| === chihchun is now known as chihchun_afk | ||
| RAOF | BAH! How am I failing at geometry? | 04:27 |
|---|---|---|
| dholbach | hiza | 09:13 |
| dholbach | hiya | 09:13 |
| dholbach | can anyone please comment on my mail to the mailing list? | 09:14 |
| alan_g | dholbach: I'm totally unfamiliar with that library, what it does and how we use it. What would constitute a useful "quick look"? | 09:17 |
| dholbach | alan_g, I don't know what mir's dependency on libxkbcommon is and how an update would affect mir | 09:19 |
| dholbach | maybe we could use the updated package and just test on desktop and phone? | 09:19 |
| * alan_g Wonders if a keymapping library is relevant to phone - it sounds more like something XMir uses. Starts digging. | 09:25 | |
| greyback | alan_g: hey, just a note: platform-api uses mir::DefaultServerConfiguration::the_shell_surface_factory which you're removing... | 09:31 |
| alan_g | greyback: Oops. Thanks for the warning. | 09:31 |
| alf__ | greyback: Hi! Any experimental results for non-blocking swapbuffers? | 09:32 |
| greyback | alf__: I'm like 95% ready to test, you just chose a mir version to branch which makes changes unity-mir needed updating for. | 09:33 |
| greyback | so just getting the bits to all worth together is tricky | 09:33 |
| alf__ | greyback: ok, I can rebase to some other version if it's easier to test | 09:34 |
| alf__ | greyback: which version is easier for you? | 09:34 |
| greyback | alf__: so can I :) Gimme 20 mins and I should be there | 09:35 |
| alf__ | greyback: ok, np :) | 09:35 |
| * alan_g realizes he needs to set ENABLE_MIRSERVER_IMPLEMENTATION | 10:14 | |
| dholbach | alan_g, do you know of anyone who knows how libxkcommon is related to mir? | 10:16 |
| alan_g | alf__: ^ | 10:17 |
| alf__ | dholbach: alan_g: I would guess racarr, xkbcommon seems to be used only in src/shared/input/xkb_mapper.cpp which was written by him | 10:20 |
| dholbach | racarr, if you could reply to my mail on the list, that'd be great | 10:20 |
| dholbach | thanks alf__ | 10:20 |
| alf__ | dholbach: np | 10:21 |
| greyback | alan_g: question - how does one convert a mir::scene:Surface to a mir::shell::Surface? | 10:23 |
| greyback | oh, shell::Surface eliminated in your MR | 10:24 |
| greyback | never mind | 10:24 |
| alan_g | greyback: If mir::shell::Surface exists on the branch then it is one | 10:24 |
| alan_g | Wow! You're ahead of development-branch? | 10:25 |
| greyback | alan_g: you're changes were not insignificant, so best ot be ahead of the curve | 10:29 |
| alan_g | greyback: I hope they'll make things easier for you in future. | 10:30 |
| greyback | alan_g: you'd be hearing from me if I disapproved :) Yes I think they'll suit us fine | 10:30 |
| greyback | alan_g: I've thrown together the platform-api fix: lp:~gerboland/platform-api/mir-SurfaceFactory-to-SurfaceCoordinator | 10:41 |
| alan_g | greyback: so have I lp:~alan-griffiths/platform-api/compatibility-with-proposed-mir-changes | 10:42 |
| alan_g | ;) | 10:43 |
| greyback | heh, you won :D | 10:43 |
| alan_g | I did? | 10:43 |
| alf__ | greyback: So looking at USC code, it seems that it completely stops the compositor if screen is off, however my fix depends on the compositor running and consuming frames | 11:01 |
| alf__ | greyback: that's probably why things still get stuck | 11:01 |
| greyback | alf__: aha, so USC also needs changing | 11:01 |
| greyback | alan_g: is lp:~alan-griffiths/unity-mir/compatibility-with-mir-changes supposed to allow unity-mir build with Mir's developer branch, as it is right now? | 11:04 |
| alan_g | greyback: that's the idea. I've not checked today | 11:05 |
| alf__ | greyback: yes, although we should provide a better interface for USC to interact with the display configuration | 11:05 |
| greyback | as last night it wasn't enough, I had to pull in your compatibility-with-proposed-mir-changes & the corresponging mir branch to get all to work | 11:05 |
| alan_g | greyback: so if we can land the Mir branch I don't need to fix it. ;) | 11:06 |
| greyback | alan_g: indeed :) | 11:06 |
| greyback | works for me | 11:06 |
| greyback | alf__: I'll hack USB to not explicitly stop/start the compositor, and let you know the results. Will take ~45 mins, had to reflash phone as upstart died | 11:07 |
| alan_g | greyback: alf__ anpok can we have some "approves" on https://code.launchpad.net/~alan-griffiths/mir/eliminate-shell-SurfaceFactory-and-Surface/+merge/214041 ^^ | 11:07 |
| alf__ | greyback: something like this should work http://paste.ubuntu.com/7202738/ (haven't tried) | 11:08 |
| greyback | alf__: why stop/start it at all? | 11:08 |
| greyback | oh, because you switch btween 2 compositor things | 11:09 |
| alf__ | greyback: it may not be needed with the current state of things. Ideally if a screen is off the mg::Display shouldn't even produce a DisplayDuffer for it. Any change in DisplayBuffers needs a compositor restart. | 11:12 |
| alf__ | greyback: actually we need to do it for mesa at least | 11:15 |
| === alan_g is now known as alan_g|lunch | ||
| === alan_g|lunch is now known as alan_g | ||
| === dandrader is now known as dandrader|afk | ||
| kgunn | bregma: this ended up being the problem... | 13:53 |
| kgunn | https://bugs.launchpad.net/ubuntu/+source/xorg/+bug/1302270 | 13:53 |
| ubot5 | Ubuntu bug 1302264 in systemd (Ubuntu) "duplicate for #1302270 systemd-logind assert failure: error.c:319: Assertion failed in nih_error_get: context_stack != NULL" [Critical,Fix released] | 13:53 |
| bregma | kgunn, so now you're using Unity 8 on the desktop for all your day-to-day needs/ | 13:59 |
| anpok | mterry: the issue from yesterday | 14:23 |
| anpok | did that one disappear? | 14:24 |
| === dandrader|afk is now known as dandrader | ||
| anpok | ok just experienced it | 14:28 |
| === alan_g is now known as alan_g|tea | ||
| === alan_g|tea is now known as alan_g | ||
| alf__ | alan_g: greyback: Is there a reason we don't have MP for platform-api and unity-mir compatibility-with-proposed-mir-changes branches? | 14:40 |
| alan_g | alf__: I was going to MP when the changes land on development-branch | 14:41 |
| alf__ | alan_g: ok | 14:42 |
| === alan_g is now known as alan_g|tea | ||
| === alan_g|tea is now known as alan_g | ||
| === jono is now known as Guest82898 | ||
| === bschaefer_ is now known as bschaefer | ||
| anpok | mterry: that issue from yesterday - greeter is visible but not reacting to input - what where your steps to reproduce it? So far it only happens for me once on start up | 15:38 |
| anpok | and it disappears when display is turned off/on again.. | 15:38 |
| mterry | anpok, I can't reproduce yeah | 15:38 |
| anpok | maybe I have a broken combination of mir related packages.. | 15:39 |
| anpok | but I get it quite frequently - restart lightdm - > display turns off and greeter is frozen.. | 15:39 |
| anpok | *display turns on | 15:39 |
| anpok | EventHub does not seem to wake up on touch presses | 15:41 |
| mterry | anpok, then maybe I lucked out in my combination of mir packages.... | 15:42 |
| mterry | anpok, ah.... I think I know what it was -- it was more of the raising-sessions regression I saw. The transparent spinner session must have been on top | 15:44 |
| mterry | anpok, are you using my split greeter branches? | 15:44 |
| bregma | kgunn, the u-s-c in landing-011 tests out OK for me: I'm not sure if it needs phone testing but I'm keen to see it land if you're ready to pull the trigger | 15:47 |
| kgunn | bregma: i did test phone it was fine.... | 15:49 |
| kgunn | just hadn't had time to sort the desktop this morning...after recovering from the bad xorg pkg issue last night...let's land it :) | 15:49 |
| bregma | kgunn, it's time to release it to the wild then | 15:49 |
| kgunn | bregma: i''m on it | 15:50 |
| anpok | mterry: yeah .. still your usc branch and mir-devel from today | 15:51 |
| mterry | anpok, OK yeah. So session-raising got broken somehow. And the (transparent) spinner session is on top of the greeter | 15:52 |
| mterry | anpok, set_focus_to() no longer raises a session and even manually calling raise on the individual surfaces doesn't seem to do the trick | 15:52 |
| anpok | yeah that was working with devel | 15:53 |
| mterry | anpok, it was? Ever since latest devel it seems to have stopped working for me | 15:54 |
| mterry | "latest devel" -- I'm not actually sure when | 15:54 |
| alan_g | mterry: which -r are you at? | 15:58 |
| mterry | alan_g, that's a great question... something from Wednesday | 15:59 |
| mterry | (the silo 002 PPA built it and it doesn't make it obvious which revision was used) | 16:00 |
| anpok | -r 1528 seems to work | 16:00 |
| mterry | anpok, but you said you were having the same input problem with the greeter? | 16:02 |
| mterry | anpok, that means whatever devel version you are using has the session-raising problem | 16:03 |
| * alan_g has touched the codepath - but there were no test breaking. Will write one. So: create a couple of sessions with surfaces and try to raise the 1st one? | 16:03 | |
| kgunn | mterry: want me to rebuild mir ? | 16:03 |
| alan_g | Has a bug been filed? | 16:03 |
| mterry | alan_g, yeah I guess. I'm using both raise on surfaces and set_focus_to, neither seems to do the trick | 16:04 |
| mterry | alan_g, no I wasn't sure until today what was even wrong | 16:04 |
| mterry | kgunn, I'm not sure devel is fixed yet | 16:04 |
| mterry | alan_g, I can file a bug, give me a sec | 16:04 |
| kgunn | mterry: ok...was just thinking if you wanted to get to the tip | 16:04 |
| anpok | mterry: no session raising issues, just user input not working | 16:04 |
| mterry | anpok, right. But that is (I believe) a symptom of the transparent spinner session sitting on top of greeter when it shouldn't be, stealing all the input | 16:05 |
| mterry | anpok, because the greeter wasn't being raised properly | 16:05 |
| mterry | kgunn, not yet, I don't think | 16:06 |
| kgunn | ack | 16:06 |
| mterry | alan_g, bug 1302689 | 16:10 |
| ubot5 | bug 1302689 in mir (Ubuntu) "Session raising problems" [Undecided,New] https://launchpad.net/bugs/1302689 | 16:10 |
| alan_g | ta | 16:10 |
| mterry | kgunn, actually... it looks like there are quite some API changes in mir/devel recently that will affect my branches. So maybe a respin of mir now would give me an easier time of updating to deal with them | 16:15 |
| mterry | I'll still need a respin later for the raise bug fix, but I can update to API now anyway | 16:15 |
| didrocks | mterry: hey, nothing related to that, but you are putting my laptop at risk! :) | 16:30 |
| mterry | didrocks, did you see a deja-dup CPU craziness? | 16:30 |
| didrocks | mterry: not more than usual: more seriously deja-dup (surely duplicity) is trying to backup, but I'm getting: | 16:31 |
| didrocks | Invalid data - SHA1 hash mismatch for file: | 16:31 |
| didrocks | duplicity-full.20140327T181132Z.vol1.difftar.gpg | 16:31 |
| didrocks | everytime | 16:31 |
| didrocks | so I don't have backup! zomg :) | 16:31 |
| didrocks | (the computed and real fingerprints are indeed different) | 16:31 |
| mterry | didrocks, that reminds me of an old old bug that was fixed years ago | 16:31 |
| didrocks | mterry: I moved my backup server to a rasperry in case the backup server impacts it… | 16:32 |
| didrocks | (but I thought the compression was client-side) | 16:32 |
| mterry | didrocks, and if you backed up 03/27, you should have a presumably-bug-free version of duplicity | 16:32 |
| mterry | didrocks, correct | 16:32 |
| didrocks | mterry: I didn't, my server went down and that's why I bought a rasperry | 16:33 |
| didrocks | so, it was my first new "full" backup | 16:33 |
| didrocks | mterry: oh no, I have one from 27/03 | 16:34 |
| didrocks | which was the one shown | 16:34 |
| didrocks | mterry: is there any logs I can throw your way? | 16:35 |
| kgunn | mterry: rebuilding | 16:36 |
| mterry | didrocks, not especially... What I need is to go back in time to 03/27 and fix it there | 16:36 |
| mterry | didrocks, find out what happened I mean | 16:36 |
| kgunn | mterry: so for usc, will you have a seperate branch i can land fo' real if we promote mir-devel ? | 16:36 |
| didrocks | mterry: so, I can remove this backup? it doesn't seem to be complete anyway | 16:36 |
| kgunn | gotta run for lunch..bbiab | 16:36 |
| mterry | didrocks, yeah | 16:36 |
| didrocks | like having 9 volumes instead of 1800+ | 16:36 |
| didrocks | mterry: ok, will do that and relaunch a backup, thanks! | 16:37 |
| mterry | didrocks, sorry man | 16:37 |
| didrocks | no worry, thanks you for unblocking me :) | 16:37 |
| kdub | bzr! | 16:38 |
| === chihchun_afk is now known as chihchun | ||
| === alan_g is now known as alan_g|EOD | ||
| davmor2 | kgunn: has anyone on the team got a manta | 18:08 |
| kgunn | davmor2: everyone i think | 18:38 |
| davmor2 | kgunn: update to latest proposed open the weather app and hit the bottom date on the right hand side apparently in the desktop version scrolling is smooth and fast | 18:39 |
| davmor2 | kgunn: I've filed a bug against the weather app https://bugs.launchpad.net/ubuntu-weather-app/+bug/1302728 but it is on the off chance that it is actually mir/qt/qml that is the issue I thought I'd give you guys a heads up | 18:41 |
| ubot5 | Ubuntu bug 1302728 in Ubuntu Weather App "extremely slow scrolling between days on manta" [Undecided,Confirmed] | 18:41 |
| kgunn | mterry: uh-oh, merge conflict on the mir rebuild in split greeter silo....for https://code.launchpad.net/~mterry/mir/no-initial-display-configuration-sent-to-hosting-server-merged | 18:43 |
| kgunn | https://ci-train.ubuntu.com/job/landing-002-1-build/25/console | 18:43 |
| kgunn | davmor2: sure | 18:43 |
| mterry | kgunn, you can drop that branch, it made it into trunk | 18:43 |
| kgunn | mterry: ack | 18:44 |
| kgunn | mterry: ...so you have no mp on mir-dev ?...e.g. i need to pick or gen one to rebuild? | 18:46 |
| mterry | kgunn, oh right. I'll make a fake one | 18:47 |
| kgunn | mterry: oh wait...you have this one | 18:47 |
| kgunn | https://code.launchpad.net/~mterry/mir/missing-links/+merge/213906 | 18:47 |
| mterry | kgunn, that made it in too. But probably won't conflict, just will be a no-op | 18:47 |
| mterry | kgunn, so try that alone | 18:47 |
| kgunn | roger that | 18:47 |
| * mterry goes afk for a little bit | 18:48 | |
| === chihchun is now known as chihchun_afk | ||
| === chihchun_afk is now known as chihchun | ||
| === chihchun is now known as chihchun_afk | ||
| === chihchun_afk is now known as chihchun | ||
| === chihchun is now known as chihchun_afk | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!