=== udsbotu changed the topic of #ubuntu-uds-client-1 to: Currently no events are active in this room - http://summit.ubuntu.com/uds-1308/client-1/ - http://irclogs.ubuntu.com/2013/08/28/%23ubuntu-uds-client-1.html === marvin is now known as Guest73214 === matteo is now known as Guest38214 === dednick is now known as dednick|lunch === smile is now known as Guest29401 === dednick|lunch is now known as dednick === udsbotu changed the topic of #ubuntu-uds-client-1 to: Track: Client | Qt 5.1(.1) migration plans and blockers | Url: http://summit.ubuntu.com/uds-1308/meeting/21934/client-1308-qt51/ === tsdgeos_uds_ is now known as tsdgeos_uds [13:59] Hangout link: https://plus.google.com/hangouts/_/368836c93bc010c273ce53686164a50dd2e47904?authuser=0&hl=en <- hangout link [14:03] * rsalveti waves [14:09] Did anyone from upstream review this patch? [14:09] sil2100: ↑↑↑↑ [14:13] Mirv: is the package with jit enabled by default on arm already available in the ppa? [14:13] we'd like to have it enabled by default if possible [14:13] it's way faster at least [14:14] ScottK: you mean the QPA themeplugin order-load patch? === shock is now known as shockwavenl [14:14] Yes. [14:14] one easy way to check, is going to https://www.webkit.org/perf/sunspider/sunspider.html [14:14] pmcgowan: ^ [14:14] ok [14:14] rsalveti, please do ;) [14:14] if it works fine with that, when testing jit, then it'll probably work everywhere [14:14] pmcgowan: sure :-) [14:15] ScottK: not yet, as I said, once I get some info from the upstream developers and then assess if it's really needed [14:15] ScottK, what do you think about getting 5.1 into saucy? [14:15] ScottK: but it seems so for now at least [14:15] Mirv: pmcgowan: do you know the status of the pixelratio patch in qtwebkit? [14:15] I think you're better of to go to 5.1. [14:15] that was the most annoying one to fix, besides jit [14:16] ScottK, yep [14:16] Whatever rough edges it has, it'll be an overall improvement. [14:16] +1 [14:18] right [14:18] ok, thanks [14:18] as long as the impact is well understood and all stakeholders are happy, i also feel we should do it [14:18] yeah, that needs to be fixed before switching I guess [14:19] asac, yes, we need a concerted testing plan, then flip [14:21] do we want a formal call for testing (if it didn't happen already)? [14:21] yeah. that would be cool [14:21] asac: can we test custom images already using the same infra used by the official image? [14:22] can we prepare stuff outside and do a binary copy? building qt takes a while, so we reduce the time where we are in an inconsitent state [14:22] I'd like to get it run over all the smoke tests we have [14:22] rsalveti: i would like to see it prepped in ppa, tested and directly/atomically binary copiued over at best [14:22] that's fine, and the builders are way faster nowadays [14:22] tahts tehe safest route i see [14:22] well, that can break the green results :-) [14:22] I wanted a parallel image, and make it tested the same way as the original one [14:22] i think there is some risk acceptable [14:22] to see if there's any other regression in the apps [14:23] if we do our homework, roughly know what to expect (e.g. what can we explain will break) [14:23] Debian has done a lot of work on Qt5 recently. I would also suggest merging their 5.1 packages and then updating to 5.1.1, if they haven't already. [14:23] asac: well, you're stopping CI today because of that issue [14:23] then i am fine [14:23] rsalveti, you think a custom image? [14:23] rsalveti: we dont stop it... we put it in manual/coordination mode [14:23] which means we just go step by step [14:23] pmcgowan: we can easily create a custom image if needed, but just wondering if that will help testing it [14:23] to ensure the big things land in an isolated manner [14:23] for easy blame :) [14:24] at best we manually test with dist-upgrade from the ppa that we will copy from, run some autopilots etc. [14:24] maybe a call for testing using the external ppa is already enough, not sure [14:24] and then if we are happy just binary copy over [14:24] asac: right, but that needs to be coordinated [14:24] Mirv: you can ask me or others in kubuntu to upload [14:24] and we already have the test infra for the default image [14:25] why not reusing that for custom images? [14:25] instead of getting someone doing manual tests [14:25] rsalveti, I will talk to francis as well, maybe the 3 of us can have a call [14:25] Did you review what changes we have that should go to Debian to reduce the diff? [14:25] Riddell: yeah, I think the upload is not an issue, the allowing of upload is they key [14:25] -y [14:25] pmcgowan: sure, sounds good [14:26] rsalveti: lets continue later ... spilled coffee all over here :) [14:26] brb [14:26] lol [14:27] With my release team hat on, it sounds like all the discussion about risk and what's being done to mitigate it could be boiled down to a nice FFe that would be easy to assess. [14:27] Understood about the packages not in Debian, but for the ones that are, it's important to keep the minimum diff for long term maintainability. [14:29] AIUI, the packages that aren't in Debian haven't been formally released by upstream yet, so Debian would actually prefer not to have them. No issue there. [14:29] right [14:29] the maintainability/minimal diff doesn't seem something that has to do with FF [14:30] No. [14:30] Since we're talking about the next three months today, it seems worthwhile to plan for at some point though. [14:33] Does WebGL work on QtWebkit in Qt 5.1? [14:35] Qb1t3r, it should afaik [14:35] Qb1t3r, there is work remaining to enable it in touch [14:36] Question Why not Qt5.1.2 ? [14:36] because there's no 5.1.2 (yet) [14:36] ThanksYouMrMiner: 5.1.1 was released today, so it's a bit hard to use 5.1.2 yet [14:37] alright, thanks! [14:38] thank you, glad to have more clarity [14:47] rsalveti: if the JIT answer was fuzzy, yes all PPA:s (the latest ready one qt5-beta2 / 5.1.1~20130820) have now the JIT enabled [14:48] ah, and adding formal test call to the tasks, balloons can probably help there [14:48] Mirv: thanks === udsbotu changed the topic of #ubuntu-uds-client-1 to: Track: Client | Scopes for 14.04 | Url: http://summit.ubuntu.com/uds-1308/meeting/21904/client-1308-scopes/ [15:01] https://plus.google.com/hangouts/_/72fd9e8502afb6b3ca99d9a43c67acd183ebbfe7?authuser=0&hl=en <- hangout link [15:02] davidcalle: ^ [15:03] sil2100, thanks, installing plugin... -_-" [15:05] Starting the session [15:10] I would like to know the current state of automated user testing, with autopilot. Please add it to the schedule :) [15:13] elopio: ACK :) [15:25] http://pad.ubuntu.com/uds-1308-client-1308-scopes <- for the pad [15:25] thostr_: ^ [15:26] It seems odd to say that we want server scopes for things like Deviant Art. [15:26] Where they'd be much stronger scopes if they could use local accounts. [15:27] i.e. I could search my private deviations [15:29] The problem I have now building the click scope is that some results are available fast (ie, the installed apps) and some are not (ie, the apps available to install, which come from a webservice) [15:30] should I be using different scopes for this, or perhaps I should be using a different api? [15:30] Q: ^ [15:30] sil2100: ^ [15:31] tedg, very true, but this rely on online accounts integration too [15:32] davidcalle, certainly, but that's not too difficult [15:35] Having results jump around while new results come in is really disrupting when you're trying to click/tap something. [15:35] jamesh: great, thanks. [15:35] jamesh: is that with the vala api still, right? [15:36] QUESTION: will the C++ api for scopes be available in 14.04 or 14.10? [15:36] Sound is lagging quite badly for me, not sure I get everything you say [15:36] pstolowski, ^ [15:36] davidcalle: I can hear him perfectly [15:36] alecu: yes [15:37] alecu: ResultSet.flush [15:37] davidcalle: ah. see last note on etherpad [15:37] grat [15:37] great [15:38] pstolowski, thanks. Looking [15:42] QUESTION: I can clearly understand how scopes running in the server return lists of results, but it's not very clear to me if those scopes can also interact with complex previews on the client side (like the scope that let's you type in your credit card or log into your account to purchase music or apps). Is that at all possible, or are interactive previews only available for client side scopes? [15:43] *lets [15:43] <__lucio__> alecu, the previews ive seen just require some piece of data to work. like music previews [15:43] <__lucio__> alecu, we can call those previews and feed them server generated data [15:53] davidcalle: This would need some kind of service that knows what's ok to request [15:53] like libvirtd or something [15:56] right, it will be slow [15:56] I just wanted to understand if it was at all possible or planned === udsbotu changed the topic of #ubuntu-uds-client-1 to: Currently no events are active in this room - http://summit.ubuntu.com/uds-1308/client-1/ - http://irclogs.ubuntu.com/2013/08/28/%23ubuntu-uds-client-1.html [16:01] Thanks everyone [16:02] thanks all! [16:02] such a nice wrap up by thomas :) [16:02] indeed [16:02] thank you! === DanChapman is now known as DanChapman|afk === magicluriv is now known as luriv === udsbotu changed the topic of #ubuntu-uds-client-1 to: Track: Client | Feedback session on mir | Url: http://summit.ubuntu.com/uds-1308/meeting/21936/client-s-1308-mir-feedback/ [18:01] https://plus.google.com/hangouts/_/129ef8287d8f93045020a8dff5ef577b473f200a?authuser=0&hl=en <- hangout link [18:03] jibel: want to join in? ;) [18:05] Session started! [18:05] finally lol [18:06] Can't wait wait for MIR. :3 [18:07] kgunn, what is the sense of the feedback from the call for testing? [18:07] is everything cool for FF? [18:08] Question - how does Mir work with wine? [18:08] gah, I logged out of the server [18:08] i'll like to talk some about what system76 would like to see in the mir input layer, once it's written (that's not in place yet, correct?) [18:08] also, is it possible to get an invite to the hangout? [18:08] does it fullscreen properly? [18:08] kgunn, if you answered my question, I think I missed it [18:09] rickspencer3, you didn't miss it [18:09] olli, kgunn I can't seem to get the youtube video to load now :( [18:10] sil2100, ^^ invite to hangout request [18:10] https://plus.google.com/hangouts/_/129ef8287d8f93045020a8dff5ef577b473f200a?authuser=0&hl=en <- here you go! [18:10] rickspencer3, http://www.youtube.com/watch?v=uhNCVh18vn8#t=306 [18:11] oops [18:11] looks like chromium barfed [18:11] no, i mean once the x input stack is replaced [18:11] hmm, okay [18:11] xubuntu will delay shipping xmir 13.10. Are any other flavors working with it now, or is it going to be a "wait and see" from Ubuntu, then filter to derivatives in +1? Any insights in communications with partners appreciated. [18:12] I'm back [18:12] kgunn, the problem was that chromium was slowly crashing for me :( [18:13] olli: can you join in ^? [18:13] Or is there not much to say? [18:13] sil2100, I think kgunn is doing a great job [18:14] ;) [18:14] rik-shaw, re flavors, the current situation is that only UbuntuKylin is planning to have xmir in 13.10 [18:14] fair response: maybe more practical is do you guys test out other desktops along the way to see where they may stand? [18:15] Will 14.10 have Compiz? [18:15] rik-shaw, we will once a flavor commits to use the technology [18:15] @olli: thanks for the replies [18:15] rik-shaw, it's in our best interest as an upstream to make sure our customers/consumers are covered [18:17] rik-shaw, with regards to other partners, we are talking to a group of partners that we have identified as critical stakeholder in the Mir Ecosystem [18:17] The only flavour that I have not had running on xmir is Ubuntu Gnome [18:17] this group does include ISVs, OEMs, GPU vendors [18:18] and focuses commercial partners [18:18] Jono and team are working with the community partners, such as other upstreams (DEs) and flavors [18:19] proprietary drivers: I heard things are going on with discussions with Nvidia and AMD to support Mir. I also read this may help Wayland as the requirements would be the same. Do you know anything about this (if true)? [18:19] rik-shaw, this is a very sensitive topic as these discussions are under NDA [18:19] so we would prefer to not talk about it at this stage [18:20] olli: thank you for the reply. [18:22] yes, thank you. w/o delving into the wayland discussion, just hoping there is a chance that mir could benefit even though that are skeptical / critical of it. [18:23] Thank you everyone! [18:23] * sil2100 was just a decoration this time [18:23] Thank you [18:23] thanks again. [18:23] thanks! [18:24] thanks! [18:24] * john-mcaleely time for food in the UK === DanChapman|afk is now known as DanChapman === udsbotu changed the topic of #ubuntu-uds-client-1 to: Currently no events are active in this room - http://summit.ubuntu.com/uds-1308/client-1/ - http://irclogs.ubuntu.com/2013/08/28/%23ubuntu-uds-client-1.html === bschaefer_ is now known as bschaefer