[00:06] balloons, I may or may not be able to help you [00:06] depending upon the nature of your query [00:07] and timezones don't suck as long as you are in the right one [00:09] retoaded: :-) I appreciate the offer. I'm concerned about what's happening with the clock app runs [00:10] balloons, ahhhh, not really my area but I can take a peek (as long as it's note code related). [00:14] retoaded: yea, I think this might be something to ask tomorrow [00:15] balloons, ack [00:15] retoaded: if you can confirm the runners look fine, I think that's enough. [00:15] balloons, where are the runners located? [00:16] retoaded: http://91.189.93.70:8080/computer/ [00:16] they look fine and run for the other core apps.. sadly it's not a quick fix I thought it was anymore [00:16] retoaded: oO http://91.189.93.70:8080/job/mediumtests-trusty-slave-reboot/ [00:17] thatt's cool, so you have a reboot job :-0 [00:17] balloons, ok, I don't think I have much access to that one. I need to fix that one of these days [00:17] or perhaps it's not what i think it is.. regardless. ty [00:17] no worries. ttyl [00:18] ack [08:01] good morning [08:01] could anybody publish silo 3? [08:42] thostr_: hi [08:42] sil2100: hi, in hangout right now... [08:42] thostr_: so, regarding indicator-network - robru confused everything up ;/ [08:43] thostr_: there was no indicator-network landing because there were no changes to release [08:43] thostr_: so I'm not sure what he was trying to rebuild and release ;/ [08:43] sil2100: mhhh, ok. so what needs to be done now? [08:43] Nothing, I published everything that was needed yesterday [08:43] Not sure what robru wanted to accomplish? [08:46] and robru used overrides as well :/ [08:46] so, everything is screwed, we don't have the metadata that we use (the branch is still in trusty-proposed, but well…) [08:47] Ok, now this is confusing - so in the end what did he do? Did he just re-build the silo, or tried to re-publish it? [08:48] not sure what he tried to do… [08:48] sil2100: just look at job history [08:50] he tried to republish [08:50] I see 'ignore missing' he set [08:50] yeah [08:50] not sure what he tried to do at all… [08:51] thostr_: the packages were migrating, did anything happen/did you ask him to do anything? [08:51] http://162.213.34.102/job/landing-005-2-publish/18/console was pretty clear that it was already published :/ [08:52] at least, the publication published nothing [08:52] didrocks: yes, but there was still a strange error message in the ci sheet [08:52] didrocks: now, I'm really lost there [08:52] http://162.213.34.102/job/landing-005-3-merge-clean/11/console [08:52] thostr_: that was your error ^ [08:53] thostr_: which is legit, it's telling you indicator-network was not built at all, and you set it in the configuration [08:53] thostr_: if you look at the build job, it told you why it didn't build it: [08:53] waht I got yesterday was something like v [08:53] "jettisoned without merging" [08:54] 2014-02-17 18:02:46,091 INFO === Trying to prepare indicator-network === [08:54] … [08:54] 2014-02-17 18:03:08,820 INFO No new useful revision published compared to dest, no need to upload this component [08:54] thostr_: that was someone putting that in the comment [08:54] ah [08:54] thostr_: not in the status field AFAIK [08:54] so you proposed indicator-network, which introduce no change [08:54] I guess robru forgot or didn't know about this corner-case [08:54] sil2100: and didn't try to understand :/ [08:54] so, should I now just rebuild and before remove the network=iindicator MP? [08:55] thostr_: no, there is some way to ignore that [08:55] so, what you should have done [08:55] is [08:55] Yeah... since why did he publish the second time when he could have already noticed that the silo doesn't have indicator-network for the second time [08:55] telling "ok, I know some projects are missing, that's on purpose" [08:55] IGNORE_MISSINGPROJECTS [08:55] Ignore if some projects that are going to be merged back were not published. No merge will then be processed for those [08:55] thostr_: this is the flag for that case ^ [08:56] that's when building the project [08:56] but what do I do now? [08:56] thostr_: that was when merging [08:56] now, robru removed the whole state [08:56] the only way is to free the silo [08:56] set landed manually [08:56] and bzr pull lp:~ps-jenkins//trusty-proposed && bzr push [08:57] for every component that were released [08:57] am I supposed to do that now? [08:57] thostr_: yeah [08:58] well "now" == when you have time to reconciliate the trunk [08:58] we are going to free the silo for you meanwhile [08:58] * didrocks runs merge and clean with ONLY_FREE_SILO [08:58] and set "Landed" to the spreadsheet [08:58] sil2100: does this make sense to you? ^ [08:59] didrocks: ok, right... since normal merge and clean would miss one revision? [08:59] I mean, release [09:00] Ok, right [09:00] didrocks: let's do it [09:00] sil2100: well, he destroyed the whole bzr branches === psivaa changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: psivaa | CITrain support - US: robru, cyphermox, rsalveti - EU: sil2100, Mirv, didrocks | CITrain support no answer: use mup bot after 30 minutes, but chooose right timezone | Landing instructions: http://goo.gl/8H1Du3 | Known Issues: - [09:00] didrocks: should I do the pulling and pushing? [09:01] sil2100: you can, please keep thostr_ in touch [09:01] sil2100: look at your publish job [09:01] you have the url to the branches [09:02] didrocks: ACK [09:02] thostr_: I'll do some bzr push magic on your projects now [09:02] sil2100: great, thanks!!! [09:03] thostr_: all the confusion (and people using overrides without knowing…) was due to indicator-network added to the list without having anything to flush. Can you check this doesn't reproduce please? [09:03] this is the second time from your team btw [09:03] the system tries to only rebuild if it's needed (which is available through an option as well) [09:29] didrocks: I'll be in the meeting in like 1-2 minutes [09:32] ogra_: around? [09:32] ogra_: meetonng [09:32] oops [09:36] thostr_: we have a unity-scope-loader crash (not reliably though): http://jenkins.qa.ubuntu.com/job/trusty-touch-mako-smoke-daily/54/artifact/clientlogs/camera_app/_usr_bin_unity-scope-loader.32011.crash/*view*/ [09:37] do you mind having a look at that as we landed a new version? [09:37] thostr_, the list of changes is at http://people.canonical.com/~ogra/touch-image-stats/20140217.1.changes [09:42] ogra_: looking [09:54] didrocks: the crash report is useless... we haven't changed that code in ages, all recent scopes changes are only applying to new scopes framework [09:55] and the new scopes don't utilize that unity-scope-loader any more [09:55] thostr_: ok, so let it die in limbos [09:56] didrocks: yes [10:02] thostr_, if it is not used, what keeps it in the image then ? [10:02] (it is not seeded) [10:02] ogra_: it's still used today [10:02] ah, k [10:02] but will go away when we switch to new scopes (probably next week) [10:02] right, understood [10:32] ;) [10:36] didrocks, maguro doesnt look good [10:36] unity hangs hard for me, some click apps seem not installable [10:38] WTH [10:38] hm [10:38] * ogra_ reboots === thostr_ is now known as thostr_lunch [10:48] reboots always fix everything :) [10:50] yeah, doesnt seem to hang anymore [10:50] ogra_: see, fixed! [10:50] ship it :) [10:50] * didrocks wonders where the icon theme is set… [10:50] woah [10:50] the camera on maguro always flickered ... [10:51] but now it is like a strobe [10:51] completely unusable [10:52] asac: maybe we should let people using maguro on the previous image and move on and promote the new one. What was the deprecation plan for this? ^ [10:52] nah, we shouldnt do that [10:52] are we in a hurry ? [10:52] you prefer to ship it with a broken camera? [10:52] else i would propose to wait for QA to get up [10:52] question is if anyone will look at the camera issue [10:52] * asac doubts it [10:53] i dont htink we should promote more "broken" stuff [10:53] based on the feedback we had with the other issues, I doubt we'll have support [10:53] to maguro [10:53] rather turn that channel off [10:53] asac: +1 [10:53] but would prefer to do that when moving mover [10:53] over' [10:53] asac, that will happen anyway as soon as we switch [10:53] lets see where we stand today wrt 4.4 switch [10:53] which is due this week [10:53] asac: well, so we can't promote anymore image until we switch with 4.4? [10:53] ogra_: right, but do we accept that we dont promote until 4.4? [10:53] thats not clear to me yet [10:54] asac: can you clarify that out with the management? [10:54] didrocks: lets wait for the south americans to come on [10:54] asac, well, i only did phone/sms smoketests the last days ... i dont even know if we didnt promote with a broken cam already [10:54] didrocks: management will not take that decision from me :) [10:54] ogra_: plz check [10:54] ogra_: dont complain about non-regressions :) [10:55] that causes confusions [10:55] hehe [10:55] then i cant do any other work [10:55] So, no promotion today :'( ? [10:55] anyone else tested maguro? [10:55] dogfooded? [10:55] asac: nobody else has a maguro around [10:55] sil2100, i would prefer to get some input from QA [10:55] davmor2: :) [10:55] asac: he's in the US timezone this week [10:55] which doesnt mean no promotion :) [10:55] do you know if old maguro images had camera broken [10:55] asac, right, once oakland gets up [10:56] which is like ... 10h from now :) [10:56] 5 i think [10:56] well ... in ~5 they should be available [10:56] didrocks: what issues do plague us currently on maguro? [10:56] do you know? [10:56] i dont see why we are in a hurry with promoting [10:56] asac: the only one I know, apart from tests being broken is what ogra_ told, the camera [10:57] asac, i had a hard hang when testing ... but that didnt come back afetr a reboot [10:57] psivaa: any idea why the demo dashboard shows "syncing" http://ci.ubuntu.com/smokeng/trusty/touch_custom_demo/mako/31:20140218:20140115.1/6657/ [10:57] asac: looking [10:57] asac, and the camera is definitely bad ... but its over a week ago i tried it the last time [10:57] ogra_: so you feel the camera was super stable a week ago? [10:57] no [10:57] or was it always buggy on maguro? [10:58] it was always flickering [10:58] and surely you sometimes ended up in a bogus state from time to time, right? [10:58] but in a state that you could still use it and actually see the pic [10:58] today it is like a strobe ... you cant really see the pci (it takes photos still though) [10:58] *pic [10:58] hmm [10:59] that phenomenon i had 3 month ago [10:59] when i was using maguro [10:59] i havent had hard hangs in quite a while, but yes, indeed i had some over the last year [10:59] sounds like its not a regression [10:59] ogra_: cant you downgrade and just use camera all over the house for 10 minutes? [10:59] :) [10:59] you could show me pics of your livingroom, cats, etc. :) [10:59] lol [10:59] well, it regressed as in it flickers so badly that it makes the app unusable [10:59] ogra_: that could be MIR induced [10:59] wheres it flickered before but you could still use it [11:00] ogra_: ok, try comparing to last promotion [11:00] *whereas [11:00] thats all that matters :) [11:00] guess no need to hurry as we are unlikely to take steps to address/fix this issue. just support decision what to do with maguro going foward [11:00] * ogra_ puts the bootspeed work aside and installs a new image [11:01] yeah sorry :) [11:01] no worries [11:08] asac: this appears to be a bug in the dashboard sync script. the results are in 'syncing' state right from the custom_demo builds. [11:09] psivaa: kk. please file ticket or whatever to have someone look i guess [11:10] as long as results are correct [11:10] then its not that important [11:10] but still buggy :) [11:10] asac: yea, filing a bug for it. [11:10] probably the reason why it's showing 0% on the KPI page [11:13] right [11:13] thats how i found it :) [11:13] wondered whats going on [11:13] but then saw "synching" [11:14] is mirv on vac? [11:15] * asac thinks so [11:18] asac: yes, he'll be back next week [11:19] didrocks: can I assign a silo for unity8 now anyway? I guess it won't affect maguro's camera-app I think? ;) [11:20] sil2100: sure [11:36] asac, didrocks, so on 188 there is the same fast flicker when i cover the lens or do fast movement, but it isnt there if i hold the camera still and focus on something ... in the latest proposed image it was there all the time [11:37] oh, and it seems to be completely gone when i lock/unlock the screen with the camera running in 188 [11:39] * ogra_ upgrades to proposed again to make sure [11:42] let's see :) [11:43] geh === thostr_lunch is now known as thostr_ [12:06] could anybody reconfigure silo3? [12:07] o/ [12:09] thostr_: done [12:16] ok, going for a run, see you later guys [12:17] asac, didrocks, so the flickering is worse on the new image, but the trick with lock/unlock (that i hadnt discovered before) makes it a bit better at least [12:25] ogra_: can you ask that in your meeting with rsalveti? [12:25] sure [12:25] to know if he has time to have a look [12:25] that will help to decide I guess [12:26] he knows about the flickering in general [12:26] its an issue on the android side [12:26] ogra_: just to confirm, it's possible to promote images per device, right? [12:26] yes, ugly but possible [12:26] why ugly, you have to put an hack for that? [12:26] (IIRC, you have a for … in loop) [12:27] no, it is just ugly to have one device released with known bugs that got fixed in other images on other devices [12:27] we usually dont do that [12:27] technically it isnt ugly ... [12:27] but it breaks an established philosophy [12:28] but given that the device will fully vanish from the channel by end of the week and not be installable at all anymore then i'm not even sure we should bother [12:29] yeah, that's why I think it's acceptable (due to the vanishing) [12:29] (with holding it back) [12:29] but let's see what rsalveti is able to tell (I'm sure he's more focused on mwc and 4.4) [12:29] right [12:30] and we shouldnt distract him to much ... to get that done [12:30] right [12:30] but let's get a confirmation first [12:31] popey, on your flo, could you check the last three lines of ~/.bashrc ? [12:32] export GRID_UNIT_PX=20 [12:32] i wonder if something went wrong with the patch that forces landscape ... and thus you get wrong default values for the resolution [12:32] export QTWEBKIT_DPR=2.5 [12:32] export NATIVE_ORIENTATION=portrait [12:32] oh nope [12:32] thats manta [12:32] * popey untangles usb cables [12:32] hmm, the first one should be 21 i think [12:32] * ogra_ chacks the code [12:32] fi [12:32] export GRID_UNIT_PX=21 [12:32] export QTWEBKIT_DPR=2.5 [12:32] ^^ thats flo ogra_ [12:33] ah, then it is right [12:33] you shouldnt get these giant icons with a value of 21 though [12:34] hmm, but you should have NATIVE_ORIENTATION=landscape in there too with the latest image [12:34] i fudged that into /etc/environment iirc [12:34] ah [12:34] yes, i did [12:35] i need to apt-get update this one for latest demo unity [12:35] it is in /etc/ubuntu-touch-session.d/flo.conf now [12:35] groovy [12:35] nexus 4 still dead ☹ [12:36] btw, i think we need to re-runn all the AP tests to make sure landscape doesnt break anything [12:36] popey, try detaching the battery cable [12:39] asac, ^^^ [12:40] (we should be re-testing in protrait on flo) [12:40] err [12:40] landscape [12:42] ogra_: what? take it apart? [12:42] popey, yes [12:42] not that hard [12:42] k [12:43] there is only one cable ... underneath a metal brace you have to unmount first [12:43] and all howtos i read suggested to leave the brace off when re-assembling it since it can cause unwanted grounding somewhere [12:43] i havent had any issues with mine since [12:56] ogra_: back off, unplugged, replugged battery, back still off, plugged in usb, flashing red light \o/ [12:56] (that's good right?) [12:56] yay [12:56] yep, it is [12:56] thanks [12:56] leave it charging [12:56] k [12:57] it should come out of that state on its own [12:57] if it doesnt, 5min button press should help === gatox is now known as gatox_brb === cjohnston changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: cjohnston | CITrain support - US: robru, cyphermox, rsalveti - EU: sil2100, Mirv, didrocks | CITrain support no answer: use mup bot after 30 minutes, but chooose right timezone | Landing instructions: http://goo.gl/8H1Du3 | Known Issues: - === alan_g is now known as alan_g|lunch [13:33] ogra_: yeah, however, we have never done that before, so would be good for now to check whats up with the APs [13:34] right [13:35] gah [13:35] so why did my flo come up in portrait now [13:35] root@ubuntu-phablet:/# tail -3 /home/phablet/.bashrc [13:35] export GRID_UNIT_PX=21 [13:35] export QTWEBKIT_DPR=2.5 [13:35] export NATIVE_ORIENTATION=landscape [13:36] hmpf [13:37] ogra_: maybe root session needs that env? [13:37] or we need something from the PPA [13:37] i just used the daily tarball [13:37] under the assumption it would work though [13:40] ogra_: most likely you need ppa then. yes [13:40] hmm [13:40] though people said just the env hack was needed in ppa [13:40] *shrug* [13:44] in PPA ... thats the point :) [13:45] * ogra_ cares for the images we will push to endusers ... there are enough people looking at the others :) [13:50] ogra_, seems those vars should be set system wide in /etc/environment or somewhere [13:51] they are exported by the session manager on session startup [13:51] the bashrc entries are just for AP === gatox_brb is now known as gatox [13:52] ogra_, I see [13:53] its just the easy place to check which values they have after a fresh install [13:53] didrocks: I'm shocked that the spreadsheet still works ok! Maybe your yesterday's optimization helped? === alan_g|lunch is now known as alan_g [14:05] sil2100: not sure, maybe that 5 writing in different cells in a short period of time confuses them [14:05] sil2100: as it seems to start when we have multiple silos [14:12] ogra_: while I was running, I was thinking about this: can you open a bug with the camera issue and workaround? If we go on and we promote without maguro, we can still reference and open for contribution [14:13] i would really like to promote all or not promote all instead of doing poecemeal [14:13] *piecemeal [14:18] didrocks: hi! In Online Accounts, we have a few merge proposals which have been merged into trunk, but have not landed yet. This week we started with the CI-train; so, do I add these packages to the Landing Asks document or...? [14:20] cjohnston: Hi, could we enable running the tests as part of https://jenkins.qa.ubuntu.com/job/mir-clang-trusty-amd64-build/ please? [14:21] mardy: you will be bootstrapped today, so we can flush trunk with an empty MP [14:21] mardy: hi! [14:21] alan_g: please file a bug against ubuntu-ci-services-itself [14:21] mardy: right, as didrocks said ^ [14:21] ogra_: that doesn't prevent to open a bug whatever decision we take, right? :) [14:21] didrocks, sil2100: thanks, I will! [14:21] cjohnston: ack [14:21] mardy: yw ;) [14:21] didrocks, indeed [14:22] sil2100: BTW, any news about raw 240 in the Landing Asks? (u-s-s-o-a) [14:24] mardy: we have dropped cu2d already and used it only for specific cases (mostly desktop-only ones), so we'll have to release this with citrain ;) [14:26] sil2100: Ah, OK, will do [14:59] sil2100: does this look OK? https://code.launchpad.net/~mardy/signon/packaging/+merge/206952 [15:00] sil2100: I mean, the way that the MP is presented, not the code [15:02] * sil2100 looks [15:04] mardy: hm, I'm not sure if we need to include this info in the project trunks - usually we only provide a wiki page for both the merge policy and test cases ;) [15:06] mardy: I guess nothing bad would happen if we did though! [15:07] sil2100: I'm following tedg's idea of including that in trunk, and then generate the wiki page automatically [15:07] sil2100: well, half-automatically :-) [15:08] sil2100, The big advantage is that the tests can be reviewed the same time as the feature. [15:08] * didrocks likes this idea [15:08] Indeed [15:08] Not that I'd talk about features so close to feature freeze ;-) [15:09] mardy: or rather than generating in the wiki page, you can say: "please refer to in trunk" [15:09] on that wiki page [15:09] kenvandine: are you going to be the "lander" for OA? [15:09] didrocks, The tool does both, it points out how to get it in the wiki page, and then includes it below. [15:09] ok :) [15:09] mardy, i doubt it [15:10] didrocks, I figured I had to do that to keep certain pro-wiki groups happy :-) [15:10] mardy: upstream is the lander, for all your team, I guess it's dbarth [15:10] tedg: interesting… ok :) [15:12] kenvandine: OK; and what about the code reviews? I can ask alex-abreu, if you prefer [15:12] kenvandine, didrocks: hi [15:13] dbarth, an say: "please refer to in trunk" [15:13] on that wiki page [15:13] oops [15:13] in the case of mardy's landing [15:13] dbarth, https://code.launchpad.net/~abreu-alexandre/qtcreator-plugin-ubuntu/add-inspector-option-to-html5-run/+merge/206955 [15:13] we have a dependency noted in the ci pipeline to land signon-ui in the archive first [15:13] kind of stuck between 2 worlds [15:13] alex-abreu: sweet [15:14] dbarth: why stuck between 2 worlds, you will land signon-ui as well, right? [15:14] mardy, which review? [15:14] for me, that's part of the whole "webapps team" responsability, right? [15:17] that should go in the same silo right? [15:17] yeah, it's a possibility [15:18] alex-abreu: for instance, https://code.launchpad.net/~mardy/signon/packaging/+merge/206952 [15:18] alex-abreu: but the question is also about all the future MP which I'll create :-) [15:19] sil2100: I think with the bootcamp, I won't be able to do the evening meeting + email, can you cover me? [15:19] kenvandine: what is a silo? I've been hearing this word many times, but I'm not sure what that means in the CI-train context [15:20] didrocks: ok, sure ;) Will try that - just to make sure, we didn't promote any image today due to the maguro issues, right? [15:20] sil2100: well, I guess ogra_ will have more info from rsalveti. No promotion yet. You can mention though the crash that we ignore and the work on barry's crash [15:21] mardy, the ci train gives you a landing silo, which includes a PPA where all the components of the landing get published [15:21] for review and testing [15:21] mardy, I could help w/ the reviews if you want [15:21] didrocks: right, Barry still didn't answer, seems not to be around [15:21] Not sure what JFDI means [15:21] ;) [15:21] alex-abreu: thanks; I guess that means that you should join the ~online-accounts team [15:22] heh :) [15:22] kenvandine: I got feedback from a lot of native speaker even who didn't know what a silo represents in fact [15:22] if I knew, I would have just called it cylon :p [15:22] mardy, will you be my manager? [15:22] kenvandine: if that means that a failure in one of the packages means that the whole bunch is rejected, I'd rather have OA in a different silo, as OA and webapps have very few contact points [15:22] yeah... silo is a good term for it though [15:23] kenvandine: of course it's a good one… because it exists in French! :) [15:23] mardy, no... the silo should contain all the components that need to change for the landing [15:24] kenvandine: OK, then it doesn't seem to be harmful to be in the same one :-) [15:24] mardy: silo != stack. stacks were stucked together forever [15:24] mardy: silo is dynamic, it's per landing [15:24] mardy, it's just for the one landing [15:24] so, you can say [15:24] "I'm going to land a new Mir" [15:24] then, you get a silo with mir + unity-mir + platform-api + unity-system-compositor + xorg [15:24] for instance [15:24]