| === nudtrobert1 is now known as nudtrobert | ||
| === dpm is now known as dpm-afk | ||
| Saviq | Mirv, so, about {add,insert,move}Items... upstream's doing a weird thing there IIUC, they just add the new methods to the 5.6 import, marking them \since 5.7 | 09:30 |
|---|---|---|
| Saviq | Mirv, so sounds like there won't be a 5.7 import of multimedia, or rather that 5.6 and 5.7 will be equal | 09:30 |
| Saviq | Mirv, but well, if that's what's upstream, I'm fine with that | 09:36 |
| Mirv | Saviq: yeah, I'm not sure how important that is as long as everything continues to work when we eventually switch to 5.6 and 5.7 for real. that's my only concern. so if there's anything that _needs_ to be done because they have been backported to 5.4/5.5, then it's useful to do, but only if it's really required for that special case | 09:42 |
| Saviq | Mirv, nope, it's fine with me, we can advertise we have those features starting with next framework, and then we'll just be dropping patches as we go | 09:43 |
| Mirv | yep | 09:45 |
| Saviq | Mirv, so one remaining question, how do we modify our Depends: to be sure we have the 5.6 import there? :) | 09:52 |
| Saviq | virtual package? | 09:56 |
| Saviq | Mirv, can't see any other way since we have it in >= 5.4.foo *and* >= 5.5.foo | 09:56 |
| Saviq | but not between 5.4.foo and 5.5.foo | 09:57 |
| Mirv | Saviq: virtual package might a bit heavy solution. maybe >= 5.4.1-1ubuntu20~ is enough - all 5.5 packages in xenial have had the core playlist support/import already, this newer 5.5 landing just adds a couple of more patches | 10:05 |
| Saviq | Mirv, well, what if we need the "a couple more patches"? | 10:06 |
| Mirv | Saviq: one option is to wait that 036 lands eg today or tomorrow | 10:06 |
| Saviq | Mirv, no worries, we can't land unity8 because of oxide-qt anyway, *but* I'm trying to do the right thing here ;) | 10:07 |
| Mirv | Saviq: I know you're trying to do the right thing, I'm not sure if it's worth it other than for the overlay since otherwise both landings would need reworking to add the virtual package and then we would need to hang to that delta from Debian either forever or a good while (and if we eventually remove it, it'd again be technically incorrect to depend on something else) | 10:08 |
| Saviq | Mirv, well, not if we depend on >= 5.7 at that point | 10:09 |
| Saviq | tsdgeos, do we use the "new" playlist bits ({add,move,remove}Items) in our audioCard? | 10:10 |
| Mirv | Saviq: true. anyway, I'm not sure if we need to adhere to the people who specifically retrieve an archived Qt version. but also true, if you require the three new functions, the >= 5.4.1-1ubuntu20~ is a wrong thing in xenial | 10:10 |
| tsdgeos | Saviq: we actually do, becuse jim convinced me it was so much better to use those than the existing ones | 10:10 |
| Mirv | technically. sometimes I'm not so sure if we need to follow all rules if it's technically complex and practically can't affect anyone | 10:11 |
| Saviq | Mirv, it's not about "archived Qt versions", rather that things can migrate in front of one another, people might not dist-upgrade | 10:11 |
| Saviq | dunno | 10:11 |
| Saviq | in this case we might just be lucky, because Qt can land/migrate before unity8 does | 10:12 |
| Mirv | Saviq: yeah. the correct thing from Ubuntu point of view would be to depend on >= 5.5.1-ubuntu2. however, you've the dual landing to a PPA that is not Ubuntu, so that'd require a change in Ubuntu that has nothing to do with Ubuntu the distro... | 10:12 |
| Mirv | or one can rewrite control from rules though, like some other projects | 10:12 |
| Saviq | that sounds like the "technically complex" bit, adding a virtual package wouldn't, although I understand why you're refraining from it | 10:13 |
| Saviq | Mirv, simples, then, land 5.5 in overlay ;) | 10:13 |
| Mirv | Saviq: ok, landing 028 as approved by Saviq. DONE! https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/landing-028/+packages has 5.5 for vivid :D | 10:14 |
| Saviq | ;) | 10:14 |
| Saviq | Mirv, anyway, we'll do whatever you tell us to do, you're the packaging master after all, but note I will deflect davmor2's blame^Wwrath unto you if shit hits the fan ;) | 10:15 |
| Saviq | and there's a *lot* of that (blame, not shit) | 10:16 |
| Mirv | Saviq: it's not davmor2 who cares about that, but the other packaging masters. but if we get this 5.5 update swiftly into xenial, I suggest just using >= 5.4.1-1ubuntu20~ | 10:17 |
| Saviq | Mirv, davmor2 cares about everything | 10:18 |
| davmor2 | Mirv: I'll blame saviq if the image breaks it was fine | 10:18 |
| Mirv | davmor2 oh, xenial image isn't broken yet? | 10:18 |
| Mirv | davmor2 how nice of you to approve xenial image as green! | 10:18 |
| Mirv | I'll announce that Dave has approved it | 10:18 |
| davmor2 | Mirv: Xenial is constantly broken and it's all Saviq s fault | 10:19 |
| Saviq | oh oh oh... but we actually test stuff when releasing to xenial ;P | 10:19 |
| Saviq | and we dual-land everything | 10:19 |
| Saviq | we're the good citizens here | 10:19 |
| davmor2 | Saviq: no you don't Developer mode was completely screwed | 10:20 |
| === dpm-afk is now known as dpm | ||
| Saviq | davmor2, not by us :P | 10:21 |
| davmor2 | Saviq: is it on the phone? Yes so it's your fault, you know the rules ;) | 10:21 |
| Saviq | dammit! | 10:21 |
| Saviq | so close... so close! | 10:21 |
| * Saviq goes back to pretending he EOY'd | 10:21 | |
| davmor2 | Saviq: admit it you missed all this while you were gone didn't you ;) | 10:22 |
| Saviq | tsdgeos, good news, with your last branch got a PASS again on vivid | 11:55 |
| tsdgeos | Saviq: :) | 12:04 |
| Saviq | now checking xenial | 12:15 |
| Saviq | kdub, hey, I'm afraid things went wrong in proposed migration and there's no one that's able to resolve it, not until pitti's back in January... | 12:22 |
| kdub | Saviq, thats unfortunate :( | 12:23 |
| kdub | Saviq, could retrying the whole process with those MP's work? | 12:29 |
| Saviq | kdub, unlikely | 12:30 |
| kdub | Saviq, alright. I still might ask around though, even if to understand what's blocking us better | 12:38 |
| === maclin1 is now known as maclin | ||
| tsdgeos | jhodapp: there? | 13:38 |
| jhodapp | tsdgeos, yes but in a meeting, be with you in a few mins | 13:38 |
| jhodapp | tsdgeos, hey | 13:41 |
| tsdgeos | jhodapp: i can make dash video playing crash | 13:42 |
| tsdgeos | the bactrace is in 0xb42bfe2c that belongs to /android/system/lib/libEGL.so | 13:43 |
| tsdgeos | but all i have is ?? | 13:43 |
| tsdgeos | any idea how to get a better bt? | 13:43 |
| jhodapp | oh interesting, sounds like a NULL graphic buffer issue | 13:43 |
| tsdgeos | paste.ubuntu.com/14137574/ | 13:44 |
| jhodapp | tsdgeos, https://wiki.ubuntu.com/Touch/Core/UbuntuDebugAndroid | 13:44 |
| tsdgeos | thread 26 seems to be the one with the crash | 13:44 |
| tsdgeos | it's actually a Qt thread :? | 13:45 |
| tsdgeos | * 26 Thread 0xa7614410 (LWP 3717) "QSGRenderThread" 0xb24cc72e in ?? () | 13:45 |
| tsdgeos | let me install some more qt dgb then | 13:45 |
| jhodapp | tsdgeos, oh geez, not the scene graph renderer again | 13:45 |
| jhodapp | that thing is very complicated | 13:46 |
| tsdgeos | yep | 13:47 |
| tsdgeos | jhodapp: in that wiki there's a step i don't know how to fullfil | 14:09 |
| tsdgeos | it says | 14:09 |
| tsdgeos | "If you have a local setup of our CyanogenMod based phablet builds" | 14:09 |
| tsdgeos | i don't | 14:09 |
| jhodapp | tsdgeos, that's slightly out of date, it should read AOSP and you can get one via phablet-dev-bootstrap | 14:10 |
| tsdgeos | jhodapp: so "phablet-dev-bootstrap mydir" should be enough? | 14:12 |
| jhodapp | tsdgeos, I did: phablet-dev-bootstrap -v mako aosp | 14:12 |
| jhodapp | that'll make sure you can generate boot, recovery and system images for mako specifically | 14:13 |
| tsdgeos | jhodapp: what if i'm not using a mako? | 14:14 |
| tsdgeos | ERROR:phablet-dev-bootstrap:Vendor device arale not supported | 14:15 |
| jhodapp | tsdgeos, we don't maintain the vendor specific code/blobs with that tool for arale or krillin, so it's best to just target mako or flo | 14:15 |
| jhodapp | tsdgeos, the binaries you need should work on arale too | 14:16 |
| tsdgeos | k | 14:16 |
| jhodapp | you can adb push them over | 14:16 |
| tsdgeos | jhodapp: is that supposed to take like forever? it's still downloading stuff | 14:48 |
| jhodapp | tsdgeos, yeah it's rather large | 14:49 |
| Mirv | Saviq: tsdgeos: qtmultimedia is now published to xenial | 15:21 |
| tsdgeos | k | 15:21 |
| Saviq | kdub, yeah, I'm sure them guys will be interested to know what happened | 15:25 |
| Saviq | Mirv, do you think rebuilding qtmir/-gles and republishing it would kick autopkgtests in migration? ours are stuck since yesterday | 15:25 |
| Saviq | tsdgeos, FYI, these are the failures in xenial http://pastebin.ubuntu.com/14138615/ (LazyImage is likely flaky, should be better with cimi's branch hopefully) | 15:28 |
| Mirv | Saviq: I don't know, but that could work if the armhf are just not completely broken. they seem to pass for some tests so it's just the problem of some runs getting stuck. | 15:28 |
| Saviq | Mirv, ok, let's try, /me kicking rebuild | 15:29 |
| tsdgeos | Saviq: that's locally? | 15:30 |
| Saviq | tsdgeos, yes, in adt-run | 15:30 |
| Saviq | kdub, oh, but you should be able to build u-s-c in a silo, the mir silo was forced to merge/clean | 15:34 |
| Saviq | so while it's stuck in proposed, it's all fine in vivid/overlay and lp:*/ubuntu\ | 15:38 |
| Saviq | -\ | 15:38 |
| kdub | hmm | 15:38 |
| tsdgeos | Saviq: does that run have https://code.launchpad.net/~aacid/unity8/testIndicatorsMenuFix/+merge/280165 ? | 15:42 |
| Saviq | kdub, worst case scenario, a version gets skipped in xenial | 15:42 |
| Saviq | tsdgeos, prolly not | 15:42 |
| kdub | yeah, so maybe just try a vivid-only landing for the fix affecting the image in silo 45 | 15:42 |
| Saviq | kdub, nooo | 15:42 |
| kdub | ah, :) totally misunderstood then | 15:42 |
| Saviq | kdub, land to both vivid+xenial, worst case it will get stuck in xenial proposed again | 15:42 |
| tsdgeos | Saviq: that branch should fix test_verticalVelocityDetector | 15:42 |
| dandrader | Saviq, are you landing https://code.launchpad.net/~nick-dedekind/qtmir/qtmir-test-build/+merge/272932 ? | 15:42 |
| Saviq | tsdgeos, ack | 15:42 |
| kdub | Saviq, ah, alright | 15:42 |
| Saviq | kdub, but we'll unstuck it when things go back to normal | 15:43 |
| Saviq | dandrader, can't really land qtmir or unity8 atm, all kinds of things went wrong | 15:43 |
| Saviq | dandrader, so no | 15:43 |
| kdub | Saviq, ack, will try to rebuild the fix there and see how far that can go to fix the images | 15:46 |
| kdub | may as well keep moving where we can | 15:46 |
| * voxel_ xenial sooooooooo broken :'( | 16:10 | |
| voxel_ | not sure what happened but i don't get a black screen anymore u8/xenial yay | 16:11 |
| voxel_ | it's more than 1 month since the last time web browser worked and xmir... i can't even remember... 15.10 | 16:12 |
| voxel_ | oh well :( | 16:12 |
| === voxel_ is now known as vox_aw | ||
| Saviq | vox_aw, did you file bugs? | 16:25 |
| vox_aw | yep | 16:26 |
| Saviq | vox_aw, then they're probably a better audience than here | 16:27 |
| Saviq | as neither browser nor mir folk generally read this channel unless pung | 16:27 |
| Mirv | Saviq: tsdgeos: ok merry xmas silo 028 seems to boot to Unity 8 properly now finally. so, vivid-overlay + Qt 5.5. pinning required. | 16:55 |
| Saviq | Mirv, you're out tomorrow? have a great time off, all the best for the family :) | 16:56 |
| Mirv | that could be useful at some point | 16:56 |
| Mirv | Saviq: still here tomorrow, but gave out the present early :) anyway, likewise regardless! | 16:57 |
| Saviq | :) | 16:57 |
| tsdgeos | Mirv: :) | 16:57 |
| === dandrader is now known as dandrader|afk | ||
| q_ | How can I switch my keboard shortcuts Super and Super-W ? | 18:12 |
| === dandrader|afk is now known as dandrader | ||
| === attente_ is now known as attente | ||
| === davmor2 is now known as davmor2_ho-ho-ho | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!