canonical.com/~ubuntu-archive/proposed-migration/update_output.txt [12:10] * ppc64el: phonon4qt5-backend-gstreamer, phonon4qt5-backend-vlc [12:10] ah, I'm still in -excuses :p [12:10] ok, so gstreamer [12:11] seb128: it's more than that actually, right? [12:11] like on i386? [12:11] didrocks, right, as said that's an obvious one, then " there are some issues on the main archs as well" [12:12] didrocks, I'm starting a pbuilder looking at that one [12:12] sorry, I read main pocket [12:12] like release [12:12] oh, ok [12:13] seb128: phonon-backend are stuff which were blocked in proposed [12:13] that's what xnox rebuilt [12:13] so, it should be a publisher cycle away [12:13] it seems we always have stuff that are a publisher cycle away from the goal :-) [12:14] is that like we do half the remaining distance every cycle but never reach the goal? [12:15] mhr3_: ok [12:17] mhr3_: soo, we still didn't clean the 006 silo which has mediascanner2 in it - maybe you want to wait a bit so that we can assign a normal silo for this? [12:17] sil2100, ah right, qml plugin.. forgot about that [12:18] sil2100, yea, let's wait for 006 [12:18] mhr3_: or you want to have a test-build with ignore-conflicts? I would prefer waiting a bit first if it's not urgent ;) [12:18] please don't make ignore-conflicts the rule :/ [12:18] * didrocks warned… [12:18] seems you give an option, then the override is always used/proposed [12:18] * didrocks sigh [12:20] didrocks: I always don't recommend it, but give it as a possibility..! [12:20] But ok, indeed it's best not to do that even [12:21] sil2100: please don't propose it :p [12:21] especially as the thing is going to get in soon [12:21] didrocks, he knows that i can make a good decision about when to use it ;) [12:22] mhr3_: you? man! he's sooooo wrong :p [12:22] didrocks, auch! so mean! [12:22] morning [12:22] how much carnage do we have [12:23] pmcgowan: we've been almost there for 3 hours :) [12:23] it's just complex [12:23] it's the upper level magicians didier, dimitri now mostly at work [12:24] ah well then [12:24] while also the next landings are being prepared [12:26] sil2100, so... 46 instead? :) [12:27] ;) [12:27] mhr3_: ok, that's more like it I guess! [12:27] But there are no test plans! OH NOES [12:28] catching that up [12:30] sil2100, filled in [12:31] Assigning! [12:32] Mirv: meeting is in 30 minutes or 1h30? [12:32] josepht: did you happen to see the c_ihelp call for qt3d merge requests handling? I now merged manually one, and there's another one at https://code.launchpad.net/~timo-jyrinki/kubuntu-packaging/qt3d_disable_tests_on_powerpc/+merge/211042 - there used to be jenkins passing a month ago or so [12:32] didrocks: 30min [12:32] ok [12:32] mhr3_: hah... you're unlucky today! [12:32] as we are 1.5 cycle away of nirvana… [12:33] sil2100, noooo [12:33] will be short :) [12:33] mhr3_: take that! :p [12:33] didrocks, you enjoy my suffering, don't you? :P [12:33] mhr3_: indicator-network is locked! [12:33] mhr3_: well, I see you enjoying mine! [12:33] Mirv: looking [12:33] didrocks, touche :D [12:34] ;) [12:34] sil2100, also 5.2? [12:34] yep 5.2 [12:34] josepht: all I know is that it's still seen at http://91.189.93.70:8080/search/?q=opensource-src-trusty-amd64-coverage-kubuntu-packaging , so I'd assume some sort of automatic functions to be happening [12:34] damn, is there still someplace where we don't use qt? :) [12:35] hmm, maybe that's different place nowadays, qtdeclarative did have jenkins approve an hour ago, but it's not visible there at least [12:35] josepht: Chris Gagnon probably knows the most about this === alan_g is now known as alan_g|lunch [12:35] I'm not sure if it's on CI team's machines at all or not [12:35] Mirv: the latest build for that job was manually run by alesage [12:37] Mirv: I'll create an asana task and follow up with Chris and/or Allan when they come online [12:37] josepht: ok, thanks a lot! [12:38] Mirv, will you send out a mail when 5.2 finally lands? [12:38] would be nice :) [12:39] mhr3_: yes, 5.2 ;) [12:40] mhr3_: well did_rocks could take the honor by announcing it at the evening's e-mail, I don't mind. although, I believe it's worth pointing this out also on ubuntu-devel so I can respond to the earlier thread there or something like that [12:41] mitya57: ok qt3d now built fine for all 6 archs, with tests disabled on powerpc. so it's ready in the silo and will get published... at some point today [12:46] Mirv: it built everywhere, mind if I publish it myself? [12:47] mitya57: no, please don't yet [12:47] OK [12:47] mitya57: let's get 5.2 in first, it will delay again [12:48] * mitya57 just wants to get most of FTBFSes resolved before tomorrow [12:48] mitya57: yeah, but by that, you will make a lot of people (20+) blocked and delayed by another 1.5h :p [12:49] (not telling about appmenu-qt5 which already transitioned and so broke in trusty release pocket) [12:52] It migrated! \o/ [12:53] * mitya57 just got a ton of bug closing mails. [12:54] wow [12:54] Can I go ahead with qt3d now? [12:55] mitya57: some are still stuck [12:55] mitya57: better to wait to ensure that everything is in [12:55] than creating horrors :) [12:55] Even pyqt5 migrated, which depends on everything else :) [12:56] mitya57: yeah, but I would better see the confirmation of everything's being fine ;) [12:56] like http://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses.html [12:56] not listing then in the next publisher cycle [12:56] so that we can double check [12:56] Mirv: do you still have the list of all packages you had yesterday? [12:56] so that we can grep against that one [12:57] or using launchpadlib api [12:57] didrocks: http://pastebin.ubuntu.com/7089993/ that's of course only the 117 [12:58] Mirv: I guess that will be enough :) [13:00] Mirv: I'll do for p in `cat package_list`; do grep $p excuse; done [13:01] Mirv, didrocks will you have time for the standup [13:01] morning [13:01] * rsalveti reading backlog [13:01] pmcgowan: yeah [13:07] elopio, can you join the standup? [13:09] Mirv: can I close bug 1278329 now? [13:09] bug 1278329 in qtbase-opensource-src (Ubuntu) "[FFe] Qt 5.2" [Undecided,Confirmed] https://launchpad.net/bugs/1278329 [13:09] didrocks: Please ping me when it will be OK to go ahead. [13:10] mitya57: will do [13:10] I expect I'll need to do some symbols-update uploads after qt3d lands. [13:22] mitya57: sure, go ahead and close it! [13:22] update_excuses returns no results when grepping for the list of 117 - does it mean it's DONE? [13:24] mitya57: reading #ubuntu-release, we need at least one more britney run [13:24] regarding qt3d [13:25] sil2100: yes, for that page. however, there's a new page to stare http://people.canonical.com/~ubuntu-archive/nbs.html :) [13:25] that means too much stuff to copy, it will continue in the next run [13:26] I see [13:27] hah, this I didn't know [13:27] Mirv: let me stare then as well ;) [13:30] mitya57: ping, you can go! [13:30] :) [13:31] mitya57: running merge & clean? === alan_g|lunch is now known as alan_g [13:34] didrocks: are you asking me about m & c? :) [13:35] mitya57: just grab https://code.launchpad.net/~timo-jyrinki/kubuntu-packaging/qt3d_disable_tests_on_powerpc/+merge/211042 and continue from there [13:35] mitya57: you'll be quicker that way indeed than waiting for the silo to actually work [13:35] argh [13:35] yeah, sorry [13:35] since we need that merge & clean phase done etc, and then there are two builds [13:35] didrocks: ok running normal m & c [13:37] didrocks: http://162.213.34.102/job/landing-006-3-merge-clean/19/console [13:37] jdstrand: hey, around? [13:38] Mirv: yeah, you can ignore step [13:38] as publication wasn't the last one [13:38] didrocks, Mirv: I'm just going to copy a package from landing-005 to primary archive [13:38] didrocks: http://162.213.34.102/job/landing-006-3-merge-clean/20/console [13:38] mitya57: sure, fine [13:38] Mirv: oh, don't copy, let the system do it for you :) [13:38] Mirv: no? ^ [13:38] (done) [13:39] didrocks: it's quicker that way, mitya57 can get to other fixes while the silo is not ready [13:39] ok [13:39] didrocks: eg. the MP:s haven't built yet there etc [13:39] Mirv: ok, so excepted issue [13:39] on the packages that were manually rebuild [13:39] there is an optoin [13:39] option [13:39] mitya57: thanks! I'll remove qt3d from that landing [13:39] Mirv: so, ignore dest version [13:40] didrocks: so it seems [13:43] didrocks: next up http://162.213.34.102/job/landing-006-3-merge-clean/21/console [13:43] Mirv: line 47 on CI Train has MR that fixes notes-app AP failures on Qt5.2 if we desire to pick that up [13:43] bfiller: great! I could grab them to my line 44 which already has a silo and misc post-5.2 fixes? [13:44] Mirv: sure [13:44] bfiller: ok, thanks [13:45] bfiller: oh, right, it's a click package too? so it should go through the click process more probably. [13:45] sergiusens: ^ notes-app update https://code.launchpad.net/~elopio/notes-app/dont_click_header/+merge/210349 [13:46] Mirv: yes it;s a click, but we do also want to publish the deb [13:47] bfiller: aha, ok, then I can include it anyhow too for the deb [13:48] bfiller: we don't want to wait on the deb to get the click update though [13:48] didrocks: correct [13:49] so I guess it's fine to get it merged as part of the click release process [13:52] bfiller: ok so not including in my landing after all, so that it can get merged via the click release process as mentioned. also, the other packages in that other landing are not closely related. [13:53] Mirv: ack [14:30] popey: Mirv: validating? http://paste.ubuntu.com/7090461/ [14:30] I'm awake, I'm awake... [14:31] +1 [14:32] I'm not familiar with the framework names but looks good [14:32] elopio: :) I see that, you're committing! [14:32] elopio: so, landing the workaround branch https://code.launchpad.net/~elopio/messaging-app/workaround_autopilot_5.2/+merge/211018 would be enough at first? [14:32] be careful with the status of silo 7, the "track migration to release" is confused about some packages that got manual uploads [14:32] Mirv: yes, I stayed late trying to understand the messaging problem. [14:32] to the versions it's trying to watch are already superseeded [14:32] e.g qtbox from kenvandine [14:32] Mirv: it has a prerequisite branch [14:33] Mirv: https://code.launchpad.net/~elopio/messaging-app/delete_multiple/+merge/210983 [14:33] I made a small mistake. Now tests are running again to confirm it's correct. [14:34] bfiller: can you help me getting reviews for the two branches? https://code.launchpad.net/~elopio/messaging-app [14:34] elopio: usre [14:36] didrocks: well, the effect of you removing binaries is that if we don't get things ported then something will fail to migrate later and nobody will remember why ... but I guess you decided you were fine with that [14:36] Wellark: not part of any landing I know of, feel free to include it, thanks [14:37] cjwatson: I have the whole list and tracking it [14:37] re qml-box2d, sorry about that, I actually ran the removal command last night but forgot to say yes at the prompt :) [14:37] * cjwatson processes the obvious bits from NBS [14:42] bfiller, Mirv: and https://code.launchpad.net/~elopio/notes-app/skip_bug1288876/+merge/211062 [14:43] skipping a test goes against all the values of the jedi order, so don't tell my teammates ;) [14:44] elopio: I thought this other MR seemed to solved that notes problem: https://code.launchpad.net/~elopio/notes-app/dont_click_header/+merge/210349 [14:44] elopio: bfiller: you'll need to ask for the line 48 silo reconfiguration later on [14:44] after all the branches are listed there for real [14:45] bfiller: that other one isolates the problem so it's exposed only on one test. [14:46] elopio: ok so we need both of those MR's then to fully solve it? [14:46] cjwatson: ok, I will take care of iit [14:48] this seems like a strange failure, any ideas? [14:48] https://jenkins.qa.ubuntu.com/job/generic-mediumtests-runner-mako/5823/console [14:49] bfiller, do you already have an MR bumping the framework versions for the apps or should I get one? [14:50] elopio: did the messaging-app retest finish, so those two branches are ok? [14:50] sergiusens: I don't have one [14:50] sergiusens: ok, so on the deb side, we have messaging-app and we are done [14:51] sergiusens: tell us when gallery-app, terminal-app, music-app and notes-app are in the store [14:51] bfiller, ok I'm updating gallery and note to use ubuntu-sdk-14.04-dev [14:51] sergiusens: tell me too ☻ [14:51] sergiusens: camera too? [14:51] see, I was told camera didn't need to be rebuilt [14:51] didrocks, ? [14:52] I have no idea honestly [14:52] neither do I [14:52] I wasn't following Qt 5.2 transition closely; just following orders as they came [14:52] sergiusens: I don't know, pmcgowan should knows. I just came to rescue the list :p [14:52] sergiusens: same, just there for the past 2 days… [14:52] camera is a click app now, right? [14:52] didrocks, yes [14:53] didrocks, but does it need a rebuild? [14:53] didrocks, or is the consensus; anything with compiled bits needs a rebuild? [14:53] sergiusens: well, it will be a larger subset of what really needs to be rebuilt [14:53] sergiusens: but I guess that the recent transition can have made it slept [14:53] sergiusens: so please, to ensure, just rebuild [14:54] ok [14:54] didrocks: can I still install qt5.2.1 from the -006 ppa? if so I can install and double check apps that need rebuilding. i'm assuming their'll be no images till the qt5.2.1 one right? [14:54] didrocks, everything is in the archive now; right? [14:55] cjwatson, is click chroot already supporting the new framework or should I just fiddle with maint? [14:55] sergiusens: everything but messaging-app [14:55] no sil? [14:56] mhr3_: mup him [14:56] The fuck was that disconnec... [14:56] summoning is complete :) [14:56] t [14:56] Mirv: I don't see the new results on the MP [14:56] bfiller: we need those two MRs, yes. But that's not a fix, it's a skip :) [14:56] sil2100, so, can i start the previously-frozen silos? [14:57] sil2100, and get the ones i asked for? [14:57] mhr3_: was I summoned for some assignments? Thank you, it seems I got my internet back thanks to you [14:57] SUre === bfiller is now known as bfiller_afk [14:59] mhr3_: so, firts, let me assign those silos, and then let's rebuild 001 [15:00] sergiusens, whats the question? [15:00] elopio: ah, ok, on the MP. well, anyhow, I can build the messaging-app meanwhile, and rebuild if the branches need mor echanges [15:00] hmmm [15:00] pmcgowan, list of apps that need rebuilding or if anything with compiled bits need to be rebuilt [15:01] mhr3_: ok, so you still have to wait a liitle bit, as the silo's PPA is still getting cleaned [15:01] sergiusens, not every compiled app needed building [15:01] yeah, just the one we had issues with [15:01] just the few identified in the testing sheet [15:01] sergiusens: just fiddle. I'll land click chroot support next week [15:01] mhr3_: it might take a few moments to finish, but then we move full forward with those silos [15:01] sil2100, right, i can rebuild 001 though? [15:01] sergiusens: you could copy the chroot in /var/lib/schroot/chroots/ and /etc/schroot/chroot.d/ [15:02] it's only creation that won't work [15:02] ok' I'll see what I can get out of it [15:02] mhr3_: yes, rebuild it by mentioning unity-scope-click in the rebuild packages field [15:02] sil2100, well that's the only pkg for that line :) [15:03] mhr3_: yes ;) But otherwise it won't rebuild, as it will say: all packages have already been built, or something [15:03] sil2100, checking force [15:03] oh, right, seems that's not enough [15:04] pmcgowan: the light-sensor do you happen to know which package I need to file against I'm assuming qtsensors [15:04] davmor2, ask ChickenCutlass ^^ [15:04] mhr3_: entering the name there should be sufficient [15:04] davmor2: powerd afaik [15:08] Mirv: will we have an image with just the qt5.2 bits? [15:09] i.e. should I block publishing other landings in the meantime? [15:09] didrocks: ^ [15:09] ChickenCutlass: I need to write up a bug for the light-sensor not doing anything and need the package to file against could you confirm if it is qtsensors or powerd or something else please [15:09] sil2100: no please [15:09] davmor2, powerd [15:10] rsalveti, ChickenCutlass: thanks guys :) [15:12] sil2100: we'll wait for the messaging-app and click app fixes I think [15:13] sil2100: didrocks: landing-006 merge & clean complete! [15:13] \o/ [15:13] mhr3_: good news, assigning silos for you mister [15:13] didrocks: I also handled the u1db-qt by merging manually the one branch plus syncing changelog [15:13] sil2100, can i get #21 too? [15:13] Mirv: thanks! [15:14] didrocks, rsalveti, so my image watcher bot is now running in test mode, i'll need it to watch a few image builds over the weekend to make sure it doesnt cause issues (need some cron runs for this) before letting it into this channel ... if all goes fine i'll switch it fully live on monday morning [15:14] mhr3_: no, it would require too much scrolling through the list, sorry [15:14] mhr3_: ;) [15:14] lol [15:14] mhr3_: let me take a look after those, I don't want us to starve any other upstreams in the meantime [15:14] ogra_: awesome [15:14] ogra_: but cron is currently disabled :-) [15:14] rsalveti, sigh why ? [15:14] ogra_: because we want to get this click framework fixed [15:15] popey: a confirmation if you would please https://bugs.launchpad.net/ubuntu/+source/powerd/+bug/1292565 [15:15] Ubuntu bug 1292565 in powerd (Ubuntu) "light sensors don't seem to do anything." [Undecided,New] [15:15] rsalveti, and we wont get it done today ? [15:16] sil2100, doing thomas' stuff as he's on holiday, so i'm like 4 upstreams now [15:16] ogra_: people are trying really hard but hard to say [15:16] Awesome [15:16] heh k [15:19] sil2100: cjwatson's ubuntu-ui-toolkit arch enablement branch is not yet landable since there's a chain of unfilled dependencies (-> unity-action-api -> hud -> dee-qt). should I remove it from the landing-005 so that qtpim + qtcreator-plugin-cmake + appmenu-qt5 could be landed? [15:20] I'm commenting on the branch anyhow for now [15:21] hmmm [15:21] or I wonder of course if it's just landed, does it help something on its own [15:21] sil2100, eh, meant 26, not 21 [15:22] Mirv: I wonder, I would love getting appmenu-qt5 fixed as the copyright had some problems, but I would say: do as you think is best [15:23] Mirv: what's cjwatson's opinion? [15:23] sil2100: he's away today [15:24] didrocks, even if the tests will fail all over; can't we trigger a new image with the new click definitions and Qt 5.2? [15:24] sergiusens: what do you mean, without the rebuilds? [15:25] sergiusens: if you don't have the upload in the store with the new framework, you won't get the app working [15:25] didrocks, yes, without the rebuilds [15:25] mhr3_: hmmm, now that I check the spreadsheet: look at landing on line 9, I see some of those merges in the silo I just assigned to you [15:25] didrocks, that's why I said fail all over [15:25] mhr3_: is this landing therefore invalid? Did you take all those merges from line 9 and include it in the new one? [15:25] sergiusens: yeah, why do you want to do that without the rebuilds then? [15:25] sil2100, disregard 9, i took it over [15:25] mhr3_: ACK [15:26] didrocks, I just want an image to test that the rebuilds and new framework definitions work [15:26] sergiusens: people will upgrade to that image [15:26] and will get a broken phone [15:26] didrocks, people shouldn't be on devel-proposed without accepting that casualty [15:26] sergiusens: yeah, but people will do [15:26] and it's before the week-end :) [15:26] didrocks, that's what devel-proposed is for [15:27] I think we shouldn't let people breaking that much their phone on purpose [15:27] (for at least 2 days) [15:27] Mirv: do you know of any planned fixed for address-book-app? Or can I lock it for a landing? [15:27] if only we could do "unupdatable images" [15:27] sil2100: no, I don't know, feel free to lock [15:28] Mirv: thank you :) [15:28] sergiusens: you can flash latest and dist-upgrade [15:28] yeah, seems the best to me [15:29] rsalveti, ack [15:29] didrocks, "people" shouldnt run -proposed ... that will teach them ;) [15:32] sergiusens: you need to wait on click-apparmor 2.0 for having latest framework definition [15:32] didrocks, ok [15:34] dbarth: I'll be assigning a silo for you as well, just want to double check if all merges are bug fixes for non-FFe'd components first [15:34] sergiusens (and didrocks): 0.2. I'm not to 2.0 yet :) [15:35] jdstrand: I see you always diminish yourself :p [15:35] geeez, you are so behind ! [15:35] jdstrand: so, as told on #ubuntu-touch (but with right numbers): [15:35] ubuntu-sdk-13.10 -> ubuntu-sdk-14.04-dev1 [15:36] policy 0.1 -> 0.2 [15:36] for migrating to the new fw? [15:38] ah no, it's 1.0 -> 1.1 [15:39] from the code [15:39] popey: sergiusens: FYI ^ (but maybe wait for Jamie to confirm) [15:40] thanks [15:40] sil2100: ogra_: robru: popey: davmor2: I'll be 15 minutes late in the meeting, I really need a break now and running :) [15:40] didrocks: want me to mail the community devs with broken packages? [15:40] so they know policy and framework text for their manifests? [15:40] sil2100: \o/ we can go crazy now! :) === bfiller_afk is now known as bfiller [15:41] popey: well, let's get jdstrand confirming that I'm reading right the expected policy_version, but after that, yeah please :) [15:41] didrocks: wibble [15:41] didrocks: click-apparmor went from 0.1 to 0.2, yes. however apparmor policy is in apparmor-easyprof-ubuntu, and it is 0.1.x. it is versioned to match the policy. ie, apparmor policy major.minor corresponds to apparmor-easyprof-ubuntu major.minor.microrevision [15:41] kk [15:41] err [15:41] 1.0 [15:41] sil2100: let me review real quick but i'd like some priority bump to get a silo on our desktop landings that were locked [15:41] man, I should stop transposing numbers myself! [15:41] didrocks: sure! It's a well deserved break ;) Too bad you'll be running and getting tired instead of resting ;p [15:41] meh [15:42] sil2100: can keep some of the touch stuff aside not to monopolize too many silos, for ex. [15:42] jdstrand: hum [15:42] apparmor-easyprof-ubuntu is 1.1.x, that corresponds to apparmor policy 1.1 [15:42] dbarth: which ones? Since I saw 2 that still didn't have their FFe's approved, but there was one more I wasn't sure about [15:42] jdstrand: ah ok :) everything's fine for you popey, sergiusens? ^ [15:42] click-apparmor 0.2 isn't versioned in reference to the policy-- its versioning is separate [15:42] didrocks: no ☻ [15:43] what should the policy be in the manifest? it's 1.0 for 13.10. What should it be for 14.04-dev1? [15:43] sil2100: yup, hang on [15:43] * sergiusens is lost and waiting for the waves to calm down a bit [15:44] +1 [15:45] jdstrand: maybe you can just give popey and sergiusens a practical example of what they need to update? ^ [15:45] \o/ [15:45] didrocks, it's not just this, it's the list of everything that needs to get done fwiw [15:45] I am waiting for people to notice new things [15:46] that's what I meant with waves :-) [15:46] doanac`, plars, hey, did you guys see my question yesterday about smoketests and unlocking unity8? [15:46] sergiusens: you have a starting list with what I collected for you, right? [15:46] Saviq: not sure i did. can you repeat? [15:47] is it OK to get [15:47] "There is 1 newly obsolete package: libqt5core5" ? [15:47] * Saviq grabs the log [15:47] was the package renamed? [15:47] didrocks, yah and the framework definition changed twice already ;-) [15:47] Wellark, yes [15:47] seb128: all the indicator-datetime merges are fixes not requiring FFe's, right? (just making double suar) [15:47] it's ...a now, due to the double/real ABI break [15:47] seb128: from the landing you added [15:48] popey, sergiusens: sorry, I came on to this thing at the tail end and have been furiously uploading (my prepared, but not siloed) click-apparmor up until now [15:48] popey, sergiusens: what is the problem? [15:48] doanac`, wanted to ask about the unity8 unlock script used in smoke tests, we're removing the thing you're relying on soon, fginther and om26er already implemented the new method using unity8's helpers that we maintain, apparently you tried it and backed out around a week ago, can we help making the switch somehow? [15:49] jdstrand: we have apps which need updating in the store - some are community, some ours. [15:49] jdstrand: the community people need to know what to change their manifest to, specifically the framework string and policy version. [15:49] Saviq: i think the problem we hit was that the library the new mechanism relied on wasn't in the image [15:49] jdstrand, no problem, I'm just saying I'm not going to make changes to the click apps until eveyhing lands ;-) [15:50] jdstrand: so that when their new "only qt5.2 builds" versions are uploaded, they cannot be installed on older non-qt5.2 images of the phone, which have an older policy version and older framework string [15:50] om26er, fginther, can you say how you solved doanac`'s issue in ci testrunners? [15:50] doanac`, Saviq unity8-autopilot needs to be installed [15:51] ofcourse Saviq knows that ;) [15:51] doanac`, can the same be done for smoke tests? [15:51] popey: this has the frameworks: https://wiki.ubuntu.com/Click/Frameworks [15:51] Saviq, doanac`, for the upstream merger testing, the devices are already in writable mode, so we just 'apt-get install unity8-autopilot' [15:51] popey: that is not ratified yet I don't think, but it is close. talk to pmcgowan and lool about that part [15:51] I guess seen unity8-autopilot on the image then [15:51] *seed [15:52] or do we need to extract the helpers into a separate package and seed that? or can we maybe make phablet-test-run install unity8-autopilot to $HOME/autopilot? [15:52] hmm [15:52] popey: so that lists the frameworks. I create a query interface for the sdk to use to ask for the policy version for a given framework [15:52] doesn't it already? [15:52] Saviq: daily-image testing using the unity8 click package not the deb [15:52] popey: ie, aa-clickquery --click-framework= -q policy_version [15:52] doanac`, there's no unity8 click package ;) [15:53] popey: that will give back something like "1.1"-- ie, something suitable for the security manifest [15:53] seb128: ok, it looks like it at least, so assigning a silo [15:53] doanac`, but anyway, can't unity8-autopilot be installed into $HOME/autopilot as is done for click tests anyway? [15:53] Saviq: n/m. the issue is we don't install unity8-autopilot until we test unity8. and we remove it when we are done testing it [15:53] popey: now that 0.2 is in the archive, I'll file a bug so bzoltan1 can stop hardcoding the policy version === sil2100_ is now known as sil2100 [15:53] so it sounds like you need unity8-autopilot to always get installed [15:54] popey, jdstrand I thought for "today" we are using -dev1 [15:54] isnt that what I saw fly by [15:54] doanac`, yup [15:54] popey: but, for app developers, if they use ubuntu-sdk-13.10, they should use '"policy_version": 1.0'. if they use any ubuntu-sdk-14.04* framework, they should use '"policy_version": 1.1' [15:54] pmcgowan: there is the security-policy as well, it's the part jdstrand is answering on I guess [15:54] I see [15:54] excellent, thanks jdstrand, I think that's what popey and sergiusens needed :) [15:54] didrocks, asac: as per screen unlocking. is it okay if we install unity8-autopilot when we provision a device and leave it there for the entire image testing run? [15:55] sil2100, yes, no ffe needed, no feature/behaviour changes there, just bugfixes, less polling on eds, etc [15:55] doanac`: I think it's fine, however, can you refrain doing any infra change today? (as we are doing the Qt transition) [15:55] sure-- I just wanted to give some background so people can know about aa-clickquery and not hardcode [15:55] yeah ;) [15:55] * didrocks notes the aa-clickquery command [15:55] didrocks: yeah. i can't get around to this now. i'm swamped as well [15:55] mhr3_: regarding landing line 26... [15:56] sergiusens: popey: so, all cleared? :) [15:56] mhr3_: I don't see hud having an FFe, are all of the merges there fix-only? [15:56] popey: and click-reviewers-tools were updated last week to deal with these frameworks and apparmor policy versions [15:56] pete-woods, ^^ [15:56] mhr3_: since I see some 'use-new-libdbusmenu' etc. [15:56] sil2100: yes, they are all bug fixes [15:57] popey: so, with click-apparmor 0.2, the frameworks being shipped in /usr/share/click/frameworks and the review tools being updated, it should all work [15:57] popey: obviously, if there is a bug in the review tools, file it and we can discuss how to handle in the app review [15:57] sil2100: that branch just adds a dependency on a version of dbusmenu-qt with a bugfix we need on it [15:58] pete-woods: could you, just as a formality, attach the bug that's getting fixed to that branch? [15:58] ok. I think we're clear now. old = ubuntu-sdk-13.10 & policy 1.0, new = ubuntu-sdk-14.04-dev1 & policy 1.1 [15:58] doanac`: I think so, but do we want to change this today? :-) [15:58] pmcgowan: can you remind me what the package is that I should file the bug against to not hard code policy version in qtc/the sdk? [15:58] sil2100: absoutely [15:58] doanac`: should be fine to be changed next monday/tuesday [15:58] pete-woods: thanks! I'll assign a silo straight away once that's done [15:58] popey: right! sergiusens: all good for you as well or you need more info? [15:59] pmcgowan: I can never remember for some reason [15:59] Saviq, didrocks, rsalveti: https://bugs.launchpad.net/ubuntu-ci-services-itself/+bug/1292585 [15:59] Ubuntu bug 1292585 in Ubuntu CI Services "touch daily image test: needs unity8-autopilot installed for screen unlock" [High,New] [16:00] sil2100: you are doing the messaging-app AP tests or bfiller is doing it? [16:00] didrocks: so are requests that were blocked on 5.2 now unblocked? [16:00] bfiller: yeah, but we need to be able to build an image first [16:00] meaning, having the messaging-app query in [16:00] didrocks: was about to test what's in the silo [16:00] didrocks: I think it's bfiller, as I didn't hear anything [16:00] bfiller: excellent, then, we'll have the click store + a test run [16:01] bfiller: thanks! [16:01] bfiller: we assign silos on the rest, just refrain on publishing [16:01] thanks :) [16:01] Phew, so many silos assigned [16:01] didrocks: the best way to test messaging-app AP test is install latest proposed image and add silo 6 still for the qt5.2 stuff? [16:01] bfiller: silo 006 is cleared now, so no [16:02] bfiller: no, install latest proposed-image and dist-upgrade [16:02] bfiller: you can get the latest image and then dist-upgrade I guess [16:02] didrocks: got it, doing now [16:02] thanks [16:02] didrocks: ruuun! [16:03] doanac`, https://bugs.launchpad.net/ubuntu-ci-services-itself/+bug/1292585/comments/1 [16:03] sil2100: well, I guess it's going to be the whole meeting I'll miss :) [16:03] Ubuntu bug 1292585 in Ubuntu CI Services "touch daily image test: needs unity8-autopilot installed for screen unlock" [High,New] [16:03] sil2100: can you lead it? I think the instructions are quite clear anyway :p [16:03] didrocks, anything to discuss ? [16:03] didrocks: will try ;) No worries! [16:03] well, basically, now it's: [16:03] * ogra_ thinks we could as well just skip [16:03] - getting messaging-app in [16:04] didrocks: you'll write the landing e-mail later though? Or should I do it as well? [16:04] - getting the click side (gallery-app, terminal-app, notes-app with MP fixes, music-app with MP fixes) in [16:04] with the new framework and policy [16:04] hmm, did anyone get a calendar notification ? [16:04] and kick an image + dogfood it [16:04] ;) [16:04] sil2100: yeah, I'll do that [16:04] i dont see the meeting on my calendar [16:04] ogra_: nope [16:04] ok, should I remove it? [16:04] ogra_: any meeting? It's still one hour [16:05] oh [16:05] crap+ [16:05] It's not like now ;) Just in an hour [16:05] all my other meetings are 1h earlier now [16:05] sil2100: okay, I've linked the bug now [16:05] so, nuking it? [16:05] pete-woods: thanks, will assign in a moment! [16:05] i would be fine with that [16:05] sil2100: thanks very much! [16:06] didrocks: I guess we can - anything else we can pass on IRC, and the thigns I have to say I'll just send over by IRC [16:06] ok [16:06] let me prepare and send the email now then [16:06] that will be better for information sharing [16:06] ++ [16:06] Right [16:07] and then you can run til dawn if you like :) [16:07] ogra_: your other meetings were created by US folks [16:07] rsalveti, sure, i'm just getting used to it [16:07] cyphermox: balloons: FYI, meeting deleted. All info in landing email [16:07] (incoming :p) [16:07] since *all* other meetings i have are US ones :) [16:07] yeah, they need to learn how to use UTC at some point [16:07] robru: ^ as well ^ [16:07] ok [16:07] didrocks, oO [16:07] it's not that hard [16:07] rsalveti, ++ [16:08] yeah === josepht changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: cihelp | CITrain support - US: robru, cyphermox, rsalveti - EU: sil2100, Mirv, didrocks | CITrain support no answer: use mup bot after 30 minutes, but choose right timezone [16:08] sil2100, didrocks what? my calendar has the meeting on UTC. is the meeting not in an hour? [16:08] it was in one hour [16:09] it's not anymore :) [16:09] dead [16:09] killed [16:09] why are we not having a meeting? is there not a lot to talk about re:qt? [16:09] ok, I'm going back to my holidya [16:09] holiday [16:09] jdstrand: enjoy! ;) [16:09] jdstrand: thanks so much, enjoy! [16:09] robru: everything depends on sergiusens + popey now. I'll include the few info in the landing team email [16:10] I'll check email. file bugs if there is an issue [16:10] didrocks, oh ok. because I have people who are just screaming for silos now ;-) [16:10] robru: yeah, see the header of the spreadsheet [16:13] hi folks. could someone please rebuild https://code.launchpad.net/~barry/camera-app/py3autopilot/+merge/210922 [16:13] the error is strange [16:13] popey, can you take care of one thing? [16:14] popey, the coreapps ppa would need rebuilds of the plugins [16:14] didrocks, isn't there a landing meeting today? [16:14] sergiusens: they're triggered from jenkins I thought? fginther ? [16:15] sergiusens: no, I try to sum up everything needed, but we can hop on now if you need more info [16:15] popey, not sure; but they'd need version bumps [16:16] didrocks, if this is what needs to happen http://paste.ubuntu.com/7090970/ then I'm fine [16:16] sergiusens, rebuilds due to a new revers depends? [16:16] fginther: due to qt5.2 [16:16] didrocks, but to make it faster; I would like to manually merge the framework changes into the trunks [16:16] sergiusens: one sec [16:16] didrocks, proper review; but manual merges [16:17] sergiusens: notes-app needs https://code.launchpad.net/~elopio/notes-app/skip_bug1288876 [16:17] sergiusens: https://code.launchpad.net/~elopio/notes-app/dont_click_header/+merge/210349 [16:18] didrocks, can't someone else merge that? [16:18] bfiller, can you make sure that lands ^? [16:18] sergiusens: bfiller, mind merging those manually? (as we told we'll take the deb part later on) ^ [16:19] sergiusens: on the merge directly framework change, yeah, +1 [16:19] didrocks, that can be merged with the normal train if you want [16:19] sergiusens: and apart from those 2 mentions missing, yeah, +1 for all [16:19] sergiusens: no, let's try to not mix the .debs with .click part, we have too many things in transit already [16:19] didrocks, if it's manuall merge I'll do it [16:19] * didrocks hates having both .deb and .click for the same component, that's so annoying :/ [16:20] sergiusens: yeah, please ;) [16:21] sergiusens: then, as soon as you are done and popey or balloons publishes all of them, we can get an image kicked [16:21] rsalveti: you will probably be around I guess? ^ (and remember then to reenable the cron) [16:21] cjwatson, fwiw and I'm working around it there's a dict in /usr/lib/python3/dist-packages/click/chroot.py with only 13.10 defined (ignore if something is in flight now with that) [16:22] robru: while I'm are it, line 48 is the only one we need to publish before the next image is in (so BIG priority), bfiller is doing the testing as we speak [16:22] didrocks, bfiller ok I'm here. ping me when it's time to publish [16:22] robru: just make sure to look at it if it's yellow or not ;) [16:22] didrocks, popey: if there is an emergency issue, please contact mdeslaur and sbeattie [16:22] robru: i.e. if 'tested: Yes' [16:23] jdstrand: excellent, I'm sure everything will be fine! [16:23] * jdstrand goes on holidya for real [16:23] sil2100, you want me to *poll* like some kind of *peasant*?? I demand push notifications! [16:23] yes, me too [16:23] heh, enjoy for real :p [16:23] jdstrand, if you are on holidays, Isay, don't worry and leave :-) === Ursinha is now known as Ursinha-afk [16:24] I jdstrand thanks. [16:24] -typos [16:26] sergiusens, didrocks : notes-app trunk has those changes merged in now [16:26] great! [16:26] :O [16:26] robru: :O [16:26] ;) [16:28] sil2100, oh I just had the greatest idea... citrain should use mup to send SMS to landers the exact second when a silo is marked "testing: yes" [16:28] ;-) [16:28] didrocks, ^ [16:28] robru: patch? :p [16:29] didrocks, haha, maybe! [16:29] didrocks: yes [16:29] robru: ;) [16:29] no wait, short circuit that whole plan... we just need an IRC bot that *pings* landers when testing is set to yes.... [16:29] robru: already discussed and bug opened yesterday [16:30] didrocks, i ... i think i love you [16:30] ;) [16:30] lol [16:30] hmm I can't dist-upgrade my trusty install :-/ qtcreator-plugin-ubuntu : Depends: qtcreator-plugin-cmake but it is not installed [16:30] ;D [16:31] sergiusens: phone or desktop? [16:31] didrocks, desktop [16:31] sergiusens: sudo apt-get install qtcreator-plugin-cmake? [16:31] didrocks, I'm trying to get click chroot working with the latest an greatest :-) [16:31] sergiusens: latest for sure… [16:31] sergiusens: greatest, we'll see :p [16:31] is the morale that low? [16:31] :-) [16:31] ahah [16:31] didrocks: so once we get the new clicks should we be good to trigger a new image? === Ursinha-afk is now known as Ursinha [16:32] or you just want to wait the one done by cron? [16:32] if greatest cannot be aspired to; there was no reason to update Qt ;-) [16:32] rsalveti: exactly, I expect that messaging-app will be in first (it's the only archive-side change we wait on) [16:32] rsalveti: no please, kick in, the earlier, the better [16:32] alright [16:33] sergiusens: so, apt-get install complains about what for this plugin? :p [16:33] didrocks, I apt-get removed it for now; so no sdk for me :-) [16:34] sergiusens: that was an option :p [16:34] Who needs the SDK anyway! [16:34] whats an SDK ? [16:36] sil2100: still on it, getting back to you quickly now [16:36] dbarth: ok, just ping me once done [16:50] didrocks: tests running for messaging-app for some time now, no failures yet but not complete .. [16:50] bfiller: excellent o/ [16:50] there are a lot of tests :) [16:51] Mirv: is it build-deps that are unsatisfied for ubuntu-ui-toolkit/arch-any, or only runtime deps? if it's build-deps, then we can go ahead with that landing because it'll harmlessly dep-wait for now and be useful later [16:51] rsalveti: sil2100: robru: popey: ogra_: Mirv: davmor2: cyphermox: balloons: ToyKeeper: sergiusens: so updated the ML, tell me if anything is unclear [16:51] Mirv: if it's *only* runtime deps (i.e. it'll build, but build uninstallable packages) then we should indeed hold off [16:51] cjwatson: sounds good to me [16:51] sergiusens: yes, I know, that's what I was referring to earlier [16:51] sergiusens: I'll fix it next week [16:51] sil2100: checking silo6, reconfigure hates me, did i do something wrong: http://162.213.34.102/job/landing-006-0-reconfigure/3/console [16:51] cjwatson: I'll recheck when I get some times [16:52] bfiller: no worry ;) [16:52] dbarth: what did you try reconfiguring? [16:52] bfiller: I think you will still be faster than the click store [16:52] dbarth: I mean, what happened that you wanted to reconfigure? [16:52] didrocks,Mirv: looking at the build in landing-005, it's harmless dep-waits [16:52] so safe to land anyway [16:52] just means the dep chain is deeper than I'd thought, but it won't hurt [16:52] cjwatson: yeah [16:52] let's land it [16:52] didrocks: Ran 16 tests in 482.296s [16:52] OK [16:53] not sure about the source upload then [16:53] bfiller: \o/ [16:53] didrocks: so looks good :) [16:53] bfiller: \o/ [16:53] sil2100: do you know? [16:53] cjwatson: thanks [16:53] bfiller: put it to tested: yes [16:53] ah Mirv is still around :) [16:53] Mirv: so, can you deal with that one? [16:53] sil2100: done [16:53] didrocks: not sure, Mirv knows best ;) [16:53] bfiller, didrocks: landing messaging-app! [16:53] yeah, I was thinking he left for the week-end [16:53] sil2100: waoooowwwww [16:53] thanks bfiller for the testing [16:53] robru: ^ [16:53] dinner, tests running for qtpim will report [16:53] didrocks: np [16:53] robru: landing 007! [16:54] popey, do you know which plugins need rebuild? [16:54] sil2100: robru: ok, so for everything else than silo 005, please wait for next image now [16:54] heh "fundations" [16:54] the funny guys :) [16:54] ogra_: I didn't know how to call that :p [16:54] didrocks: ok [16:54] robru, didrocks, bfiller: published, now we just wait for clicks I guess ;) [16:54] sweet! [16:55] we can just stare at beuno, sergiusens and popey now [16:55] actually, I missed my exercise of the day [16:55] and it's too late now [16:55] \o/ [16:55] we could have our meeting :p [16:55] but nothing more than the email anyway [16:55] didrocks, message makes sense, thanks for the updates as usual [16:56] dbarth: anyway, if you wanted to reconfigure the silo by adding a new merge proposal, you have to list all the merge proposals that are to be in the final silo [16:56] sil2100: re-adding webbrowser-app bits that i had removed from the original request [16:56] balloons: yw! I except very good news from you on the incoming image of course :p [16:56] to try and land during the big qt5-2 lock [16:56] fginther: at least the konsole one used by terminal, a music too, right sergiusens ? [16:56] dbarth: then just copy-paste the MR list from the spreadsheet to the MERGE_PROPOSALS field in the reconfigure job [16:56] oh i see [16:56] vUDS behind us it's a bit easier to get back into things :-) [16:56] dbarth: and you're ready to build ;) [16:56] dbarth: (once done) [16:57] sil2100: seems I didn't break the publication with my changes :) [16:57] from this morning :p [16:57] sil2100: have you run another m & c? [16:57] didrocks: hah ;) Indeed! No, I didn't run a m & c yet after the qt5.2 [16:57] So this we'll still have to check [16:57] sil2100: and it blames me for trying to put webbrowser-app in :/ [16:58] dbarth: aaah, right, new component [16:58] sil2100: yeah, let's see :) [16:58] sil2100: is that webbrowser-app package locked already? [16:58] dbarth: in this case I have to add it myself, let me do it [16:58] dbarth: or wait, probably... hmmm [16:58] dbarth: the message should be explicit, no? [16:58] bfiller: mind if we do a quick landing for webapp-container stuff in webbrowser-app? [16:58] do you have the link? [16:58] didrocks: it is; it says i'm a naughty boy [16:58] dbarth: I don't see it assigned anyway... [16:59] didrocks: email looks good to me :) [16:59] dbarth: that's fine, you can add it to the existing line I have already for that [16:59] * bfiller looking [16:59] bfiller: i was on line 40 [16:59] dbarth, didrocks: right, I have to reconfigure the silo [16:59] popey, hmm, music uses qtpowerd and grilo; iirc that's in the archves [16:59] So all is ok [16:59] sil2100: let's see what bfiller says [17:00] sergiusens: correct [17:00] popey, that was rebuilt; correct? [17:00] dbarth: we requested line 38 as well for a bunch of browser changes, wonder if we should combine? [17:00] bfiller: if you want to combine, you can reconfigure yourself to add the MPs on the same component [17:00] bfiller: checking with alex-abreu if we need it to complete testing or not [17:01] * sil2100 is waiting for decisions [17:01] ;) [17:01] sergiusens: https://lists.ubuntu.com/archives/trusty-changes/2014-March/011823.html & https://lists.ubuntu.com/archives/trusty-changes/2014-March/011755.html [17:01] (yes) [17:01] sil2100: sorry about that [17:01] (guru meditation) [17:02] ok sounds like we can split [17:02] bfiller: ok, can you take the 2 MPs in your silo; we can still our other webapp changes in the silo, without those [17:02] sil2100: i'll just remove the offending branches, and should be good by myself now [17:03] bfiller: will send the 2 orphan MPs to adopt [17:03] dbarth: ok :) [17:03] dbarth: ack, please add them to line 38 then === retoaded changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: retoaded | CITrain support - US: robru, cyphermox, rsalveti - EU: sil2100, Mirv, didrocks | CITrain support no answer: use mup bot after 30 minutes, but choose right timezone [17:03] that's so lovely said :) [17:04] didrocks, bfiller, dbarth: in the meantime I'll assign a silo for the webbrowser-app changes as well, once you add the additional merge to that [17:04] sil2100: sure, but let bfiller self-reconfigure if needed (not sure he tried already) [17:04] sil2100: thanks, can I also have a silo for line 19 gallery-app please [17:04] didrocks: have not tried that yet [17:05] bfiller: hm, isn't gallery-app already click? [17:05] bfiller: me and Mirv were wondering what to do with that one actually ;) [17:05] sil2100: it is but we still need deb for the desktop [17:05] * bfiller doesn't understand the process for clicks [17:05] Same here, a bit... still confused [17:05] bfiller: if you can feel better, we don't as well :p [17:05] :) [17:05] bfiller, sil2100 : yes I think that's because there's *no* process for clicks. they go into the click store with no testing... [17:06] robru: not good [17:06] bfiller: MPs sent [17:06] bfiller: my plan was to discuss that at vUDS, but we didn't get to that, so will rehash with sergio to have the same process next week [17:06] bfiller, I guess we test certain click apps after they're in the image, but that means they have to get into the click store untested first. [17:06] fginther, hey, can you have a look at the dependency error in https://jenkins.qa.ubuntu.com/job/autopilot-testrunner-otto-trusty/3427/console ? [17:06] didrocks: I figured for now we still use the train as we still needs the debs in the archive [17:06] (to support running on desktop) [17:07] sil2100: we're good for silo 6, could we also get our html5 fixes re-activated on line 16; that's only small bug fixes [17:07] bfiller: yeah, we really need to be able to factorize both on the same silos or to have click running on desktop, it's a pain to do 2 releases for the same app basically [17:07] sil2100: or should we add to silo 6? [17:07] was hoping that magically everytime the deb hit the archive the click would get uploaded ot the store [17:07] dbarth, whoa whoa, careful who you say "silo 6" to... I am traumatized by that number ;-) [17:07] Mirv, hm, i still don't see the rebuild* branches marked as merged... is that still going to happen? [17:07] robru: really? ;) [17:08] bfiller: too much magic :p but yeah, it's an idea I've in my mind [17:08] dbarth, (it was the silo that had qt5.2 in it recently) [17:08] ;) that was a good one; it's landed [17:08] click archiving practices are a bit lol [17:09] cjwatson: you mean, as you supposively day off? :) [17:09] shockingly it turns out that the twenty years we've spent refining how .deb archiving works were actually worth something [17:09] like* [17:09] didrocks: I've spent most of it driving round in a van as intended, sitting at a computer constitutes relaxation by comparison :) [17:10] cjwatson: ok, in that sense… ;) [17:10] sil2100: so I'd say until that click magic is figured out we need to continue to build the deb, so we'd need the silo for line 19 [17:10] sil2100: at least that allows us to land on the desktop [17:10] the autoloading is the easy part; the machinery to make sure it doesn't break anything is what needs to work [17:11] cjwatson, tedg do the click hooks need any updating to take into account the new framework? [17:11] not afaik [17:11] aside from the apparmor hook which is being dealt with [17:11] is being / has been [17:11] cjwatson, so are your springs shiny and polished now ? [17:12] :) [17:12] dbarth: let me see line 16 [17:12] bfiller: sil2100: I guess now we can get notes-app a silo as well? (maybe an empty MP to deliver it?) [17:12] ogra_: no, but marginally less grubby and cluttered ... [17:12] (ready isn't set to yes) [17:12] bfiller: mind joining our hangout for a very quick chat? [17:12] cjwatson, ah, pkcon hides the error :-) ERROR: Could not transform 'com.ubuntu.notes_notes_1.4.253.json' to AppArmor easyprof. Skipping [17:12] might still be in flight [17:12] bfiller: ok, I'll assign for webbrowser-app and gallery-app for now [17:12] dbarth: sure [17:12] bfiller: would like to arbitrate oxide/content-hub priorities [17:13] dbarth: let's maybe deal with that once others are in, ok? ;) Since we're already using up most of the silos! [17:13] sergiusens: Yes, that's from aa-clickhook [17:13] dbarth: link? [17:14] bfiller: alex-abreu sent it [17:15] libqt5graphicaleffects5 <- ubuntu-ui-toolkit is showing up on http://people.canonical.com/~ubuntu-archive/nbs.html. Is somebody looking into that? [17:16] ogra_: German efficiency, as always… :p [17:16] didrocks, Mirv what is the deal with libqt5core5 being renamed, is that an upstream change? [17:16] oh, that's an alternate dependency, it seems [17:16] pmcgowan: defines "upstream" [17:16] qt project [17:17] pmcgowan: hum? no, they broke the ABI [17:17] pmcgowan: that was requested by Steve for us to force the rebuilds for the ABI/qreal change http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=731261 [17:17] Debian bug 731261 in release.debian.org "transition: Qt5 switching qreal == double for all platforms" [Normal,Open] [17:17] slangasek didn't want that we rebuild without changing the binary package name [17:17] Mirv, but doesnt that force all apps to rebuild [17:17] to deal with the transition [17:17] pmcgowan: well, the ABI transition does [17:17] no [17:17] actually did [17:17] it doesnt [17:17] pmcgowan: it doesn't force all click apps to rebuild [17:17] it's all dealt now [17:18] pmcgowan: it forces anything that's a deb to rebuild, but most of that is already done [17:18] but every app with c++ code? [17:18] (apparently with the exception of the core apps PPA) [17:18] pmcgowan: it's a .deb thing only [17:18] not just the few effected by the qreal change? [17:18] it is hardly worth going back now! [17:18] debs are easy to rebuild [17:18] maybe I am confused [17:18] source upload, bang, done [17:18] pmcgowan, the .so name didn't change [17:18] pmcgowan: well, seems nobody did the impact of what was impacted by the qreal change [17:18] pmcgowan, only the package name did [17:18] ah ok nm me then [17:18] there wasn't a way to break only the few affected by the qreal change [17:18] sorry [17:18] in terms of .deb dependencies [17:19] cjwatson, I get it now [17:19] all good [17:19] and for those it's better to be conservative [17:19] yes [17:20] didrocks, so no silo publishing today yet? so no point for me to expedite testing of the unity8 silo... [17:20] Saviq: well, until we kick an image. I think it's better to put changes on changes [17:20] Saviq: so yeah, no pressure :) [17:20] fginther, can you do a rebuild of things in the core apps ppa so they will install again aftr qt5.2 hit the archive? [17:20] Saviq: or you can publish, once the image is there, it will be published for you [17:20] Saviq: as you prefer :) [17:20] sorry [17:20] Saviq: or you can *test*, once… [17:22] didrocks, right, maybe I'll do that [17:23] balloons, yes, popey asked a little bit ago and I'll have this started soon [17:23] didrocks, heh [17:24] cjwatson: am I right in saying "pkcon install-local foo.click" is the *right* thing to do, and "click register ..." is the *wrong* thing to do, to fudge a click package onto a phone? [17:24] popey: man click [17:25] bash: man: command not found [17:25] :þ [17:25] on your laptop [17:25] ☻ [17:25] there exists an invocation of click register which works [17:25] or rather of click install [17:25] oh, groovy. I'll try that then [17:25] but it's a bit hard to get right, and there isn't much indication when you get it wrong [17:25] so pkcon is now bad mmkay? [17:25] huh? I never said that, I was mid-typing [17:25] so you should use pkcon install-local instead, because it's harder to get wrong [17:26] ah okay. [17:26] didrocks, pmcgowan: we analyzed the impact of the qreal change on the app store, and we know we are lying when we say that the 13.10 framework is still supported. For deb dependencies, which are based on package name, there was no reason to *not* accurately declare the ABI break, because we were in full control of the reverse-dependency tree [17:26] bfiller, https://code.launchpad.net/~sergiusens/notes-app/framework-update/+merge/211104 [17:26] and non-packaged reverse-dependencies will be broken, but that's upstream's doing [17:26] the apparmor change is in the archives (release) :-) [17:26] the reason I ask is because some apps appear to break interestingly when I use pkcon and wanted to make sure it wasn't "notabug, use click" [17:26] slangasek, understood [17:27] popey: "This is a low-level tool; to install a package as an ordinary user you should generally use pkcon install-local PACKAGE-FILE or some higher-level user interface instead, which take care to use the correct set of options. (Do not use sudo when invoking pkcon, as it needs to know the calling user.)" [17:27] popey: that's what "man click" says of click install [17:28] cjwatson: ok, good. I'll file bugs then. [17:28] popey: if you're using pkcon without sudo and things break, I certainly want to investigate (but when I'm actually at work) [17:28] if you're using pkcon with sudo I will laugh :) [17:28] hah [17:28] I am using pkcon install-local *.click [17:29] and yes, this is not a friday evening panic, just wanted to make sure I file bugs in the right place [17:29] sergiusens: merged [17:29] sergiusens: need for gallery as well? [17:29] robru: just to confirm, if I remove one MR from a component, can I just click Reconfigure to prepare the silo again? [17:30] popey: "pkcon install-local foo.local" translates to "sudo click install --user= foo.click" under the hood [17:30] didrocks: so, anyway! Since there was no evening meeting today: [17:30] s/foo.local/foo.click/ [17:30] didrocks: the unity8 silo has a 'workaround/fix' for the failure in unity8 (url-dispatcher one) [17:31] didrocks: system-image-dbus we should have fixed soon as well [17:31] cjwatson: right, so click doing the legwork, and bugs should be filed in click? [17:31] didrocks: also, as mentioned last week, I will not be around on Monday sadly - I might be able to appear afternoon, but I will be away for the whole morning [17:31] popey: yes, click implements the packagekit backend in question here [17:31] sweet, thanks [17:31] bfiller, yeah, and I'm also doing camera [17:33] sil2100: great! [17:33] sil2100: yeah, enjoy :) [17:33] boiko, yeah, exactly [17:33] robru: nice! thanks [17:34] boiko, rebuild is for new commits on the same MPs, reconfig is for when you add or remove MPs, and you only need landing team to do the reconfig if you want to add a whole new project to the silo. [17:34] boiko, you're welcome [17:36] [6~/wg 23 [17:36] argh [17:37] Saviq, all the tests are failing. There's probably an upgradable package combination. I'll try creating a new container to resolve the problem [17:38] s/upgradable/un-upgradable/ [17:38] sil2100: robru: landing-005 tested, please publish (I understood it's ok to publish, but since I'm the lander let's pretend I'm not in the landing team) [17:38] also, I'm done and ready to fall asleep [17:38] Mirv, on it! [17:38] robru: so, I clicked reconfigure, and clicked build in the jenkins page [17:39] robru: but got this error: http://162.213.34.102/job/landing-011-0-reconfigure/3/console [17:39] Mirv, you wanna do the packaging ack for me ;-) [17:39] boiko: ERROR:root:You need at least specify a MP or a source [17:40] boiko, so you have to manually copy&paste the MP list from the spreadsheet into the reconfigure job [17:40] boiko, see here: http://162.213.34.102/job/landing-011-0-reconfigure/3/parameters/ your MP list is empty [17:40] robru: the UITK was proposed by a core dev, appmenu-qt5 was agreed by did_rocks for sil [17:40] robru: aaaahhhh, I thought it would be automatically, ok [17:40] fginther, thankls [17:40] boiko, yeah, that would be nice ;-) [17:40] hi there CI people! does anyone have an idea how long it should take for a build-wait to resolve inside one of the silo PPAs? [17:40] robru: so you can click ack [17:41] Mirv, ok, done! [17:41] robru: thanks! [17:41] boiko: robru: I think the first one who finds a jenkins plugin to pass those parameters reliably in will win a beer! [17:41] didrocks: lol [17:41] (I have one package dependent on a new version of another, both inside the PPA) [17:41] didrocks, jenkins plugin? just have the jenkins python script query the google doc ;-) [17:41] boiko: really, I looked for that :/ [17:41] pete-woods: ~20-30 minutes in average by experience [17:41] didrocks: thanks :) [17:42] robru: it's hard as well, I need to send an uid to identify the line [17:42] robru: so same issue :p [17:42] didrocks, fair. what we really need is a webapp frontend, not hacking a spreadsheet beyond what it was meant to do [17:42] robru: yep === mhr3__ is now known as mhr3 [17:43] robru: but that won't change the jenkins issue [17:43] you need to pass something to the jenkins job [17:43] pete-woods: the cron job runs every 30 minutes: [17:43] alphecca.canonical.com-lp_buildd:25,55 * * * * /srv/launchpad.net/codelines/current/cronscripts/buildd-retry-depwait.py -q --log-file=INFO:/srv/launchpad.net/production-logs/buildd-retry-depwait.log [17:43] pete-woods: anyone in the team that owns the PPA can manually retry, too [17:43] didrocks, yeah but if the info was stored in JSON somewhere, it'd be way easier for the jenkins job to just query that, without needing any special API for accessing a google spreadsheet [17:43] cjohnston: thanks for the insider information :) [17:43] cjwatson: whoops, wrong person [17:44] that's in lp:lp-production-crontabs FWIW [17:44] robru: the question is rather "what request are you configuring?" [17:44] which I think anyone in Canonical can see [17:44] robru: and for that, there is no way around, you need to pass something to the jenkins job [17:45] didrocks, but the jenkins job knows what silo it's configuring. if the information was stored in json and displayed in a web UI instead of in a google spreadsheet, it would be trivial for the jenkins job to just query the JSON by silo to see what MPs are supposed to be there. [17:45] robru: you can have races if multiple configurations happens at the same time [17:46] didrocks, so make a lock for it to only run once at a time ;-) [17:46] robru: yeah, just telling, it's not that trivial :p [17:53] balloons, popey, I've triggerd rebuilds for konsole-qml-plugin and nemo-qml-plugin-folderlistmodel [17:53] fginther: thanks. [17:53] balloons, popey, those are the only two plugins I see that are still being built by the core-apps-jenkins [17:53] balloons, popey, please let me know if I missed one [17:53] fginther: as an aside, what's the IP for s-jenkins? I have never been able to get the VPN + DNS working. [17:54] i think thats all fginther, terminal, music and filemanager use those two. [17:54] popey, 10.98.3.13 [17:54] reminders will need a rebuild but that's not in the default image [17:54] thanks [17:54] ty fginther [17:59] robru: so, now it says that some revisions are not in the package branch, I guess that's because merge and clean was not done on landing-006 for the Qt 5.2 stuff [17:59] boiko, show me the error [18:00] robru: http://162.213.34.102/job/landing-011-1-build/45/console [18:00] robru: and checking the package changelog on distro, it says it was copied from landing-006 PPA [18:01] didrocks: so what is happening with qt5.2? [18:01] boiko, right, yeah, seems so. in that case you should get the changelog from distro and add it to your MP [18:01] davmor2: well, read the ML + #ubuntu-touch [18:02] robru: wouldn't it be more correct to do the actual merge and clean those branches? other branches might suffer from the same problem [18:03] didrocks: so not yet then :'( === alan_g is now known as alan_g|EOW [18:03] … [18:03] davmor2, once the apps landed [18:03] boiko, normally I would say yes, but qt5.2 was a huge disaster and broke a lot of stuff (citrain couldn't handle it in some ways). so for now we have to just deal with it on a case by case basis :-/ [18:03] robru: got it, no problems [18:04] boiko, actually I checked the changelog, it's just a no-change rebuild, so in this case you can just force rebuild [18:04] boiko, so if you check FORCE_REBUILD it will make the release anyway, even with the missing changelog. [18:05] robru: well, I already asked renato to add the changelog entry anyways, thanks [18:05] boiko, oh ok. [18:07] guys, "call back" in a missed call in the indicator crashes the telephony indicator backend, anyone saw that before? [18:08] bfiller, ↑ [18:09] Saviq: no have not seen that before [18:12] rsalveti: sergiusens: do you need anything from me now? [18:12] bfiller, looks like when the url has escaped + (i.e. tel:///%2B48508xxxxxx), url-dispatcher fails on it [18:12] tedg, ↑ [18:12] * tedg avoids unicode [18:12] didrocks, time [18:12] :-) [18:12] tedg, ** (process:26240): WARNING **: Unable to dispatch url 'tel:///%2B48508xxxxxx':GDBus.Error:com.canonical.URLDispatcher.BadURL: URL 'tel:///%2B48508551786' is not handleable by the URL Dispatcher [18:13] sergiusens: :p [18:13] sil2100: around? emssaging-app seems stuck in proposed [18:13] Saviq, Yeah, the old URL dispatcher only allowed certain characters in tel URLs. The new one is less restrictive. [18:13] Saviq: the indicator shouldn't crash though, let me see [18:13] boiko, it might not be crashing, it's a .crash report [18:14] ah ok [18:14] tedg, when's "new one" coming? [18:14] Saviq, If you would have asked me in January (when I wrote the code) I would have said soon. I have no clue when something will land anymore. [18:15] Saviq, Here's the regex for tel urls: http://bazaar.launchpad.net/~indicator-applet-developers/url-dispatcher/trunk.14.04/view/head:/service/dispatcher.c#L415 [18:15] tedg, well, we've landed 5.2 now :D [18:15] sil2100: seems not around, I'll remove the binaries due to that === Ursinha is now known as Ursinha-afk [18:15] as it's again the sdk [18:15] didrocks, wait === Ursinha-afk is now known as Ursinha [18:15] seb128: it's qtdeclarative5-ubuntu-ui-toolkit-plugin not being available, right? [18:16] yeah, just time ;-) [18:16] boiko, tedg, bfiller, bug #1292657 then [18:16] bug 1292657 in url-dispatcher (Ubuntu) ""Call back" in indicator fails the indicator service with "+"-prefixed numbers" [Undecided,New] https://launchpad.net/bugs/1292657 [18:16] Saviq: thanks [18:17] not sure if the dialer app will grok the escaped url, so just in case [18:19] I don't think you need to escape "+" in a URL path, no? [18:20] tedg, well, + is often used to represent spaces, so if you want a literaly +, you should really escape it [18:20] yeah [18:20] * tedg is confused [18:20] Don't you have to escape space? [18:20] tedg, you do [18:20] Then how does + represent space? [18:20] tedg, but it's often used as "?q=a+b+c" [18:20] tedg, spaces can be escaped as %20, but they are also often escaped as a + [18:21] "Within the query string, the plus sign is reserved as shorthand notation for a space." [18:21] tedg, real world example: https://www.google.com/?q=foo+bar note how it puts "foo bar" in the search bar [18:21] according to http://www.w3.org/Addressing/URL/4_URI_Recommentations.html [18:22] Sure, but that's a specific format for Google, not a standard. [18:22] tedg, ↑↑ [18:22] yay for that, too :) [18:23] it's on w3.org, it must be a standard ;D [18:23] That's only in a query string. [18:23] tedg, agreed, but will you split it to not encode it in the path? [18:24] tedg, I could imagine tel:///Michał Sawicz/, which should be escaped, should it not === Ursinha is now known as Ursinha-afk [18:25] Saviq, Sure, but some lib that telephony service is using is doing this. [18:25] tedg, sure, I'm just saying that it should be "allowed" to encode it [18:25] tedg, well, the url escaping library in python offers "quote" and "quote_plus", which replaces spaces with + signs. so it's a pretty widespread standard. === gatox_ultra is now known as gatox [18:27] Saviq, Eh, okay. In the end I don't care because we're not looking at the path anymore in the next version :-) [18:27] Seems odd that I library would do that though. [18:27] that a [18:29] ok o/ [18:29] tty Monday [18:30] bfiller: I unblocked messaging-app FYI [18:30] didrocks: should I publish? [18:31] bfiller: it was already published, but stuck in proposed === Ursinha-afk is now known as Ursinha === gatox_ultra is now known as gatox [18:37] sergiusens: any idea why I'm seeing this build error trying to build gallery in train? http://162.213.34.102/job/landing-015-1-build/16/console [18:37] hoping not click related [18:38] something with can't find the source tarball [18:39] sergiusens: nm, I see the failure of tests [18:41] bfiller, yeah; the two line click thing we did was a version bump, nothing more === bfiller is now known as bfiller_afk === gatox_ultra is now known as gatox [18:52] popey, mir took most of the executors; so I'm starving (as are the builds), will bbs [18:52] sergiusens: kk [19:02] cjwatson, hmmm, just wondering about qtcreator-plugin-cmake seems to be stuck in -proposed due to arm64, powerpc, and ppc64el. You mentioned something about allowing those failures through for now... this is the last piece of qt5.2 and then we can kick an image build, mind taking a look? [19:02] balloons, https://code.launchpad.net/~sergiusens/music-app/version_fix/+merge/211123 [19:03] robru, thats interesting I go a failure on that during upgrade but -f install fixed it or so I thought [19:04] pmcgowan, must be unrelated, unless you had -proposed enabled... [19:04] robru, quite a coincidence, I now have 3.0.1-0ubuntu2.0 from universe [19:06] pmcgowan, yeah, i'm talking about -0ubuntu4 stuck in proposed [19:06] hah sergiusens I did that twice last night with calc. Can we please fix the store silliness? [19:07] balloons: No!! Silliness is good for you ;) Store issue you're allowed to have fixed :) [19:08] I should have known you were behind this davmor2 [19:08] balloons, just be extra careful as you promised me you would be until the store is fixed ;-) [19:08] It wasn't me [19:09] sergiusens, :-) extra careful means let someone else trip up right? [19:09] anyways, you are approved [19:13] balloons, fwiw I thought the update mentioned on the mailing list would do this for you.. seems not [19:16] fginther, E: File /var/cache/pbuilder/trusty-armhf.tgz does not exist http://s-jenkins.ubuntu-ci:8080/view/click/job/terminal-app-click/42/console [19:16] fginther, ah nvm [19:17] robru: we're not ready to kick an image... [19:17] just to be clear, based on your "last piece of qt5.2 and then we can kick an image build" [19:17] popey, oops, yeah. I meant "the last silo to be published for the image". [19:17] cool [19:18] Preparing to unpack .../android-tools-adbd_4.2.2+git20130218-3ubuntu22_armhf.deb ... [19:18] alan@deep-thought:~$ [19:18] thats annoying [19:18] alan@deep-thought:~$ adb shell [19:18] error: device not found [19:18] afaik rsalveti holds the virtual keys for image kicking (as he has a eagle eye view on what's missing still) [19:18] popey, ugh [19:19] ogra_, postinst on upgrade needs an adb kill-server [19:19] * popey reboots and hopes for the best [19:19] popey, just adb kill-server [19:19] too late ☻ [19:38] * robru -> lunch [19:49] sergiusens: hey [19:49] popey: yeah, that's really annoying [19:50] I'd really need to shut down computer... [19:50] yes, yes you do. [19:51] popey: what are we missing still? [19:51] rsalveti: autopilot testing gallery & music, currently doing notes [19:51] cool [19:51] camera passed okay [19:52] do you know if we already have the proper target at our store? [19:52] I know beuno was working on that [19:52] we do [19:52] oh, hang on [19:52] the proper proper one? [19:52] 14.04 [19:52] i only see our interim one [19:52] ubuntu-sdk-14.04-dev1 [19:53] yeah, that's the one [19:53] not the ones cjw said we really should have [19:53] cool, so we're at least able to upload stuff [19:53] yes [19:53] all uploaded, just pending ap tests [19:53] wow [19:53] such nice people [19:53] much testing [19:53] still need to get my dodge weather app approved [19:54] notes passed [19:54] sergiusens: you need to upload notes to the store ☻ [19:54] great [19:54] i grabbed the click from jenkins. [19:55] actually, i can.. [19:55] hah, i can't [19:55] You just tried to access a feature which you don't have permission to use. [19:55] Only the owner of an application can do that. [19:55] separation of roles ftw [19:55] hm, qtcreator-plugin-cmake is still in proposed [19:56] not sure if that will cause us any issues [19:56] hahah [19:58] are you guys aware that the gallery-app is broken? [19:58] define broken [19:58] popey: does not start. [19:58] just grey screen [19:58] after rebuilding it? [19:59] as if only the app backcolor is drawn but nothing else [19:59] what version? [19:59] Wellark: on the current image? [19:59] * popey is testing com.ubuntu.gallery_2.9.1.925_armhf.click here on qt 5.2 [19:59] and yes, it's broken [19:59] 0.0.67+14.04.20140307-0ubuntu1 [19:59] gallery is a click now, not a deb [19:59] robru: do you know if we really need latest qtcreator-plugin-cmake? [19:59] ok. [20:00] robru: if we so we better ping someone in #ubuntu-release [20:00] right. [20:00] sergiusens: gallery is indeed broken. [20:00] robru: or infinity if he is around [20:00] popey: which dir should I look? [20:01] davmor2: where can I get the image number? [20:02] can't remember anymore... [20:02] on the device that is [20:02] /etc/something... [20:02] Wellark: setting→about phone→ [20:02] system-image-cli --info [20:02] Wellark: build [20:02] sorry hit return [20:03] http://paste.ubuntu.com/7092064/ gallery app log [20:03] current build number: 237 [20:03] device name: mako [20:03] channel: devel-proposed [20:03] did apt-get dist-upgrade on it [20:04] file:///opt/click.ubuntu.com/com.ubuntu.gallery/rc/qml/GalleryApplication.qml: File not found [20:04] popey, that is concerning [20:04] not only that it can't find the file [20:04] its missing "current" [20:04] but the click path doesn't include the version [20:04] my /opt/click.ubuntu.com/ is empty [20:04] yeah [20:05] Wellark: you're using a deb probably, i'm testing a click [20:05] when did gallery migrate to click? [20:05] popey, do you know when we might have an image built with 5.2? [20:05] Package 'gallery-app' is not installed, so not removed [20:05] popey, last week i think [20:06] kenvandine: when we get these issues fixed [20:06] so, dive in ☻ [20:06] * kenvandine needs to test content-hub stuff.. but it all deps on qt5.2 now [20:06] i guess i need to dist-upgrade :) [20:07] kenvandine: no the qt5.2 image can't roll till all the issues are fixed the gallery app is one that was supposedly fixed by rebuilding against qt 5.2 [20:07] bfiller_afk: ^^ [20:07] popey, what do you have in /usr/share/click/preinstalled/com.ubuntu.gallery [20:07] root@ubuntu-phablet:/# ls /usr/share/click/preinstalled/com.ubuntu.gallery [20:07] 2.9.1.922 current [20:07] ok, me too [20:08] i have .923 in /opt [20:08] as current [20:08] but i built that myself :) [20:08] popey: odd question but the manifest is right for build and security right? Just double checking that aa isn't preventing the file from existing :) [20:09] popey, clearly the click package that is installed is broken, just not sure where that bug lies [20:09] davmor2, the path is wrong there [20:09] no version in the path [20:10] popey, do you have gallery in /opt at all? [20:10] yes [20:10] with version and current? [20:10] root@ubuntu-phablet:/opt/click.ubuntu.com/com.ubuntu.gallery# ls -l [20:10] total 8 [20:10] drwxr-xr-x 5 clickpkg clickpkg 4096 Mar 12 17:56 2.9.1.922 [20:10] drwxr-xr-x 5 clickpkg clickpkg 4096 Mar 14 19:56 2.9.1.925 [20:10] lrwxrwxrwx 1 clickpkg clickpkg 9 Mar 14 19:56 current -> 2.9.1.925 [20:11] and what's in the .desktop file? [20:11] in .local/share/applications [20:11] not the one in /opt [20:11] http://paste.ubuntu.com/7092099/ [20:11] specifically the past [20:11] path [20:11] Path=/opt/click.ubuntu.com/.click/users/phablet/com.ubuntu.gallery [20:12] ah [20:12] and that should be a link to the current version too [20:12] what does that path point to? [20:12] you sure? [20:12] others dont and they work [20:13] maybe that error message is misleading [20:13] http://paste.ubuntu.com/7092106/ camera for example [20:13] right... that's a link to the current version [20:13] lrwxrwxrwx 1 phablet phablet 50 Mar 14 19:57 /opt/click.ubuntu.com/.click/users/phablet/com.ubuntu.gallery -> /opt/click.ubuntu.com/com.ubuntu.gallery/2.9.1.925 [20:14] although the error message shows clearly the path isn't in .click [20:14] robru: we need your help [20:14] our lander is on holiday [20:14] grep DENIED /var/log/syslog |grep gallery [20:14] ooh [20:14] and our spare-lander decided to EOD [20:14] http://paste.ubuntu.com/7092112/ [20:14] robru: could you help me and pete-woods out [20:15] popey, so that's old [20:15] robru: if you could be kind enough to mark silo 2 as testing done [20:15] popey, dunno then :/ [20:15] yeah [20:15] robru: and ready for landing [20:15] balls [20:15] file:///opt/click.ubuntu.com/com.ubuntu.gallery/rc/qml/GalleryApplication.qml: File not found [20:16] is clearly concerning... [20:16] yeah [20:16] no idea where it's getting that path from [20:16] should this be on the latest image? [20:16] 237? [20:16] Wellark: what image do you have? [20:16] 237 [20:16] and what version of gallery? [20:16] popey, where did you get that version of gallery? [20:16] kenvandine: from sergiusens ☻ [20:17] its built in jenkins against qt5.2 [20:17] again, I'm saying that my /opt/click.ubuntu.com/ is empty [20:17] ok... maybe that is what's broken :) [20:17] and gallery-app deb is not installed [20:17] Wellark: I didnt ask that [20:17] Wellark, that is only because you haven't updated any of the preinstall clicks [20:17] click list [20:17] I have no idea where to check the version of a preinstalled click [20:17] Wellark, ^^ [20:17] ls /usr/share/click/preinstalled [20:17] or look in /usr/share/click/preinstalled [20:17] ok, got it [20:17] com.ubuntu.gallery2.9.1.922 [20:18] same as my preinstalled [20:18] i have .923 and it works [20:18] with image 237 [20:18] works for me on 237 [20:18] but that was my own rebuild i think [20:18] ok. you guys seem to have it under control then. [20:19] popey, so are you saying if i dist-upgrade my phone i won't have a working gallery? [20:19] gallery is one of the things i need to test with my content-hub branches :) [20:19] yes [20:19] sigh [20:19] popey, I just did a build with what's in the archives [20:19] and all my branches are built against 5.2 :-D [20:19] oh joy [20:19] sergiusens: seems busted [20:20] popey, and you installed it the right way right... with pkcon? [20:20] my build is from yesterday, just ran the click build script and installed [20:20] Wellark, hey sorry, i'm just on lunch. what's going on? is testing really done? sounds like there's a problem. [20:20] of course different qt [20:20] robru: not our component [20:21] but can't imagine that is causing the weird path failure [20:21] robru: I just reported something I noticed [20:21] Wellark, also, it can't land until after the next image build, so there's no rush either way [20:21] robru: if you could just mark the silo as testing done [20:21] Wellark, sure [20:21] popey, fwiw did camera work? [20:22] robru: thanks! [20:22] robru: cheers! [20:23] sergiusens: camera has some weird things going on.. the image starts to shift to left and wraps around from the left corner to the right [20:23] Wellark, the click? [20:23] sergiusens: I also see this [20:23] sergiusens: the actual camera feed [20:23] it shifts and glitches [20:23] Wellark, pete-woods again... the newly built camera from the click store? [20:24] Wellark, pete-woods : you're welcome! [20:24] sergiusens: this is on image 237 and version is 2.9.1.250 [20:24] sergiusens: all I did was dist-upgrade from the build 237 [20:24] I don't know if click store has never version [20:24] and just FYI [20:24] now my n4 keeps rebooting it self randomly [20:24] Wellark, you need 2.9.1.250 for the qt 5.2 rebuild [20:24] 2.9.1.253! [20:24] sergiusens: ok. just saying waht I'm seeing [20:25] but that's under control too [20:25] ok. [20:25] Wellark, try updating it [20:25] good. [20:25] sergiusens: I need to EOD [20:25] Wellark, from the updae manager [20:25] Wellark, ah, ok; [20:25] but clearly I have an outdated version [20:25] so all good [20:25] sergiusens: yes [20:25] popey, https://myapps.developer.ubuntu.com/dev/click-apps/118/ [20:25] that's notes [20:25] sergiusens: notes worked [20:26] have you updated it again? [20:26] but seriously.. I see n4 reboots like every 30 minutes or so. oh,well.. I will just wait for the next image. there is probably just breakage as I did dist-upgrade on top of 237 [20:26] popey, yes [20:26] Wellark: maybe you should have mentioned that earlier [20:26] yes, you broke it by dist-upgrading [20:27] popey, terminal: https://myapps.developer.ubuntu.com/dev/click-apps/157/ [20:27] popey: 22:03 < Wellark> did apt-get dist-upgrade on it [20:27] I did tell that :) [20:28] Wellark: very sorry, missed that [20:28] np. [20:28] popey, music https://myapps.developer.ubuntu.com/dev/click-apps/143/ [20:29] popey, going to see what's wrong with gallery now [20:29] thanks [20:30] bfiller_afk, on line 47 you mention fixes for qt5.2, I want to assign a silo for that, but 'ready' is not marked as 'yes'. Will it be ready soon? [20:30] bfiller_afk, or is that click-only? [20:32] kenvandine, popey http://paste.ubuntu.com/7092185/ [20:32] ralsina, gave you silo 7 for line 11. [20:32] there is no dbus-launch on the image [20:33] I think xnox removed it; are the hooks depending on it? [20:33] rsalveti, ^ [20:33] hah [20:33] nick bfiller [20:33] robru: I am not handling those landings anymore, can you reassign to thostr? [20:33] ralsina, sure [20:33] robru: thanks [20:33] ralsina, you're welcome === bfiller_afk is now known as bfiller [20:34] robru: those were merged manually as they were needed to generate the click for the image, I'll make another MR to release trunk cause we need the deb as well [20:37] sergiusens: but that might just be a warning [20:38] sergiusens: what are we still missing then? gallery-app and notes-app? [20:38] gallery broken, notes, terminal, music to-do [20:39] popey: so not landing tonight then? is it safe to call it a night? [20:39] not my call ☻ [20:39] thought sergiusens just did music and terminal [20:39] gave them to me to test, yes [20:39] I'd like to be able to build an image today if possible, but you don't need to kill yourself :-) [20:39] i can only test one at a time [20:40] bfiller, oh alright [20:40] * popey is eating a spicy pizza, so killing myself very slowly [20:40] popey: oh, I can access the links he gave, so I thought it was published already [20:40] or are you just validating whatever is already published? [20:41] * davmor2 hears popey 's arteries hardening from here [20:41] process is: he uploads, I download, ap test, app rove [20:41] you didnt click app-rove did you rsalveti ? [20:41] * rsalveti is trying to find out if he can help with something [20:41] nops [20:41] cool [20:41] I don't even have the power to do so [20:41] notes passed last time, expect it to pass again [20:41] then i just have terminal and music to do [20:42] awesome [20:42] wont take long, maybe 30 mins if it all works [20:42] I believe sergiusens is already testing gallery [20:42] ok, notes passed [20:43] great [20:43] terminal now [20:43] Wellark: when stuff is broken this way it's just better to give up lol [20:44] wait a new image to be published and flash with bootstrap [20:44] i recommend pizza [20:44] pizza is always a good idea [20:45] popey: where can I find latest click for gallery? [20:45] the broken one? [20:45] http://popey.com/~alan/com.ubuntu.gallery_2.9.1.925_armhf.click [20:45] not sure if broken, but the latest :-) [20:45] thats the one I tested [20:45] great [20:46] rsalveti: I blame popey (TM) [20:46] :-) [20:47] i will be delighted if it turns out I messed up the test and gallery is fine [20:48] terminal passed [20:48] music next [20:48] go popey go popey [20:52] rsalveti, no, gallery is busted [20:52] pmcgowan: no it's go go gadget popey you were close though :) [20:52] my mistake [20:53] rsalveti, you can track the gallery problem if you want as well [20:53] others seem easy [20:53] gallery is a beast [20:53] pmcgowan: there are a few people saying that unity on the desktop is broken today after a dist-upgrade :( [20:53] rsalveti, http://s-jenkins.ubuntu-ci:8080/view/click/job/gallery-app-click/lastSuccessfulBuild/artifact/out/com.ubuntu.gallery_2.9.1.925_armhf.click [20:53] davmor2, how so? [20:54] sergiusens: do you have any idea why? [20:54] rsalveti, also QObject::connect: Cannot connect GalleryApplication::mediaLoaded() to (null)::onLoaded() [20:54] rsalveti, going to see if some fiddling with gallery happened... [20:54] pmcgowan: https://bugs.launchpad.net/ubuntu/+source/unity/+bug/1292396 [20:54] Ubuntu bug 1292396 in unity (Ubuntu) "After upgrade Unity session is broken" [Undecided,New] [20:55] davmor2, unless I am mistaken, he upgraded before the packages arrived [20:55] it seems so [20:56] rsalveti, popey so nothing changed in gallery at all [20:56] davmor2, not our packages! he says he upgraded yesterday and filed the bug today europe morning [20:57] he may have gotten a partial push, it happens [20:57] sergiusens: let me try to rebuild the deb instead and see [20:58] pmcgowan: ah fair enough :) [20:59] davmor2, fwiw I updated my other laptop and its fine (but not this one, I'm no fool) [20:59] hm [20:59] rsalveti, just use the existing deb [21:00] jeez, the music autopilot tests take ages [21:00] * popey makes another note to replace the songs in the music app autopilot tests again [21:00] popey, yeah, we need to move away from autopilot for most of the tests that are run [21:00] also.. [21:01] we don't do any autopilot runs of these apps until *I* manually run them before they hit the store.. or.. it's too late and the image is baked [21:01] surprised that all AP tests in jenkins only run on desktops, not actual phones/tablets [21:01] popey, there's bugs open for that since october [21:02] what? to run on real devices in the DC? [21:02] or some other solution? [21:04] rsalveti, the deb works fwiw [21:04] sergiusens: interesting [21:04] still updating here [21:04] don't see any apparmor blocks [21:05] sergiusens: I still blame popey (TM) [21:05] Ran 13 tests in 824.278s [21:05] FAILED (failures=3) [21:05] gnnnnnn [21:06] * popey reboots and runs again [21:06] popey: did the phone lock up? [21:06] no [21:06] ran all the way to the end [21:06] :( [21:06] going to run it again [21:06] SQLite error: QSqlError(-1, "Unable to fetch row", "No query") [21:06] might be that [21:06] confirmed it's no app apparmor [21:07] brb, going to get beer, back in 824s [21:09] robru: line 47 ready for silo [21:15] rsalveti, for some reason the gallery can't populate the database it creates [21:17] * popey returns [21:19] davmor2: volume work on mako? [21:19] trying to turn down the volume on bkerensa and it's not working [21:19] davmor2: oh, nvm you're not on qt5.2? [21:20] sergiusens: maybe it's not compatible somehow with the previous database? [21:20] would be quite weird anyway [21:20] popey: it does on 5.0.2 if that helps :) [21:21] ☻ [21:21] rsalveti, nah, I wiped it to test; it creates it; but then spits out that error [21:21] Ran 13 tests in 821.592s [21:21] FAILED (failures=3) [21:21] bah! [21:21] popey: stop breaking them [21:21] sergiusens, are you still see that connect error? [21:21] yeah, volume up&down with the hw buttons are not working [21:21] http://paste.ubuntu.com/7092412/ eyeballs welcome [21:22] rsalveti, I wonder if I built against the wrong deps while it was still in flight [21:22] * popey pokes balloons with the above pastebin [21:22] pmcgowan, which one? [21:22] sergiusens, the one you pasted above [21:22] sergiusens: yeah [21:22] wondering if it doesnt get the loaded signal it wont do shit [21:23] sergiusens: rebuilding it here to see [21:23] +1 [21:23] * balloons looks [21:23] pmcgowan, right; [21:23] ok, who broke music? [21:23] :-) [21:24] balloons, was music ever tested with qt5.2? [21:24] balloons: for context that's image 237 dist-upgraded so it has 5.2 [21:24] it was.. I've got 237 on my devices too [21:24] davmor2: what level of testing did you do for core apps on 5.2? [21:24] I'm seeing calendar issues with 5.2 [21:24] was it shakedown test or full AP? [21:24] balloons, 237 alone does not have qt 5.2 [21:25] sergiusens, I'm not seeing on things without 5.2.. [21:25] yeah, I believe it should be good to build a new image even without having everything fixed [21:25] so for instance my flo is 237 only and works [21:25] rsalveti, thank you [21:25] rsalveti, that's the idea of devel proposed :-) [21:26] sergiusens: we just wanted to make sure that some of the stuff was rebuilt at least [21:26] rsalveti: gimmie 10 mins to test music app. manually [21:26] popey: sure [21:26] will give ~1h at least [21:26] ok, cool [21:26] I'll dance my ass off to some fat choons testing music app [21:26] rsalveti, yeah, let popey approve and have those landed in the click list (so they are pulled in the image) [21:26] yeah [21:26] popey: shakedown and ap but the ap was fairly early on and iirc there were only a few apps that failed mostly due to the app not opening. Terminal gallery and then there were about 5 other fails iirc. But since then it has all been manually testing stuff [21:27] ok [21:27] ta [21:27] sergiusens: just rebuilt it and was able to open it at least [21:27] davmor2, are you on the latest terminal? [21:28] rsalveti, built as click? [21:28] sergiusens: i tested and approved terminal [21:28] rsalveti, different options in cmake [21:28] sergiusens: deb [21:28] let me take a few pictures [21:28] rsalveti, yeah, installing the deb also works ;-) [21:29] oh, thought you also had issues with it [21:29] sergiusens: how to build click natively? [21:31] phablet 1988 3.4 0.8 106464 15860 ? Ssl 21:29 0:02 mediascanner-service [21:31] phablet 2002 17.9 0.4 66624 8348 ? Ssl 21:29 0:12 mediascanner-service-2.0 [21:31] uh [21:31] that doesn't look right [21:32] why does mediascanner2 not have a conflicts or replaces? [21:32] oh, there's a click-build.sh in there [21:32] popey: yeah, it's "fine" [21:32] both are installed since weeks already [21:32] wat [21:32] guess waiting new scopes [21:32] ok [21:33] rsalveti, nah, don't use it [21:33] i see no music in the dash [21:33] rsalveti, or use it, but on armhf [21:33] sergiusens: haha, what should I use it then? [21:33] sergiusens: yeah, building natively [21:34] phablet@ubuntu-phablet:~/Music$ find . | wc -l [21:34] 414 [21:34] rsalveti, gallery is a mess [21:34] haha [21:34] rsalveti, can you access http://s-jenkins.ubuntu-ci:8080/view/click/job/gallery-app-click/configure [21:34] rsalveti, if you look at the cmake definitions you'd go out and cry [21:35] guess I don't have access [21:35] bfiller, whooops, asleep at the wheel, sorry. got you silo 16, please build [21:35] rsalveti, basically you just want cmake ../trunk_dir -DINSTALL_TESTS=off -DCLICK_MODE=on -DBZR_REVNO=$bzr_revno [21:35] make [21:35] great [21:36] make install DESTDIR=tmpdir [21:36] click build tmpdir [21:36] so that script should be fine [21:36] rsalveti, great [21:36] rsalveti, and here's the hack if in a pristine env mk-build-deps -i trunk_dir/debian/control [21:37] rsalveti, makes me wonder if we need to add a plugin to the seeds [21:37] rsalveti, nah, shouldn't be the case [21:37] yeah, otherwise deb would complain as well [21:37] right guys I'm calling it a night. I'll keep my phone with me and check it periodically for a new build if there is one I'll test it if there isn't I wont :) Catch you all Monday other wise night all [21:38] thanks davmor2 [21:38] thanks, enjoy [21:38] have a great weekend [21:40] balloons: having difficulty breaking music app here in normal use. [21:41] balloons: getting dangerously close to approving it into the store with failing tests ㋛ [21:41] yeah [21:41] rsalveti, pmcgowan so onLoaded is defined in file:///opt/click.ubuntu.com/com.ubuntu.gallery/rc/qml/GalleryApplication.qml which can't be found because the path is wrong [21:41] better than a completely broken one [21:42] sergiusens: hm [21:42] it's supposed to be file:///opt/click.ubuntu.com/com.ubuntu.gallery/${version}/rc/qml/GalleryApplication.qml [21:42] maybe galley would just work if installed during fist boot? [21:42] rsalveti, maybe, I can try and preinstall it ;-) [21:42] sergiusens: yeah [21:43] as nothing changed [21:43] rsalveti, but it may be related to click and the way gallery sets it's organization name [21:43] hm, right [21:43] rsalveti, this worked 4 images ago for sure [21:43] sergiusens: ok, got the click [21:43] pkcon install-local *.click? [21:44] rsalveti, as phablet, yes [21:45] m_view->setSource(Resource::getRcUrl("qml/GalleryApplication.qml")); [21:45] it's click or that getting messed up [21:46] yeah, just a white screen [21:46] ** (process:14579): ERROR **: Unable to find keyfile for application 'com.ubuntu.gallery_gallery_2.9.1.922' [21:46] wtf [21:47] rsalveti, didn't you just built 925? [21:47] com.ubuntu.gallery_2.9.1.latest_armhf.click [21:47] which is 925 [21:47] rsalveti, ah, pass in _2.9.1.latest [21:48] that's the actual version... [21:48] SQLite error: QSqlError(-1, "Unable to fetch row", "No query") [21:48] http://paste.ubuntu.com/7092526/ [21:49] yeah, same issue [21:49] file:///opt/click.ubuntu.com/com.ubuntu.gallery/rc/qml/GalleryApplication.qml: File not found [21:49] rsalveti, yeah, and above the signal problem, right? [21:51] sergiusens: now how is this path translated? [21:52] Resource::getRcUrl [21:52] sergiusens: did you try preinstalling it? [21:52] rsalveti, it's ugly [21:52] rsalveti, config.h.in [21:52] haha [21:52] why everything is so complicated with gallery [21:53] if we had deb still we would be fine already [21:53] rsalveti, Resource::getRcUrl uses galleryDirectory defined there [21:54] return QString(QDir::currentPath() + QDir::separator() + [21:54] was it ever tested without being pre-installed? [21:57] rsalveti, yes it was [21:57] rsalveti, I'm guessing the latest click updates did this [21:59] rsalveti, let me try something [22:11] rsalveti, no other reason why QDir::currentPath would return the wrong path [22:11] rsalveti, anyways I'm making it relative to test [22:11] ok [22:13] rsalveti, do you have a build still? [22:13] rsalveti, if so, try and beat me building with http://paste.ubuntu.com/7092626/ [22:32] sergiusens: rsalveti hows it going? [22:36] popey, installing and dist upgrading to be able to build for the past 30' :-/ [22:36] popey, whatdaya know; it just finished [22:37] popey, I have a hard stop when my guests arrive to start the fire for some bbq :-P [22:38] should we have rebuilt cordova-ubuntu-2.8 ? [22:38] http://ubuntuone.com/0YOQ3KMcg0WLXWQAmuKHVL [22:38] its built against libqtcore5 [22:38] libqt5core5 (>= 5.0.2) [22:39] yes then [22:39] popey, wrt rsalveti can probably just push something [22:39] thats going to screw developers over the weekend [22:39] http://paste.ubuntu.com/7092713/ [22:39] wants to yank the sdk [22:39] hello nik90 [22:40] okay so do I postpone my system update to next week? [22:40] i would, but this is less than ideal. [22:40] pmcgowan: ^^ [22:40] considering that it otherwise going to remove my SDK :) [22:40] well indeed [22:41] hey [22:41] yo [22:41] sergiusens: sorry, back now [22:41] rsalveti, no worries [22:42] rsalveti, also; yo can avoid most args by just running click-buddy with no args in the dir you want to build [22:42] for native builds it just works [22:42] xcompile of gallery has some weird qmake calls inside still [22:42] right [22:42] sergiusens: still building in there? [22:42] rsalveti, I just finished dist-upgrading my builder [22:43] rsalveti, soooo slow [22:43] yeah [22:43] triggering a new build then [22:43] build is 50% done [22:43] rsalveti, but I think my patch should work [22:43] as the workdir when running a click app has to be the clickdir itself [22:44] as everything is relative [22:44] sergiusens: indeed [22:44] popey: did we get the rest of the apps published already? [22:44] rsalveti, not sure why currentPath would give a bogus path though [22:44] yeah, we can debug on monday [22:44] rsalveti: except music, yes [22:44] rsalveti, maybe the new qt doesn't follow symlinks? [22:44] popey: can't we also publish music? [22:45] popey, and gallery I hope :-) [22:45] popey: better than having one that is completely broken [22:45] well, i figured you guys knew that one [22:45] popey, yeah, but still good to make sure :-P [22:45] sure ☻ [22:45] you've been drinking ;-) [22:45] lol [22:45] so I can at least build a new image [22:45] haha [22:45] and we can debug on a fresh env right on monday morning [22:46] I'm afraid dist-upgrade is not safe enough [22:46] rsalveti, it isn't [22:46] rsalveti, btw; the cordoba html stuff, wouldn't a no rebuild push fix that? [22:47] approved music [22:47] sergiusens: sorry, not sure I know what is wrong with it [22:47] popey: great [22:47] rsalveti, it depends on qt 5.0.2 [22:47] rsalveti, so needs initially at least a rebuild [22:47] right, wouldn't a rebuild fix it? [22:47] righ [22:47] rsalveti: once you have kicked an image can we talk about the issue with some sdk packages not built for 5.2? [22:48] popey: which packages? [22:48] yeah that ^ [22:48] http://paste.ubuntu.com/7092713/ [22:48] once sec, about to test gallery [22:48] on desktop it wants to remove the sdk [22:48] this is a big problem for our developers [22:48] if they dist-upgrade over the weekend they will lose the sdk [22:48] e.g. cordova-ubuntu depends on libqt5core5 (>= 5.0.2) [22:49] and for qt5.2 we now have libqt5core5a, not libqt5core5 [22:49] right, so I guess we just need to rebuild it [22:49] ya [22:49] and possibly others [22:50] ubuntu-html5-container too [22:51] actually looks like just those two that were missed [22:51] sergiusens: failed same way here [22:51] the rest are ppa things [22:51] great, let me trigger a rebuild for those folks [22:52] rsalveti, so the dir we are being changed to is bogus [22:52] cordova-ubuntu was already built against qt 5.2.1 [22:53] but the deps? [22:53] they're wrong. [22:53] rsalveti, ls $(click pkgdir com.ubuntu.gallery)/rc/qml/GalleryApplication.qml [22:53] rsalveti, maybe it's that new libclick? [22:53] doanac`: feels kind of wrong; would say depends on the level of hack we are using [22:53] anyway [22:54] /opt/click.ubuntu.com/.click/users/phablet/com.ubuntu.gallery [22:54] sergiusens: that is correct [22:54] * sergiusens invokes tedg [22:54] ah, not around [22:54] sergiusens: did you try preinstalling it? [22:54] doanac`: if all other autopilot depends on unity8-autopilot (e.g. that thing provides the framework with base classes, then its ok - otherwise we should put it where the autopilot base classes our tests use are living [22:54] rsalveti, not this one, but the previous one I have [22:55] asac: feel free to weigh in here: https://bugs.launchpad.net/bugs/1292585 [22:55] Ubuntu bug 1292585 in Ubuntu CI Services "touch daily image test: needs unity8-autopilot installed for screen unlock" [High,New] [22:55] popey: dependency is fine, it's >= [22:56] * doanac` has to go get eat [22:56] rsalveti: but thats no good [22:56] right, something else is broken [22:56] qt5.2 ships libqt5core5a [22:56] not libqt5core5 [22:56] but that's set as a shlib depends [22:56] hmm [22:56] by the build [22:56] so if you rebuild it should work [22:56] right [22:57] unless someone manually set that dep [22:57] :-/ [22:57] not in cordoba [22:57] cordova [22:57] * sergiusens advertises that he lives in cordoba [22:58] libqt5core5a (>= 5.2.0) [22:58] Depends: qtdeclarative5-cordova-2.8-plugin, geoclue-gypsy, libc6 (>= 2.14), libgcc1 (>= 1:4.1.1), libqt5contacts5, libqt5core5a (>= 5.2.0), libqt5feedback5, libqt5gui5 (>= 5.0.2), libqt5multimedia5 (>= 5.0.2), libqt5network5 (>= 5.0.2), libqt5positioning5, libqt5qml5 (>= 5.0.2), libqt5quick5 (>= 5.0.2), libqt5sensors5 (>= 5.1.1), libqt5systeminfo5, libqt5widgets5 (>= 5.0.2), libstdc++6 (>= 4.5), libxcb1 [22:58] for cordova-ubuntu-2.8 [22:58] libqt5widgets5 (>= 5.0.2) ??? [22:58] was it rebuilt when only qtbase was in? [22:58] although qt5widgest is in the same source package [22:59] seems weird [22:59] libQt5Widgets.so.5.2.1 [22:59] used this guy [22:59] it seems I can install ubuntu-sdk here just fine [23:00] let me do a dist-upgrade and check [23:01] rsalveti: as popey, when I tried a dist-upgrade, it wanted to remove ubuntu-sdk ubuntu-sdk-libs ubuntu-sdk-libs-dev and so on ..which was a bit scary [23:02] rsalveti: however I have the phablet-tools and the core apps ppa installed [23:02] that paste above is from nik90 [23:02] weird [23:02] trusty? [23:02] rsalveti: yes trusty [23:02] http://paste.ubuntu.com/7092820/ is mine [23:02] just updated again [23:02] nik90: update again, i wonder if your mirror was not synced [23:02] mine no longer wants to remove the sdk [23:03] yay [23:03] probably something was missing in release [23:03] and still stuck in proposed [23:03] "Archive skew" ☻ [23:03] i love that phrase [23:03] popey: still get the removal of sdk [23:03] hm, which mirror are you on? [23:03] popey: may be I will switch to the main archive [23:03] instead of the netherlands mirror [23:04] https://launchpad.net/ubuntu/+archivemirrors [23:04] "One day behind" [23:04] http://mirror.sov.uk.goscomb.net is what I'm using which is super zippy [23:05] popey: I switched to the main server and did a update. But still removes the SDK [23:05] hm [23:06] can you ppa-purge the core apps ppa? [23:06] trying now [23:07] same thing :/ [23:07] should I go ahead and purge the ppa none the less and may be start fresh? [23:08] might be a good idea [23:08] as it's fine for me and popey [23:09] sergiusens: any new idea for gallery? [23:09] * nik90 takes a deep breath and does that [23:09] i would apt-get autoremove --purge ubuntu-clock-app music-app ubuntu-filemanager-app [23:09] PWD=/opt/click.ubuntu.com/.click/users/@all/com.ubuntu.gallery [23:09] ls: cannot access /usr/share/click/preinstalled/com.ubuntu.gallery/rc/qml/GalleryApplication.qml: No such file or directory [23:09] that's preinstalled [23:09] no clue how that's being done [23:10] sergiusens: seems it's not working well after colin did all those changes to link into all or users [23:10] rsalveti, I need to look at this later [23:10] sergiusens: right, should we just create a new image now? [23:10] or are we waiting for something else? [23:10] rsalveti, that's old I think; might be the libclick stuff [23:11] I say build now [23:11] yeah [23:11] popey: ^? [23:11] rsalveti, gives me a more relaxed time to look at it instead of on the run [23:11] sergiusens: yeah, need to go as well [23:11] in theory I'd also be off today [23:11] but yeah, just theory [23:11] what does building an image get us? [23:11] rsalveti, need to host a bbq and still have to go and buy the meat! [23:11] popey: a new image? [23:11] we disabled cron right? [23:11] heh, right [23:12] and a weekend of pain for people on -proposed potentially? [23:12] sergiusens: just go :-) [23:12] popey: not necessarily pain [23:12] just gallery is broken [23:12] but it's really hard to debug stuff if we're asking people to run dist-upgrade [23:12] true, my phone dist-upgraded seems okay ☻ [23:12] and getting tons of packages [23:12] rsalveti, enjoy the rest of the day off? [23:12] yeah, I'd punt an image. [23:12] pun! [23:12] great [23:12] sergiusens: swapping it to another day next week [23:13] rsalveti, sounds good; I signed up for pycamp [23:13] :-) [23:13] cool [23:13] deli is about to close; I have to leave [23:13] cheers [23:13] later! [23:14] === Image 238 building === [23:14] nice work sergiusens rsalveti [23:14] nik90: hows it going? [23:14] o/ [23:14] rsalveti: still fetching the packages and now it seems to be installing them [23:14] popey: ^^ [23:15] eh, hang on [23:15] what did you do? [23:15] i thought you were *removing* the ppa? [23:15] I ppa-purged the ppa [23:15] ah [23:15] ok [23:15] which does the installing and removing...I guess a dist-upgrade itself [23:16] what command are you running right now? [23:16] popey: I ran sudo ppa-purge ppa:ubuntu-touch-coreapps-drivers/daily [23:17] ah, thats running now? [23:17] popey: and that recommended removing the core apps which were installed along with some other stuf [23:17] popey: yes that's running now [23:17] I'd expect that to just removing just a few packages [23:17] not lots [23:17] popey: no but that's the thing. When I did the ppa-purge I got http://paste.ubuntu.com/7092713/ [23:17] popey: which is basically installing qt 5.2 and removing 5.0.2 [23:18] along with the SDK I suppose [23:18] ok [23:18] lets see when that finishes if we can reinstall the sdk [23:19] The following extra packages will be installed: [23:19] qtcreator qtcreator-plugin-cmake qtcreator-plugin-ubuntu [23:19] so its odd.. [23:20] haha, another flappy bird clone hits the store [23:20] alrite the ppa purge is complete [23:21] popey: my qtcreator and sdk is gone since I cannot find it in the dahs [23:21] ok [23:21] dash* [23:21] I am going to dist-upgrade now [23:21] sudo apt-get update [23:21] sudo apt-get dist-upgrade [23:21] and paste pls [23:21] popey: http://paste.ubuntu.com/7092874/ [23:22] ok, go for it [23:22] then when it finishes, see what happens with an install of ubuntu-sdk [23:22] i think you may need the sdk ppa [23:22] but lets see [23:23] ok [23:25] popey: so I am going to run sudo apt-get install ubuntu-sdk -> http://paste.ubuntu.com/7092891/ [23:25] go? [23:26] looks good [23:26] * popey switches networks, brb [23:29] popey: http://imgur.com/isuL3Zq ;) [23:29] :D [23:29] popey: so I am running qt 5.2? [23:30] looks that way ☻ [23:30] like the colour scheme you have there [23:30] and the dots for spaces [23:31] thnx :) [23:31] also [23:31] whats going on with your cog icon in the top right? [23:31] * nik90 takes a moment to register the fact that he is running Qt 5.2!!!! [23:32] popey: that's the numix-icon-theme [23:32] nice [23:32] popey: just one caveat...no more global menu for qtc 3.0.1 [23:33] but I guess that will be fixed sooner or later [23:33] wait what, qt 3.0.1? [23:33] ooh [23:33] qtcreator 3.0 [23:34] nice [23:35] right, now you've done it I feel safer to upgrade ㋛ [23:35] popey: lol..cheeky === jasoncwarner___ is now known as jasoncwarner