=== 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 [14:00] didrocks: can we make an attempt at libmirclient3 again ? we will need this for dpms [14:00] kgunn: we'll deliver first current Mir [14:01] kgunn: I ask Mirv to do it for tomorrow [14:01] then, we can merge your transition, but your team has to handle it [14:03] didrocks: so, handle it means...we send anounce on ubuntu-devel, we ping you (or someone like you), we place it on the ask sheet [14:03] didrocks: is there anything else ? [14:03] racarr: ^ [14:03] kgunn: and you ask your team to promptly upload the driver parts [14:03] as you break ABI, everything that deps on libmirclient has to be rebuilt [14:03] didrocks: yes...xmir dependency [14:03] and some parts are not under CI [14:03] RAOF: ^ [14:03] so your team needs to upload that promptly to the ppa to not block everything [14:04] hence the "we need to coordinate" [14:04] Yup === dandrader_ is now known as dandrader_|afk === dandrader_|afk is now known as dandrader|afk === dandrader|afk is now known as dandrader [15:48] hmm [15:49] how does one get the boost libs to appear in /usr/lib like the mir makefiles seem to expect [15:49] they're installed in the arch-prefixed directories [15:49] (symlinking doesnt count) [15:53] Mirv: hey was going to ask didrocks...but someone was noticing there is no saucy proposed (i haven't checked this) [15:53] Mirv: we were discussing the "ask" process...i indicated, trunk will still go to proposed [15:53] but won't be promoted to archive until its been approved via the ask process [16:24] sil2100: ^ can you see what i was saying to Mirv .... he might be mising in action [16:24] sil2100: there should be proposed with latest trunk from Mir right ? [16:24] kgunn: I think he's long gone, since he's EOD since 3 hours already [16:24] But looking [16:25] hmmm, no, I don't think it's like that [16:25] We're running releases of stacks manually now, so even daily-build doesn't have latest Mir [16:25] We'll be trying to release new Mir around tomorrow [16:31] sil2100: ug...so....to make sure i get it :) there will be a daily-build based on mir trunk latest [16:31] ? [16:31] which will actually be going thru testing...(per what didrocks indicated) [16:32] sil2100: just to let you know we are here in lexington at a mir sprint....so we're going to be pushing quite a bit of bug fixes thru [16:36] kgunn: we'll probably have a detailed talk about Mir tomorrow in the UTC morning, but the deal is - we're spinning releases manually (not automatically every 4 hours) [16:36] kgunn: so, essentially, there won't be a new Mir built in daily-build every 4 hours [16:37] kgunn: but whenever asac and didrocks decide that we will be testing a new Mir for the image, we'll be spinning manually a release of Mir which will land into daily-build [16:38] kgunn: go through our testing [16:38] kgunn: and if all is green we will release it to -proposed === dandrader is now known as dandrader|lunch [16:41] sil2100: thank you...i realize you are the messenger so i won't shoot :)...but i am a little worried [16:42] this might cause us a bit of slow down...at least a little concerned we're not getting build/integration feedback [16:42] sil2100: ...any chance someone could help launch a manual build for latest trunk eod our time in US ?? [16:42] we have some key things we want in today [16:43] is there some sort of EGL_MIR_PLATFORM #define ? [16:43] so that I get the right typedefs [16:43] kgunn: hmmm [16:43] sil2100: even now...we're 5 days diff from what's in archive to whats in trunk [16:43] kgunn: I personally doubt it, as asac is the person that needs to say 'OK!' when doing any building [16:44] kgunn: since we want to release Mir tomorrow [16:44] Not sure if he would give a green light for building things today ;p But maybe? [16:44] sil2100: thanks i'll ask [16:44] kgunn: it's best to poke asac as he's the guy in charge [16:44] asac: ^ we're in the midst of mir sprint in lexington (we traded places) [16:45] asac: i understand ur in manual build mode....any chance someone could kick us a manual build eod US time ? [16:45] ...we are hoping to have some key things lined up for merging/landing into archive today [16:45] and i was just pointing out...we're 5 days old in archive vs trunk [16:51] kgunn: so we work on a unity update that cannot be done with an updated mir [16:51] kgunn: from what i understand, one thing that would be soo much easier if you guys would commit to maintain an abi/api properly. [16:52] asac: we are committed to that [16:52] asac: already discussed with didrocks as well [16:53] asac: and we do need to bump the api [16:53] asac: afaik that means we will need to land the corresponding xmir changes with it [16:53] kgunn: its also libmirserver [16:53] that one seems to be even more of a mess [16:53] asac: but....it will require a full rebuild of the stack [16:53] you say you will maintain that ? [16:54] kgunn: at the moment we have to rebuild the full stack before landing [16:54] asac: yes [16:54] kgunn: with proper versioning we can do that step by step (e.g. first land mir, then update the rest of the stack in a second shot) [16:54] asac: yes, i understand you have to rebuild the whole stack [16:55] sil2100: so from what i understand we have a daily-release [16:55] asac: yes...that's the plan, we have libmirserver3 in waiting, we are making sure this afternoon our changes on the xmir side correspond [16:55] and a daily-release-next ppa [16:55] correct? [16:55] asac: other mir clients simply need to rebuild (they don't use the new method) [16:55] sil2100: you think we could pipe kgunn into whatever we dont use right now? [16:56] kgunn, asac: we can use the -next PPA I guess... [16:56] asac: is there anyone in late U.S. time that could help here ? (realizing poor sil2100 wants to go to sleep sometime) [16:56] kgunn: kenvandine i guess [16:56] asac: ah...thanks [16:56] kgunn: dont you have a ppa you used in the past? [16:57] for the mir preview images? [16:57] etc.? [16:57] kgunn: and cyphermox as well [16:57] he knows the system pretty good [16:57] asac: we're shooting for archive...we can test all we want off trunk, but what I do want is the integration test result from being on a current desktop/touch image [16:58] does that make sense ? [16:58] kgunn: i am not sure :) [16:58] asac: so olli_ is going to murder me if we aren't default xmir on desktop by thursday... [16:59] ok... that i understand :) [16:59] ;) [16:59] kgunn: everything on desktop you can just get in without being coordinated [16:59] asac, the question is, will you prevent it [16:59] :) [16:59] if it doesnt touch touch, i dont care [16:59] kgunn: I can propose a switch so that the new Mir stuff is going to land to the daily-build-next PPA, but you'll have to poke kenvandine or cyphermox to get the releases rolling anyway [17:00] asac: mir is the same component for both [17:00] kgunn: there's also the problem of needing a rebuild of unity8 and platform-api... [17:00] sil2100: that's awesome....thanks... and we have greyback & Saviq for any unity8 issues [17:01] RAOF: racarr ^ so i figure if we can square dpms away this afternoon...we can get a daily build going that we can put on all the systems here in lex [17:15] asac: reading my own scrollback...just to make sure, mir is obviously a touch & desktop component...so when we make changes it needs to happen in both (esp in the case of libmirclient3...we were bumping for desktop, but obviously touch components will need to accomodate) [17:16] racarr: libmirclient relevant touch components are unity-mir & platform-api ? right? [17:21] kgunn: so unity mir fixes are scheduled tomororow. having you guys ramp up in the -next ppa is probably right [17:22] you guys can then test and work with cypher and ken on getting udpates intot he ppa so we can release tomorrow morning the next shot [17:22] kgunn: but my understanding is that if you pick up the libmirclient bump you also need new drivers etc. [17:22] sil2100: is that on your radar? [17:25] asac: right....only Xmir driver (which is RAOF 's and he is here) will be new...it uses the additional methods, but all other mir clients only need to be rebuilt [17:25] as the mir clients (other than xmir) don't use the new api [17:57] RAOF, https://code.launchpad.net/~hikiko/mir/mir.1226144 === FunnyLookinHat_ is now known as FunnyLookinHat [18:08] asac, we need quick iterations to land stuff in time fro 9/19 [18:09] sil2100: just to make sure i'm not crazy...is the autobuilding off for both desktop and touch ? or just touch ? [18:09] can't afford to block on someone manually approving it, e.g. during the night shifts the team might be pulling here [18:09] we need to have builds available when needed [18:11] olli_: you have a ppa [18:11] kgunn: define 'autobuilding' ;) [18:11] olli_: you can stage stuff there now [18:11] olli_: and kenvandine and cyphermox can help you getting updates in there === dandrader|lunch is now known as dandrader [18:35] sil2100: "autobuilding" as opposed to the manual building that you're doing now [18:36] kgunn: ah, for daily-release you mean? Since the touch images take packages which we release, so basically both desktop and touch are in manual building mode... [18:38] sil2100: that's what i needed [20:56] hikiko: http://unity.ubuntu.com/mir/building_source_for_android.html [21:35] alf_, https://bugs.launchpad.net/mir/+bug/1226144 [21:35] Ubuntu bug 1226144 in Mir "switching to the "unity" (graphical) vt after exiting lightdm and mir causes a crash" [Critical,In progress]