[03:55] <robert_ancell> ochosi, you might want to check if your g_unix_signal_add is working correctly in lightdm-gtk-greeter. I just found a bug in unity-greeter where because we return false in the callback the process is terminated with the signal. I changed it to returning true and now it works. Because you've used gtk_main_quit (returns void) I think you might be returning false.
[03:55] <robert_ancell> I noticed it by adding some logging on shutdown (i.e. after the gtk_main) and saw it wasn't being executed
[07:06] <Sweetshark> Happy the-internet-broke-day!
[07:11] <Sweetshark> Ohai, golem.de found my blog and wrote an article about it.
[07:31] <mlankhorst> :o
[07:49] <Sweetshark> mlankhorst: ?
[07:55] <mhr3> larsu, ping?
[07:58] <mlankhorst> morning
[07:59] <larsu> mhr3: what's up?
[07:59] <mhr3> larsu, hey, any chance to expose list of installed schemas to gsettings-qt?
[08:00] <mhr3> larsu, or at least have bool isInstalled(schema)?
[08:01] <larsu> mhr3: like this? https://bugs.launchpad.net/gsettings-qt/+bug/1209235
[08:01] <ubot2> Launchpad bug 1209235 in gsettings-qt (Ubuntu) "Having a way to check if a schemas is available would be nice" [Undecided,New]
[08:02] <larsu> mhr3: I agree that that would be a nice idea. Do you need this urgently or can it wait until after the freeze?
[08:02] <mhr3> larsu, i needed it yesterday :)
[08:03] <mhr3> larsu, but i can work on it, if we just come up with the api we want
[08:03] <larsu> mhr3: ok. Why do you need the list instead of just a isInstalled()?
[08:03] <mhr3> i really just need the isInstalled
[08:04] <larsu> I'm much more comfortable adding that ;)
[08:04] <seb128> good morning desktopers!
[08:04] <Laney> hallo
[08:06] <Laney> hey seb128 Sweetshark mlankhorst larsu mhr3, wie gehts?
[08:06] <mhr3_> grr @ broadcom
[08:06] <mhr3_> larsu, so my first thought was to just add Q_INVOKABLE static bool isInstalled(QByteArray schema_id)
[08:06] <larsu> Laney: great thanks! How are you?
[08:06] <Laney> pretty good!
[08:06] <seb128> Laney, good! (though slightly hurt my back), you?
[08:07] <larsu> mhr3_: yeah ... I guess it would mostly be used with a loadable?
[08:07] <Laney> I went climbing yesterday and almost gave up annoyed because it wasn't going so well
[08:07] <Laney> but then I was like "NO" and carried on, then did two decent new routes
[08:07] <Laney> :-)
[08:07] <larsu> mhr3_: also I don't like the name. GSettings.isSchemaInstalled() is slightly better
[08:07] <Laney> sorry to hear about your back, was it tennis related?
[08:07] <mhr3_> larsu, loadable?
[08:08] <mhr3_> not sure what you mean by that
[08:08] <mlankhorst> Laney: good, you? :P
[08:08] <larsu> mhr3_: sorry, Loader is the right name: http://qt-project.org/doc/qt-5/qml-qtquick-loader.html
[08:08] <Laney> mlankhorst: it's sunny!
[08:08] <mhr3_> Laney, outdoor climb?
[08:08] <Laney> nah, bouldering centre
[08:09] <larsu> mhr3_: a component that loads other components based on a condition
[08:09] <mlankhorst> lucky you, it's rainy here
[08:09] <mhr3_> Laney, ah, what grade was that?
[08:09] <Laney> V5
[08:09] <mhr3_> larsu, yea, now i know
[08:10] <mhr3_> larsu, the other option is to make it non-static, but i'm not sure whether that would make it "better" :)
[08:10] <larsu> ha
[08:10] <Laney> some guys next to me were trying a V7 on which I could barely move
[08:10] <Laney> one day
[08:10] <larsu> mhr3_: you mean have an empty object if the schema is not installed
[08:10] <larsu> instead of ... crashing
[08:10]  * larsu waves fist at desrt
[08:10] <mhr3_> Laney, i sometimes can't do V0 :)
[08:11] <mhr3_> larsu, well, it could just do the check first and then do g_settings_new :)
[08:11] <Laney> practice ;-)
[08:11] <mhr3_> Laney, indeed :)
[08:12] <larsu> mhr3_: I know, but I'm not convinced that that is a good idea (because you don't notice errors right away)
[08:12] <larsu> everything will just be undefined
[08:12] <larsu> an argument _for_ that is that this is basically how js works
[08:12] <seb128> Laney, yeah, happened during training on a simple "change of direction" while moving (not even hitting the ball or anything)
[08:12] <larsu> seb128: what's your opinion from a system settings perspective?
[08:13] <mhr3_> larsu, it would kinda introduce huge api change where you first need to check the instance's .isValid
[08:13] <seb128> larsu, let me read backlog, I'm listening podcast/catching up with email
[08:13] <larsu> ya, same here :)
[08:13] <larsu> mhr3_: or you just go full on dynamic language and ignore it
[08:14] <mhr3_> larsu, also a possibility... not liking it though :)
[08:15] <larsu> me neither. I'm pretty much of desrt's opinion in that regard
[08:15] <larsu> so a static method it is?
[08:15] <mhr3_> larsu, yep
[08:15] <seb128> larsu, what's the suggestion/question?
[08:15] <mhr3_> larsu, i'll do a branch and ping when rdy to review
[08:15] <larsu> mhr3_: cool. I'll hack that up this morning
[08:15] <larsu> oh, awesome
[08:15] <larsu> thanks :)
[08:16] <larsu> seb128: the api bug #1209235
[08:16] <ubot2> Launchpad bug 1209235 in gsettings-qt (Ubuntu) "Having a way to check if a schemas is available would be nice" [Undecided,New] https://launchpad.net/bugs/1209235
[08:16] <mhr3_> will be nice to do something other than scopes :)
[08:16] <seb128> larsu, right, I followed so far, I'm not sure what concrete suggestions you had/asked if that would work for settings
[08:17] <larsu> seb128: options are (1) a statiic method GSettings.isSchemaInstalled(schemaId) or (2) settings.isValid()
[08:17] <larsu> where (2) would entail that the object would return invalid for all keys if it can't find the schema
[08:17] <seb128> (2) looks nicer
[08:17] <larsu> heh, we just decided on (1)
[08:17] <seb128> lol
[08:17] <larsu> because (2) makes it harder to spot errors
[08:17] <larsu> I think we shouldn't crash though, but print a warning instead
[08:18] <seb128> right
[08:18] <seb128> that works for me as well
[08:18] <Laney> does changing the schema work?
[08:18] <larsu> but in the qml code, I think it would help readability if we had an explicit check for the schema if it might not be installed
[08:18] <Laney> It seems weird in a QMLish way
[08:18] <larsu> and don't bother about the check if the schema is a hard dep
[08:18] <larsu> Laney: no.
[08:19] <larsu> Laney: you mean settings.schema = new.schema.name, right?
[08:19] <Laney> I could imagine an isValid property if that works
[08:19] <Laney> yes
[08:19] <didrocks> you know what time it is?
[08:19] <larsu> no, it doesn't
[08:19] <Laney> but if not then yeah, don't give that possibility
[08:19] <didrocks> it's time to remove some binary packages!
[08:19] <didrocks> :)
[08:19] <larsu> Laney: I'd love to make schema a construct-only property, but qml doesn't have those
[08:19] <Laney> all of them?
[08:19] <Laney> indeed, it's not the qml way
[08:20] <mhr3_> larsu, so how do you actually instantiate the Settings component in QML?
[08:20] <larsu> mhr3_: GSettings { id: x; schema.id: "org.gnome.desktop.interface" }
[08:20] <larsu> mhr3_: there's an example in the examples/ dir
[08:21] <mhr3_> larsu, ok, i was only looking at the cpp class, and it's constructor param there
[08:22] <larsu> mhr3_: ya. Someone needed to access it from raw qt too. The qml bindings are in the GSettings dir
[08:22] <mhr3_> in this, i'd rather go with the isValid
[08:22] <mhr3_> this case*
[08:22] <larsu> hm, why?
[08:22] <mhr3_> feels more natural in the qml world
[08:23] <mhr3_> and as you said... undefined is normal in js
[08:23] <larsu> I don't speak qml well enough to know what would feel natural
[08:23] <larsu> so I'll take your word for it
[08:23] <mhr3_> it's not like i was qml expert
[08:23] <larsu> if the settings guys and Saviq agree
[08:23] <larsu> mhr3_: like I said, I can see both sides
[08:24] <larsu> from a purely js perspective, I totally agree with the isValid() thing
[08:24] <larsu> but then I'd love to have what Laney suggested
[08:24] <larsu> not sure how it would be useful, though
[08:24] <mhr3_> larsu, so be able to set new schema id?
[08:24] <mhr3_> sure that should totally work
[08:25] <larsu> yeah, but it won't be worth the effort
[08:25] <larsu> anyway, don't worry about that for now
[08:25] <larsu> we can always add that later
[08:25] <mhr3_> very well
[08:25] <larsu> seb128, Laney: I assume you're okay with the isValid() thing as well?
[08:26] <mhr3_> larsu, and imagine the isValid being fully dynamic and it becoming true when you actually install the schema ;)
[08:26] <seb128> larsu, yes
[08:26] <Laney> larsu: property rather than method, no?
[08:26] <seb128> larsu, I've no strong opinion either way
[08:26] <Laney> so you can bind to it
[08:26] <larsu> mhr3_: lol. Have fun convincing desrt to add a "changed" signal to GSettingsSchemaSource
[08:27] <larsu> Laney: very good point. of course
[08:27] <larsu> mhr3_: ^^
[08:27] <mhr3_> yea +1 for prop instead of method
[08:27] <larsu> api break \o/
[08:28] <mhr3_> addition ;)
[08:28] <larsu> well, and the fact that we stop crashing when you give it an invalid schema
[08:28] <larsu> I presume some people will call this a "bug fix"
[08:29] <mhr3_> larsu, surely desrt won't be one of them :)
[08:29] <larsu> haha, probably not :)
[08:34] <larsu> Sweetshark: ha, nice article ;)
[08:34] <larsu> (the golem.de one)
[08:35] <mhr3_> larsu, i'll do the static for the cpp class though and do the isValid for the qml one, k?
[08:35] <larsu> ya, that's fine by me
[08:37] <Sweetshark> larsu: definitily a high mark measured in 'marketing impact per line of code' ...
[08:38] <larsu> Sweetshark: heh, yeah. Sadly amount of work is not at all proportional to marketing impact
[08:39] <Sweetshark> larsu: (cant compete with the 'marketing impact per line of code' of the wideopenssl/heartbleed thing though)
[08:39] <pitti> seb128: bonjour, ça va ?
[08:39] <seb128> pitti, salut, oui, et toi ?
[08:39] <pitti> seb128: ça va bien, merci ! (j'entends le radio de Canonical maintenan :) )
[08:39] <larsu> Sweetshark: right. Vulnerabilities are much cooler!
[08:39] <larsu> hi pitti!
[08:39] <seb128> pitti, j'écoute aussi le podcast
[08:39] <pitti> seb128: I noticed that rhythmbox-ubuntuone is still in trusty; is that supposed to be, or was it just forgotten?
[08:40] <larsu> probably an oversight - that didn't even work right in the latest versions
[08:40] <seb128> pitti, forgotten I think, let me check
[08:40] <pitti> seb128: oh, 'écouter' est mieux que 'entendre' ici ?
[08:40] <pitti> hey larsu, wie gehts?
[08:41] <seb128> pitti, écouter = listen to, entendre = hear
[08:41] <pitti> ah
[08:41] <larsu> pitti: sehr gut danke!
[08:41] <seb128> pitti, yeah, rb-u1 is an overlook, you can clean it out
[08:41] <larsu> und diR?
[08:41] <pitti> seb128: ack
[08:42] <pitti> seb128: removed; cleaning up autopkgtest jobs, too
[08:43] <seb128> pitti, danke
[08:47] <pitti> larsu: bisschen muede, aber gut, danke!
[09:02] <mhr3_> larsu, hm, what to do about relocatable schemas?
[09:03] <mhr3_> larsu, should isSchemaInstalled check for both?
[09:04] <larsu> mhr3_: yes, but only after someone asks for them for the first time
[09:04] <larsu> (hasn't happened yet)
[09:04] <mhr3_> larsu, or does g_settings_schema_source_lookup check that too?
[09:04] <larsu> that's a question for desrt
[09:05] <mhr3_> hm, i'll just assume it does :)
[09:20] <om26er> seb128, Hi! when is the plan to release next ubuntu-system-settings ? asking so that I could make sure to fix that MR.
[09:20] <om26er> after 14.04 ?
[09:20] <seb128> om26er, not sure, but probably this week
[09:21] <om26er> seb128, also regarding your comment, about the test failing if it takes more than 10 seconds to calculate. We can probably pole if the calculation have been done, for 2 minutes maybe ?
[09:21] <om26er> if it doesn't sound insane
[09:21] <seb128> om26er, I think I put a comment about that on the mp?
[09:22] <seb128> but bottom line is "no"
[09:22] <seb128> first we don't want to make the testsuite take 15 minutes to run
[09:22] <seb128> or people are going to stop running it
[09:22] <seb128> then 2 minutes is not enough if you have a full disk
[09:23] <om26er> *wonders*
[09:25] <om26er> Storage.qml have a 'ready' property which is True when the calculation is done. but again we have to wait in a certain way.
[09:26] <seb128> well, the issue is that we wait on the slowest item to display counts
[09:26] <seb128> the slowest being "space taken by Ubuntu"
[09:26] <seb128> which is basically calling "du" on the userdir I think
[09:27] <seb128> on a dev machine that can take easily some minutes (and that's not on an old rotational disk)
[09:27] <om26er> skip ?
[09:27] <Laney> Yes
[09:28] <seb128> well, or mock the test
[09:28] <seb128> like create small dirs in a temp dir and change the xdg defines
[09:29] <seb128> though that's probably not going to help for the "space taken by ubuntu" issue
[09:29] <seb128> but yeah, skip for now if we don't have a good solution
[09:30] <om26er> seb128, HOME=tempfile.mkdtemp()  ?
[09:30] <Laney> try it with that
[09:30] <Laney> Not sure if you'll have to do the xdg dirs too
[09:30] <seb128> om26er, you can try
[09:31] <Laney> http://paste.ubuntu.com/7220839/
[09:31] <Laney> switching contexts
[09:31] <seb128> Laney, thanks for the bt, can you open a bug with it? I guess it's going to be easier if charles has a look since he knows the codebase
[09:32] <mhr3_> larsu, review pls https://code.launchpad.net/~mhr3/gsettings-qt/fix-1209235/+merge/214703
[09:32] <Laney> Will do, just looking quickly
[09:49] <larsu> mhr3_: needs NULL checks everywhere now :/
[09:49] <larsu> in GSettingsQML
[09:49] <mhr3_> larsu, already handled cause priv->settings will remain null
[09:49] <Trevinho> seb128: so for the lockscreen.... I think we found also a way to make OSK to work, but in case, https://code.launchpad.net/~3v1n0/unity/lockscreen-disable-with-osk/+merge/214653 this just allows to use default gnome-screensaver if any user setting might interfere with that. And that's something we can now handle easily at unity level rather than at gs one
[09:50] <seb128> Trevinho, hey
[09:50] <seb128> Trevinho, nice
[09:50] <larsu> mhr3_: eek you're right - that's already handled for the case that noone sets "schema.id". Nice work, thanks!
[09:51] <mhr3_> larsu, can land with train it with the other changes we needed it for
[09:51] <mhr3_> ...land it with train :)
[09:52] <Trevinho> seb128: this allows to get rid of all the gs patches we did recently, while unity might or might not lock the session.
[09:52] <seb128> Trevinho, what happens if g-s starts before unity ?
[09:52] <seb128> can that happen?
[09:53] <Trevinho> seb128: not in general, as gs is not auto-starting now
[09:53] <Trevinho> seb128: but in case can't do much more than killing it
[09:53] <Trevinho> as I don't know a cleaner way to stop the dbus service
[09:53] <larsu> mhr3_: please do. I'll top-approve as soon as I've heard back from Saviq
[09:53] <seb128> Trevinho, you can take over a dbus service/replace it no?
[09:53] <Trevinho> seb128: well, we might overide its name, but gs is not using gdbus, and this would make things harder to be done
[09:53] <seb128> hum, k
[09:54] <Trevinho> seb128: not that I like this shutdown solution, but it's the only one that it come to my mind
[09:54] <mhr3_> Saviq, pls ^ blocker fix
[09:55] <Saviq> mhr3_, see #ubuntu-unity
[09:55] <Trevinho> seb128: the only other would be to move g-s to upstart initialization
[09:55]  * Saviq not sure why we moved to #ubuntu-unity ;)
[09:56] <Trevinho> seb128: I saw somewhere it's possible to do the same of a dbus-service start file using upstart, so we might use that way... and use again ubus to stop the service, but not sure how it's doable
[09:56] <seb128> Trevinho, yeah, I'm unsure how I feel about it, just too much hack/change for the freeze week
[09:56] <seb128> Trevinho, the less change the better for release
[09:57] <larsu> Saviq: because I think of unity when I want to contact you :)
[09:57] <Trevinho> seb128: to what you refer, to the whole thing or to upstart?
[09:58] <seb128> Trevinho, the screen locking/all those changes
[09:59] <seb128> Trevinho, it feels like lot of moving part and not-tested-in-production-yet changes
[09:59] <seb128> when the freeze is this week
[09:59] <seb128> but I don't have a good idea for a better way out
[10:01] <Trevinho> seb128: i see, but going back would be even worst both for removing something that in 99% works way better than before, and also it would still need quite big changes to get removed. So, the changes have been late, and that's true, but not having to interact with different services made all the process much more simple and eventually fixable
[10:01] <seb128> Trevinho, right, I'm just wondering what we should do next
[10:01] <Trevinho> seb128: so, could we just we try to get a landing ppa, and leave you guys to check how is that?
[10:02] <seb128> we could
[10:02] <seb128> Laney, ^ wdyt?
[10:03] <Laney> Yeah, let's try it out
[10:03] <Laney> is that with g-s fixed to not start on unity?
[10:03] <Trevinho> Laney: yes
[10:04] <Laney> ok
[10:04] <Trevinho> Laney: and if really a11y guys want it, it can be re-enabled by unity (yeah, hackish way but works really well)
[10:04] <Trevinho> Laney: https://code.launchpad.net/~3v1n0/gnome-screensaver/not-autostart-in-unity/+merge/214655
[10:07] <seb128> Trevinho, I see TheMuso opened https://bugs.launchpad.net/unity/+bug/1304111
[10:07] <ubot2> Launchpad bug 1304111 in Unity "UnityTextInputAccessible needs to implement AtkText." [Undecided,New]
[10:08] <Trevinho> seb128: yes, I've read it
[10:08] <Trevinho> andyrock might be more into the thing btw
[10:09] <seb128> k
[10:09] <seb128> I would talk to him, but he's not often online :p
[10:09] <Trevinho> seb128: it still doesn't explain much what's exactly broken, or how is the priority of that (considering how bad the rest of unity is on a11y side - and this is not something I like, but just it has never been a priority)
[10:10] <seb128> right
[10:10] <Trevinho> seb128: yeah, he's moving between home and univesity and he has no connection
[10:10] <Trevinho> seb128: i should meet him soon btw (or in case I'll ping him by sms :))
[10:11] <seb128> ;-)
[10:11] <seb128> well, in any case let's get those changes in a landing slot today
[10:11] <seb128> need to coordinate with bregma though
[10:11] <seb128> I'm unsure what are the plans for unity/compiz landings this week
[10:12] <Trevinho> seb128: do we want in this landing the fallback for the whole a11y to gs for now?
[10:12] <Trevinho> seb128: I mean both if osk and screen reader is enabled?
[10:14] <Trevinho> seb128: as I've supported it only for osk for now (as  orca per se works quite well in it, from my not-so-specialized-on-a11y pov)
[10:14] <Laney> [[10:14] <Laney> [  PASSED  ] 3 tests.
[10:15] <seb128> Trevinho, I don't know enough about the specific, but I would say only in case the screereader is on should be good enough
[10:15] <seb128> Laney, well done!
[10:15] <Trevinho> seb128: ok
[10:16] <Laney> it was easy, he forgot to initialise some pointers
[10:16] <seb128> cool
[10:16] <Laney> the function is weirdly complex but it's not commented so can't tell what it is doing
[10:16] <seb128> well, if the fix makes sense and works, let's land it ;-)
[10:17] <Laney> ya
[10:17] <Laney> test building & MPing
[10:28] <Trevinho> mh, for some reason I got disconnected... anyway
[10:29] <Trevinho> seb128:  so, I leave the ball to bregma; as soon he can provide a ppa I hope you can test it.... I should be offline for next hours as I've to take a flight back, but I'll be online tonight
[10:29] <seb128> Trevinho, ok, have a good trip back!
[10:30] <Trevinho> seb128: thanks
[10:31] <Trevinho> seb128: to summarize, relevant branches are http://pastebin.ubuntu.com/7221006/
[10:33] <mhr3_> larsu, oh, gsettings-qt isn't on train?
[10:33] <Cimi> seb128, mpt, searching "file" on the dash and returning as first result "personal file sharing" (instead of file manager) is a known bug?
[10:34] <seb128> Trevinho, thanks
[10:40] <seb128> Cimi, I don't think we have a defined way to rank exact matches, they both have "file" in their description so match perfectly the string you typed, how would you determine which result should be first?
[10:43] <Laney> seb128: https://code.launchpad.net/~laney/indicator-datetime/initialise-pointer/+merge/214716
[10:45] <seb128> Laney, thanks, approved
[10:45] <Laney> ta
[10:45] <Laney> going to train it?
[10:46] <seb128> yes
[10:46] <Cimi> seb128, I bet file manager should be prioritised over file sharing preferences :)
[10:46] <seb128> let's see if we manage to get a silo, we are running pretty low on those recently
[10:46] <seb128> Cimi, sure, but what technical solution do you recommend to achieve that?
[10:46] <Laney> charles: That split_settings_location function seems pretty hairy for what it's doing
[10:47] <Laney> charles: If I understand it right (display formatting Time/Zones) then g-c-c does the same thing in a simpler way
[10:47] <Laney> g_strdelimit and g_strsplit
[10:51] <larsu> mhr3_: no clue
[10:54] <mhr3_> seb128, you're landed for settings related stuff, was there gsettings-qt landing with train yet?
[10:54] <mhr3_> lander*
[10:57] <seb128> mhr3_, I don't know, seems like you just commited to trunk?
[10:58] <seb128> mhr3_, https://docs.google.com/a/canonical.com/spreadsheet/ccc?key=0Au6idq7TkpUUdC05a2ZQSmgwU2NFYnJQOE9qMDRYa3c&usp=drive_web#gid=1 says it's in CI train
[10:58] <seb128> check with sil2100 or Mirv
[10:58] <mpt> Cimi, I have no idea, sorry
[10:58] <mhr3_> oh, did i push it to trunk?
[10:58] <mhr3_> i hope not
[10:59] <mpt> seb128, all else being equal, something which starts with your search string (File Manager) should rank higher than something that has your search string in the middle (Personal File Sharing)
[10:59] <mhr3_> seb128, nope, it was the bot
[10:59] <seb128> mhr3_, seems like the CI is misconfigured then :/
[10:59] <Mirv> mhr3_: no, the previous gsettings-qt upload was a manual one for the Qt 5.2 upload, the next one should be train
[10:59] <seb128> if it was under CI train the bot should not commit
[11:00] <mhr3_> Mirv, a bot just auto-merged an approved mp
[11:00] <seb128> Mirv, the bot just autocommited a mp that was approved, should that still happen?
[11:00] <Mirv> yeah, it seems it has not been moved in the sense that the automerger would have been disabled
[11:00] <Mirv> I believe robru tried to move everything that was still unmoved at one point, but it may be that the concentrated on the daily release jenkins and not the automerger
[11:01] <Mirv> yes, daily release seems disabled
[11:02] <mhr3_> so, can we fix the mess pls
[11:03] <Mirv> mhr3_: you should probably ping the CI team to disable the auto-merger. the next CI Train release can be done with an empty merge proposal against the trunk, to flush it out.
[11:03] <Mirv> I don't think I have credentials to anything automerger related
[11:59] <xnox> larsu: would you be able to review 6 line diff & screenshot at bug #1304363 ? =)
[11:59] <ubot2> Launchpad bug 1304363 in ubuntu-themes (Ubuntu Trusty) "[UIFe] enable borderless windows under metacity (e.g. for ubiquity)" [High,Confirmed] https://launchpad.net/bugs/1304363
[11:59] <xnox> larsu: pretty please =)
[12:01] <larsu> xnox: "executive decision" makes it sound important
[12:01] <larsu> xnox: approved
[12:01] <xnox> larsu: thank you, sir! =)
[12:28] <jibel> tjaalton could you have a look at bug 1304365 , I believe it is problem with the transition from libwayland0
[12:28] <ubot2> Launchpad bug 1304365 in wayland (Ubuntu) "Precise to Trusty - all of main - fails: Broken transition from libwayland0 to libwayland-client0" [High,New] https://launchpad.net/bugs/1304365
[12:54] <jdstrand> seb128: hey, I was wondering if bugs 1292451 and 1292217 were on your radar. these seems to have a number of dupes and 1292217 seems has security implications (people from my team have anecdotally seen a number of passwords in irc channels lately-- don't know if it is this bug or not)
[12:54] <ubot2> Launchpad bug 1292451 in Unity "screensaver re-locks itself after unlocking if the configured screen-off timer goes off while screen is locked" [High,In progress] https://launchpad.net/bugs/1292451
[12:54] <ubot2> Launchpad bug 1292217 in unity (Ubuntu) "lightdm screen lock has triggered but keyboard is still connected to the main session" [High,Confirmed] https://launchpad.net/bugs/1292217
[12:55] <jdstrand> seb128: if you seen it and are on top of it, cool. I personally haven't seen these, so I can't really help reproduce, but wanted to bring it up in case you didn't see it
[12:59] <seb128> jdstrand, hey, the first one is, the second one was not (and I can't confirm/reproduce)
[13:01] <seb128> jdstrand, hum, the second one was known at the time Andy report it/was fixed, I'm unsure if the recent comment is stating it's still an issue or just mixing it with the double lock bug
[13:01] <dobey> is there a separate process i can kill for the new lock screen?
[13:02] <seb128> no
[13:02] <seb128> it's unity
[13:02] <dobey> bah
[13:02] <dobey> it came back the second time when i unlocked, and there is no entry field for the password
[13:04] <seb128> :/
[13:04] <seb128> try using the session indicator to go the greeter and log back in maybe?
[13:05] <dobey> oh weird. so it was just on one screen that the entry was missing
[13:05] <dobey> moving the mouse to the other screen gave me an entry field
[13:05] <seb128> the entry is supposed to be only on the focussed monitor
[13:05] <seb128> similar to the greeter
[13:05] <seb128> one screen has the ubuntu logo, the other one has the UI
[13:06] <dobey> yeah, it is, but the text field was only displayed when the pointer was on the second screen, not on the first
[13:06] <seb128> weird
[13:06] <dobey> indeed
[13:06] <seb128> that's worth opening a bug on unity I think
[13:06] <dobey> good thing i have two screens :-/
[13:07] <frafu> There is another issue about the lock screen concerning the on-screen keyboard: the on-screen keyboard is not anymore available on the lock screen: https://bugs.launchpad.net/ubuntu/+source/onboard/+bug/1301070
[13:07] <ubot2> Launchpad bug 1301070 in onboard (Ubuntu) "onboard not clickable/working on lock screen" [Undecided,New]
[13:09] <darkxst> this all sounds like dejavu, re-inventing the bugs GNOME had a couple of years ago when they switched to an integrated lock screen ;)
[13:10] <seb128> frafu, that's being worked
[13:17] <frafu> seb128: Do we have to do some adjustments on onboard's side to get it back to the login screen, or do you plan to use onboard's support for the unity-greeter or gnome-screen-saver (which was used until now)? (Assuming that you intend to use onboard as the on-screen keyboard for the lock screen.)
[13:17] <seb128> frafu, login screen?  I though you were speaking about the lock screen?
[13:18] <seb128> there are no known bug with the login screen
[13:18] <seb128> but no chance needed from onboard that I know about, it should work on the unity lock
[13:18] <frafu> sorry: l meant the lock screen
[13:29] <frafu> seb128: Do you happen to know who is working on the issue on unity's side? I might contact him to tell him that he should ping us (the onboard devel team) if there is need for intervention from our side.
[13:29] <seb128> frafu, andyrock I think, but he's not online atm
[13:29] <seb128> Trevinho, ^ do you know?
[13:33] <Laney> andyrock was talking with me about it yesterday
[13:44] <frafu> seb128: No, I don't know him. But he is on this channel. So, in the hope that he is watching:
[13:44] <frafu> Trevinho: Don't hesitate to contact us if you need some intervention on Onboard's side to get it to appear on the lock screen. You can do so for example by using the Answers feature on Onboard's pages on launchpad or by contacting marmuta or me per email (the addresses are available in the authors file of the onboard source).
[13:44] <seb128> frafu, andyrock is not, I was asking Trevinho for confirmation of whether andyrock is looking at those issues
[13:44] <seb128> frafu, I'm going to make sure that whoever is working on that knows they can contact you
[13:44] <seb128> frafu, thanks!
[13:46] <tjaalton> jibel: so what's broken about it?
[13:46] <frafu> seb128: Ok; thanks for passing on the information.
[14:04] <jibel> tjaalton,this upgrade from precise to trusty failed because for some reason the upgrader decides to libwayland0
[14:04] <Laney> decides to what? :)
[14:05] <mlankhorst> ..?
[14:05] <jibel> Laney, to keep :)
[14:05] <jibel> sorry
[14:08] <mvo> jibel: hello, would you be able to do a test upgrade with https://launchpad.net/~mvo/+archive/eglibc-trusty/+packages on amd64 at some point ? if not, no problem, but I don't have a precise-desktop-amd64 VM right now to verify the fix, only i386 and the PPA fails to build on i386 (which appears to be a ppa builder issue, but I digress :)
[14:11] <jibel> mvo, no problem. If you have one minute can you have a look at bug 1304365, I am not sure why libwayland-client0 is held back.
[14:11] <ubot2> Launchpad bug 1304365 in wayland (Ubuntu) "Precise to Trusty - all of main - fails: Broken transition from libwayland0 to libwayland-client0" [High,New] https://launchpad.net/bugs/1304365
[14:12] <jibel> mvo, is there anything specific to verify with this package from your PPA?
[14:14] <tjaalton> i'd have thought all the wayland upgrade issues were sorted ages ago..
[14:14] <Trevinho> frafu: so, also bschaefer was working on it and he should arrive shortly
[14:14] <Trevinho> frafu: the problem with onboard is not that it doesn't show up, but that it doesn't take events
[14:15] <Trevinho> frafu: as initial workaround we're currently iusing gnome screensaver again to lock the screen if that's is enabled, but we'd prefer to get it working correclty
[14:15] <Trevinho> frafu: afaik we found another option that is to make the window to raise in any case...
[14:16] <Trevinho> frafu: but we're still open to get the proper  solution, if you've any.
[14:16] <jibel> mvo, just before you quit I said "mvo, no problem. If you have one minute can you have a look at bug 1304365, I am not sure why libwayland-client0 is held back."
[14:16] <ubot2> Launchpad bug 1304365 in wayland (Ubuntu) "Precise to Trusty - all of main - fails: Broken transition from libwayland0 to libwayland-client0" [High,New] https://launchpad.net/bugs/1304365
[14:16] <Trevinho> frafu: I thought one might be the case that the window is not a dock, and thus compiz might give it handle in a different way than unity windows
[14:17] <mvo> jibel: sure, I have a look now
[14:18] <jibel> mvo, thanks
[14:19] <mvo> jibel: it appears its wrongly blacklisting gnome-session - that is no longer required, correct?
[14:21] <seb128> Trevinho, do you know if bregma is supposed to be online today?
[14:21] <jibel> mvo, I don't think it's related to gnome-session, because if you remove gnome-session then the upgrade fails on libgcr
[14:22] <mvo> jibel: aha, ok
[14:22] <mvo> jibel: then I will look more closely
[14:22] <mvo> still, do we need gnome-session or can it be removed from the "never-remove-this" list?
[14:23] <frafu> Trevinho: Unfortunately, I am not the developer of the team and cannot be of much help with technical questions; marmuta has done most of the implementation of onboard these last years. Do you want him to contact you?
[14:23] <jibel> mvo, to me it seems somethings wrong between libegl1-mesa libwayland0 and libwayland-client0 introduced in saucy
[14:24] <mvo> jibel: something very fishy is going on there indeed
[14:30] <Trevinho> seb128: he is
[14:30] <Trevinho> seb128: but he has hardware problems
[14:30] <seb128> Trevinho, ok, speaking about who
[14:30] <seb128> bregma, hey
[14:31] <seb128> bregma, just as a fyi, I put a landing ask for compiz with ChrisTownsend's recent bugfixes (the main one being "can't use the desktop if a win is open")
[14:33] <Trevinho> seb128: ah, another thing... bregma was saying me that we need to upload directly GS to the silo... can you do that?
[14:33] <Trevinho> seb128: i can try, but I'm travelling right now
[14:33] <Trevinho> seb128: so the net can be unstable
[14:33] <bregma> seb128, thanks, I'm having hardware issues here
[14:33] <seb128> Trevinho, I sure can put a landing ask/upload
[14:34] <seb128> I was waiting for bregma to be around to sync up on what to do for the lockscreen, since that one makes me a bit nervous
[14:34] <seb128> Trevinho, do you have a silo for it already?
[14:34] <bregma> seb128, no silo yet
[14:34] <Trevinho> seb128: I've not the super-powers...
[14:35] <seb128> let me know if you need help/what you need
[14:41] <seb128> dobey, they should be listed in the sidepane?
[14:41] <dobey> seb128: but i see no way to create one on a device
[14:42] <seb128> I guess right clicking on the device entry in the sidebar doesn't do it?
[14:42] <dobey> no, there is no right-clicking on things any more afaict
[14:43] <seb128> k, I don't know then, looks like a rb upstream issue :/
[14:43] <dobey> yeah, i was just hoping there was a non-obvious way that someone knew about :-/
[14:44] <seb128> bregma, ChrisTownsend: https://launchpad.net/~ci-train-ppa-service/+archive/landing-020 has the compiz update, if you want to test it
[14:44] <seb128> bregma, not sure if you do a full autopilot test usually, or just normal user testing
[14:44] <mvo> jibel: can reproduce in a chroot here, I'm investigating further
[14:45] <dobey> i can't create a local playlist and drag it (or even choose to "sync" it in device properties) either :-/
[14:48] <bregma> seb128, I'll just do normal user testing for that change, but on both Unity 7 and Gnome-fallback
[14:51] <mdeslaur> seb128: argh, I have two network indicators again
[14:51] <seb128> mdeslaur, edit the .settings again to drop the desktop profile
[14:51] <seb128> mdeslaur, you should dpkg-divert it so it doesn't come back at every update
[14:51] <mdeslaur> seb128: what .settings?
[14:51] <mdeslaur> sorry, last time you made me uninstall a bunch of packages
[14:52] <seb128> mdeslaur, dpkg -L indicator-network | grep settings
[14:52] <seb128> mdeslaur, oh, that's another solution ;-)
[14:52] <bregma> seb128, shouldn;t the package be fixed so that everyone who installs Unity8 doesn't get two network indicators?
[14:53] <seb128> bregma, unity8 doesn't support indicator-application (which is what nm-applet is using), if you don't pull indicator-network the unity8 session has no network indicator
[14:54] <frafu> Trevinho: Should I tell marmuta to contact you per email about onboard's issues in the lock screen in case he does not show up here? We are also interested in finding a clean solution to the issue. In the meantime, I have mailed him the various approaches you talked about above.
[14:55] <mdeslaur> seb128: qtcreator-plugin-ubuntu pulls in unity8....
[14:55] <mdeslaur> seb128: so everyone that wants to develop is going to get two network indicators now?
[14:56] <seb128> mdeslaur, it feels wrong that the qtcreator is pulling it unity8, even if that didn't mean pulling another indicator
[14:56] <mdeslaur> seb128: ok, I'm confused...where is the .settings file I'm supposed to edit?
[14:56] <seb128> mdeslaur, /usr/share/unity/indicators/com.canonical.indicator.network
[14:57] <seb128> sorry, misremembered the name
[14:57] <mdeslaur> ubuntu-system-settings depends on indicator-network
[14:57] <seb128> mdeslaur, delete the [desktop] section
[14:57] <seb128> mdeslaur, yes, we use the indicator backend for some of the features
[15:03] <mdeslaur> seb128: ok, not sure why a bunch of stuff got pulled in, but managed to uninstall enough to kill indicator-network, thanks
[15:04] <seb128> mdeslaur, yw
[15:04] <seb128> mdeslaur, so it was not qtcreator-plugin-ubuntu pulling unity8 in?
[15:05] <seb128> hum, it does, through unity-scope-tool
[15:06] <mdeslaur> seb128: yes, but I still have unity 8
[15:06] <mdeslaur> seb128: I did: dpkg -P indicator-network ubuntu-system-settings ubuntu-system-settings-online-accounts account-plugin-ubuntuone unity-scope-click
[15:07] <mdeslaur> perhaps some of those are just recommends
[15:07] <seb128> mdeslaur, dpm, sil2100: https://bugs.launchpad.net/qtcreator-plugin-ubuntu/+bug/1297812 is an issue
[15:07] <ubot2> Launchpad bug 1297812 in qtcreator-plugin-ubuntu "Add dependency to unity-scope-tool" [Undecided,New]
[15:08] <seb128> dpm, should I open a new bug or keep using that one?
[15:08] <mdeslaur> ah, right, unity8 has unity-scope-click as a recommends
[15:08] <sil2100> seb128: argh...
[15:08] <seb128> mdeslaur, it's the other way around, unity-scope-click Depends on unity8
[15:09] <seb128> mdeslaur, so installing the qtcreator plugin brings unity8
[15:09] <mdeslaur> IMHO, it still is an issue that bringing in unity8 will display two indicators in unity7...a lot of people are going to want to try unity8
[15:10] <frafu> marmuta: Hi. Did you see Trevinho's approaches about solving the issues concerning onboard and the lock screen?
[15:10] <seb128> mdeslaur, they are going to be disappointed :p
[15:10] <seb128> mdeslaur, unity8 is not really ready for a desktop use yet
[15:10] <mdeslaur> hehe, well, that's beside the point :)
[15:11] <mdeslaur> "Not only did unity8 disappoint, but it broke my unity7, and killed my cat."
[15:11] <sil2100> :(
[15:11] <sil2100> ;)
[15:12] <marmuta> frafu: I probably missed most of the discussion. if it helps, onboard's window hint is either normal or dock on compiz
[15:12] <seb128> mdeslaur, that bug is easy to fix, just drop the desktop section from the indicator file (the one I just made you edit)
[15:12] <seb128> mdeslaur, but then people would install unity8 and say "there is not even a way to connect to a network in unity8"
[15:13] <seb128> bregma, Saviq, is unity8 on desktop using the desktop profiles for indicators ?
[15:13] <sil2100> seb128: should we revert this landing?
[15:13] <mdeslaur> seb128: that just means your fix is wrong :
[15:13] <Saviq> seb128, yes
[15:13] <marmuta> frafu: dock only when docking is enabled, normal is the default
[15:13]  * Saviq can't run the desktop unity8 session for some reason :/
[15:13] <seb128> mdeslaur, what's the right fix?
[15:14] <mdeslaur> seb128: don't you have a whole team to come up with fixes like that? :)
[15:14] <seb128> Saviq, do you have a solution to keep a network indicator in unity8/desktop without having 2 under unity7?
[15:14] <Saviq> seb128, no
[15:14] <seb128> :-(
[15:14] <bregma> seb128, yes, the unity8-desktop-session forces the desktop profile for indictors
[15:14] <Saviq> seb128, unless we name the desktop profile differently
[15:14] <seb128> mdeslaur, we would if the team was not busy debugging issues created by apparmor :p
[15:15]  * mdeslaur hands seb128 a violin
[15:15] <mdeslaur> :;
[15:15] <Saviq> seb128, like unity8-desktop or something, not sure it's a good enough solution though
[15:15] <mdeslaur> ;)
[15:15] <seb128> sil2100, not sure about reverting, let's wait to hear from dpm and bzoltan
[15:15] <seb128> Saviq, yet another profile, you are going to make larsu cry
[15:16] <larsu> no.
[15:16] <seb128> bregma, silo 020 looks fine to me (tested unity session only, not fallback though)
[15:17] <seb128> bregma, larsu: do you know if can blacklist an indicator from unity-panel-service/unity7?
[15:18] <larsu> seb128: from where? As a user?
[15:18] <seb128> larsu, no, from the code
[15:18] <larsu> you can just remove the desktop profile from the indicator...
[15:18] <larsu> indicator file, I mean
[15:18] <frafu> marmuta: please have a look at your email; I now sent you the whole discussion.
[15:18] <seb128> larsu, <seb128> Saviq, do you have a solution to keep a network indicator in unity8/desktop without having 2 under unity7?
[15:19] <seb128> larsu, the issue is that both unity7 and unity8 use the desktop profile
[15:19] <marmuta> frafu: reading it right now, thanks
[15:19] <seb128> larsu, unity7 uses nm-applet through indicator-application but indicator-application doesn't work under unity8
[15:19] <larsu> seb128: the right fix there is to finally drop nm-applet
[15:19] <seb128> right
[15:19] <seb128> larsu, want to ask for the ffe to replace it by the new one for trusty? ;-)
[15:19] <larsu> is anyone working on bringing indicator-network up to speed?
[15:19] <Laney> bahaha
[15:19] <larsu> seb128: lol
[15:20]  * larsu would rather eat a hat
[15:20] <seb128> larsu, yes, that's not going to happen before the lts though
[15:20] <bregma> seb128, silo 020 passes my tests here, it's good to land
[15:20] <larsu> and we need a working unity8 on the lts?
[15:20] <seb128> larsu, meanwhile users install unity8 to play with it and wonder why they have 2 indicators in unity7
[15:20] <larsu> ahhhhh
[15:20] <seb128> bregma, great, doing that
[15:20] <larsu> now I get the problem :)
[15:20] <seb128> larsu, well, the other option is to disable the new one/drop the desktop profile, but then users are going to say "unity8 doesn't even let you connect to a network"
[15:21] <larsu> right
[15:21] <Laney> aren't they upstart managed?
[15:21] <frafu> Trevinho: is marmuta's sentence about the window hint of any help?
[15:21] <larsu> our options are:
[15:21] <Laney> just add a line to the job to quit if it's not unity8
[15:21] <larsu> (1) yet another profile
[15:22] <larsu> (2) blacklist com.canonical.indicator.network in u-p-s
[15:22] <larsu> (3) put the indicator file for the network indicator into a place where unity8 finds it but unity8 doesn't
[15:22] <larsu> do unity7 and 8 share /usr/share/unity?
[15:22] <seb128> (4) what Laney suggests
[15:22] <seb128> which might be the easiest
[15:23] <larsu> and scariest
[15:23] <larsu> but yeah, that would totally work
[15:25] <seb128> tedg, ^ can you get that happening? ;-)
[15:25]  * tedg reads backlog
[15:26] <tedg> Oh, that's in the indicator-network startup branch.
[15:26] <seb128> lol
[15:26] <Laney> hahaha
[15:26] <tedg> It just looks at pre-start and stops if not unity8
[15:26] <tedg> https://code.launchpad.net/~ted/indicator-network/startup-cleanup/+merge/207771
[15:27] <Laney> did xubuntu get that sorted?
[15:27] <tedg> Oh, that has the same == problem.
[15:29] <Laney> don't you want !=?
[15:29] <tedg> Yes
[15:29] <Laney> nod
[15:29] <tedg> Laney, Last I heard they were figuring out exactly what they wanted, but were generally in good shape.
[15:30] <seb128> hum
[15:30] <seb128> it's meeting time!
[15:30] <Laney> oh
[15:30] <Laney> whoops
[15:30] <seb128> qengho, Sweetshark, mlankhorst, Laney, tkamppeter, desrt, larsu: hey
[15:30] <seb128> Laney, start writing your week summary :p
[15:31] <Laney> kenvandine?
[15:31] <desrt> seb128: me too :)
[15:31] <mlankhorst> howdy
[15:31] <seb128> Laney, indeed, thanks
[15:31] <seb128> kenvandine, hey as well ;-)
[15:31] <seb128> ok, let's get start
[15:31] <seb128> qengho, hey
[15:31]  * kenvandine waves
[15:32] <seb128> started even
[15:32] <larsu> oh. meeting. Is it Tuesday already?
[15:32] <seb128> larsu, indeed is!
[15:32] <seb128> no qengho?
[15:32] <seb128> ok...
[15:33] <seb128> Sweetshark, hey
[15:33] <Sweetshark> seb128: aye
[15:33] <Sweetshark> - LibreOffice-4.2.3~rc3 packaged for trusty
[15:33] <Sweetshark> - openssl fallout handling (its bundled upstream)
[15:33] <Sweetshark> - Document Liberation project started: http://www.documentliberation.org/
[15:33] <Sweetshark> - fixed accessodf extension for API change
[15:33] <Sweetshark> - some tricky details on API reference versions for extensions between AOO/LO upstream
[15:33] <Sweetshark> - 3.5.7-0ubuntu6 for precise uploaded for sponsoring (seb sponsoring. now in unapproved)
[15:33] <Sweetshark> - adjusted/push upstream release plan for earlier alpha1 builds (moved 3-4 weeks earlier), should give us better/earlier testing with tagged release (that I have a good rationale to push to the libreoffice-prereleases PPA and ask for testing)
[15:33] <Sweetshark> EOF
[15:34] <qengho> Meeting!
[15:34] <Sweetshark> qengho: impersonating me today?
[15:35] <seb128> Sweetshark, I guess we want another upload before release. any chance we get 4.2.3 release? you mentioned a KDE issue, you have a patch for it right? what about the menus one (bug #1296715)
[15:35] <ubot2> Launchpad bug 1296715 in libreoffice (Ubuntu) "Menu items are greyed out in Libreoffice menu." [High,Confirmed] https://launchpad.net/bugs/1296715
[15:36] <Sweetshark> oh, yeah: on 4.2.3 I could reproduce the nasty KDE bug and reverted the offending commit already in git. However, I just got a report "works fine now, thanks" WITHOUT the revert *confused*
[15:37] <frafu> seb128: Could you please tell us what package includes the lock screen?
[15:37] <seb128> frafu, unity
[15:37] <Sweetshark> bug 12967715 will be a frustrating bughunt that Ill do ~this week ...
[15:37] <seb128> Sweetshark, ok, thanks
[15:37] <seb128> Sweetshark, thanks
[15:37] <seb128> qengho, your turn
[15:38] <Laney> haha
[15:39] <seb128> qengho vanished again?!
[15:39] <seb128> bah
[15:40] <seb128> mlankhorst, your turn
[15:40] <mlankhorst> thursday sick, upstream kernel hybrid work, testing trusty with various 'accelerated' virtualization solutions i got my hand on (qemu spice, vmware, TODO: qemu bochs, virtualbox), uploaded new xserver-xorg-video-vmware, considering uploading xxv-synaptics 1.7.4+git, it might alleviate the need for a FFe.
 hi
[15:40] <seb128> Laney, yeah, "again" :p
[15:40] <mlankhorst> and gave up hope of keeping bug queue manageable, just looking at some bugs now :P
[15:40] <seb128> mlankhorst, do you use e.u.c ?
[15:40] <Laney> 'need' for FFe> final freeze in two days, not so likely
[15:40] <mlankhorst> seb128: yeah fixed some
[15:40] <Sweetshark> seb128: maybe freenode is having a severe case of heartbleed today ...
[15:41] <mlankhorst> https://errors.ubuntu.com/?user=ubuntu-x-swat
[15:41] <seb128> mlankhorst, bug #1033533 is ranked quite high
[15:41] <ubot2> Launchpad bug 1033533 in xserver-xorg-video-nouveau (Ubuntu Raring) "Xorg crashed with SIGABRT: exaMemcpyBox with src=0x0 on nouveau with SW rendering" [High,Triaged] https://launchpad.net/bugs/1033533
[15:41] <mlankhorst> seb128: yeah but that bug = nouveau locked up, system is hosed
[15:41] <mlankhorst> nothing I can do :/
[15:41] <seb128> :-(
[15:42] <seb128> qengho, mlankhorst: is any of you looking at bug #1303273 ?
[15:42] <ubot2> Launchpad bug 1303273 in chromium-browser (Ubuntu) "chromium-browser crashed with SIGABRT in __assert_fail_base()" [Medium,Confirmed] https://launchpad.net/bugs/1303273
[15:42] <seb128> brw_get_graphics_reset_status assert
[15:42] <seb128> #5  0xae489208 in brw_get_graphics_reset_status (ctx=0xb831805c) at ../../../../../../../src/mesa/drivers/dri/i965/brw_reset.c:43
[15:42] <seb128> "brw->hw_ctx != ((void *)0)"
[15:42] <mlankhorst> I saw that bug, but haven't hit it
[15:43] <seb128> it happens every time I start chromium here
[15:43] <seb128> so if somebody needs debug info...
[15:43] <mlankhorst> oh that changes things
[15:43] <mlankhorst> glxinfo?
[15:43] <Laney> after the meeting!
[15:43] <seb128> the reports are all i386 it seems
[15:43] <mlankhorst> ah
[15:43] <seb128> so it might be 32 bits specific
[15:43] <seb128> yeah, after the meeting
[15:43] <seb128> mlankhorst, thanks
[15:44] <seb128> mlankhorst, it gives you some time to look at it while the meeting continue ;-)
[15:44] <seb128> Laney, qengho did ping back, your turn I guess then
[15:44] <Laney> • u-s-s: try to optimise the time-date search without creating a new thread (QSortFilterProxyModel might have to go or be reworked...)
[15:44] <mlankhorst> or did he :P
[15:44] <Laney> • track down upstart bug setting the wrong umask after re-execing, causing a click hook to create a directory with the wrong permissions & then a crash on later logins
[15:44] <Laney> • Fix some denials in tp-mission-control
[15:44] <Laney> • Discussion about gstreamer 0.10 / IM network support on desktop, keep what we have & revert empathy that dropped them
[15:44] <Laney> • Fix ubuntuone-credentials for cross-building which was blocking u-s-s
[15:44] <Laney> • Update webkitgtk to final, blocked in proposed for manual testing (should be done today, help if you want)
[15:44] <Laney> • Fix a weird activity-log-manager bug where it was indefinitely starting/stopping whoopsie-preferences
[15:44] <Laney> • indicator-datetime arm64 testsuite failure fix
[15:44] <Laney> • Continued discussions/lobbiying about wallpaper selection dropping in unity8
[15:44] <Laney> • Patch pilot, DMB, FFes and reviews * a lot
[15:44] <Laney> ⚛
[15:46] <Laney> https://bugs.launchpad.net/upstart/+bug/1302117 btw
[15:46] <ubot2> Launchpad bug 1302117 in upstart "Session Init changes umask on re-exec" [Undecided,Confirmed]
[15:46] <qengho> Wow, really bad network suddenly.  seb128, I'm here again.
[15:46] <seb128> Laney, thanks
[15:46] <seb128> qengho, hey, ok, your turn
[15:46] <tjaalton> seb128: you have ironlake? it's probably the same bug that prevents kwin fom working, there's a preliminary patch for it
[15:47] <seb128> tjaalton, I think do (i5)
[15:47] <seb128> great
[15:47] <seb128> mlankhorst, ^
[15:47] <qengho> done: more chromium-browser testing.
[15:47] <qengho> in-progress: grow more antsy that upstream hasn't released Cr34 yet.
[15:47] <qengho> to-do: finish a few bugs. Worst ones: Tapping on menu button reveals a menu that doesn't respond to more touch events. Menu shifted over on non-1 scaling.
[15:47] <tjaalton> yeah gen4 & 5 affected
[15:47] <qengho> EOF.
[15:48] <seb128> qengho, thanks
[15:48] <seb128> tkamppeter, hey
[15:48] <tkamppeter> - cups-filters: Released 1.0.51 and 1.0.52 with fixes for newly reported bugs: Security fix against arbitrary code injection into interface scripts automatically generated when network printers are discovered, fixes in texttopdf especially to make it robust against PPDs with broken page geometry information.
[15:48] <tkamppeter> - foomatic-db: Uploaded newest snapshot to support latest Samsung PostScript printers, and to apply several fixes to the PPD file for the Fuji Xerox DocuPrint CM305 df, especially broken page geometry info.
[15:48] <tkamppeter> - Checked through and cleaned up my collection of sample printer hardware. Removed devices which have died and ordered ink for others.
[15:48] <tkamppeter> - Bugs.
[15:49] <seb128> tkamppeter, thanks
[15:50] <seb128> desrt, your turn
[15:50] <desrt> went to freedesktop summit last week and had a lot of good discussions
[15:50] <desrt> our XDG_CURRENT_DESKTOP stuff is now enshrined in the desktop entry spec and mimeapps spec
[15:50] <Laney> how's your 2048?
[15:50] <desrt> i'm currently reimplementing the new version of the mimeapps spec in GIO -- it will give a consistent mechanism (cross desktop) of specifying default apps for given mimetypes, which is something we lack until now
[15:51] <desrt> so for example, if we have a unity default of gedit for text/plain, Qt apps will now start this, instead of following the KDE priority-numbering scheme (which would have gotten us kate if it was installed)
[15:51] <seb128> desrt, is the default by desktop?
[15:52] <desrt> we also have some agreements about how we want to deal with startupnotification in the post-X11 future, etc.
[15:52] <seb128> where/how are those specified?
[15:52] <desrt> seb128: there are two files
[15:52] <desrt> mimeapps.list specifies global defaults, but there is also the possibility of unity-mimeapps.list taken from the lowercased value of XDG_CURRENT_DESKTOP
[15:52] <desrt> this allows per-desktop overrides
[15:52] <seb128> cool
[15:53] <desrt> you can even get fancy with respect to things like XDG_CURRENT_DESKTOP=GnomeClassic:GNOME
[15:53] <desrt> and have gnomeclassic do a couple of overrides but mostly take the same as GNOME
[15:53] <seb128> is there a way to override the override? ;-)
[15:53] <desrt> vendor-patch style, you mean?
[15:53] <desrt> we deliberately excluded this possibility because of the fact that desktop file caches are strictly caches
[15:54] <desrt> which would have turned the vendor override mechanism into a runtime overhead
[15:54] <desrt> (compare gsettings where the cache file is mandatory, so there is never a runtime overhead to gschema override files)
[15:54] <desrt> if we ever move to a place where the desktop file index is mandatory, we can talk about a vendor override scheme, but not before then
[15:54] <seb128> I was mostly asking for e.g oems install
[15:54] <desrt> ah -- well, this is per XDG_DATA_DIRS
[15:54] <seb128> some time those guys want to take an Ubuntu image, put e.g acroread on it and make it default over evince
[15:55] <seb128> without changing our packages
[15:55] <desrt> and i understand that in the OEM case, they have an extra XDG_DATA_DIR added
[15:55] <seb128> right
[15:55] <desrt> so of course they'll be able to do it from there -- but they could have done that already
[15:55] <seb128> k
[15:56] <desrt> Laney: this game is seriously boring
[15:56] <desrt> larsu did some rather extensive research and definitively proved that the best strategy consistent of senseless buttonmashing
[15:56] <desrt> *is consistent of
[15:56] <seb128> lol
[15:56] <larsu> desrt: extensive?
[15:56] <Laney> I look forward to reading the trip report
[15:56]  * seb128 does as well
[15:56] <seb128> desrt, thanks
[15:56] <desrt> larsu: i read it on bastien's blog... must have been true :)
[15:57] <seb128> larsu, your turn!
[15:57] <larsu> oh right, meeting.
[15:57] <larsu> I think desrt already mentioned what we did at the summit last week
[15:58] <larsu> I've added support for XDG_CURRENT_DESKTOP to gdm and lightdm
[15:58] <larsu> (where lightdm only needed the X-LightDM prefix removed)
[15:58] <desrt> bye bye vendorpatches
[15:58] <desrt> (and X-Lightdm-Namespaced keys)
[15:58] <larsu> I also helped Antti with a unitymenumodel/qml problem
[15:59] <larsu> fixed the icon in indicator-power when the icon theme changes
[15:59] <larsu> (or in the case of seb128's slow computer, doesn't change quickly enough)
[15:59] <larsu> this week I've been working on indicator-messages bugs
[16:00] <larsu> mostly small stuff, like the chat section not disappearing when there are no apps using it
[16:00] <larsu> and the whole menu no disappearing when no apps are using it :)
[16:00] <larsu> right now, I'm trying to find a solution for the totally broken empathy integration
[16:01] <seb128> larsu, (stop insulting my nice i5 :p)
[16:01] <larsu> (messages either don't show up or show multiple times)
[16:01] <larsu> I've had some approaches, but none that worked really well
[16:01] <larsu> the interplay between telepathy/empathy/telepathy-indicator/chat-window-process is really ... messy
[16:01] <larsu> anyway. that's it from me
[16:02]  * Sweetshark s Big Bertha gives seb128 i5 a cookie: "May you be big and strong one day!"
[16:02] <seb128> lol
[16:02] <seb128> larsu, thanks
[16:02] <larsu> seb128: at least you don't need to compile LO
[16:02] <seb128> indeed
[16:02] <seb128> nor webkit!
[16:02] <kenvandine> :)
[16:02] <larsu> \o/
[16:02]  * seb128 is glad Laney took over this one
[16:02] <larsu> Laney's computer sure isn't
[16:02] <Laney> took over HA
[16:03] <seb128> HA?
[16:03] <Laney> I use PPAs these days :P
[16:03] <seb128> that's cheating :p
[16:03] <seb128> kenvandine, your turn ;-)
[16:03] <kenvandine> content-hub:
[16:03] <kenvandine>  * some work in progress for download/uploads
[16:03] <kenvandine>  * Added snap decision to open downloaded, interim solution until we have the transfer indicator
[16:03] <Laney> because I am incapable of not rebooting or closing the window without having used screen or killing the build or any other number of incompetent things
[16:04] <kenvandine> i think that's all i have to report
[16:04] <kenvandine> the long weekend has wiped my brain
[16:05] <seb128> kenvandine, k, thanks ;-)
[16:05] <seb128> ok, my turn
[16:05] <seb128>  * desktop updates (totem-pl-parser, gdk-pixbuf, rhythmbox, glibmm, gtk)
[16:05] <seb128>  * updated shotwell to correctly call the online account panel
[16:05] <seb128>  * backported some warning fixes for totem
[16:05] <seb128>  * debugged nautilus-share not working, ended up to be some Ubuntu dropped during the samba4 merge, brought that back
[16:05] <seb128>  * some patch piloting
[16:05] <seb128>  * cleaned up things for the u1 shutdown
[16:05] <seb128>  * fixed gnome-icon-theme-extras being not installable
[16:05] <seb128>  * handled landings through CI train (mostly indicators and settings)
[16:05] <seb128>  * discussions about the lockscreen changes
[16:05] <seb128>  * lot of bugs triage and testing of changes for others
[16:06] <seb128> (Ubuntu *delta* dropped*)

[16:07] <seb128> is there any other topic/question?
[16:07] <Sweetshark> addendum: https://jenkins.qa.ubuntu.com/job/trusty-adt-libreoffice/322/? is all green for the first time -- not complaining about a spurious stderr output nor running out of discspace on the VM \o/
[16:07] <Laney> weeeee
[16:07] <seb128> Sweetshark, well done!
[16:07] <seb128> thanks to jibel as well I guess ;-)
[16:08] <Sweetshark> seb128: indeed: thanks jibel!
[16:08] <Laney> u-c-c migrated too
[16:09] <seb128> thanks everyone
[16:09] <seb128> Laney, \o/
[16:09] <seb128> mlankhorst, tjaalton: so the chromium issue is being tracked on your side, do you need testers for a fix?
[16:11] <tjaalton> seb128: https://launchpad.net/~tjaalton/+archive/test
[16:11] <tjaalton> it's https://bugs.launchpad.net/ubuntu/+source/mesa/+bug/1299499
[16:11] <ubot2> Launchpad bug 1299499 in mesa (Ubuntu) "kwin crashes on desktop startup with wobbly windows enabled" [High,Confirmed]
[16:12] <seb128> tjaalton, k, thanks
[16:12] <seb128> tjaalton, what is blocking upload to trusty?
[16:12] <tjaalton> i've been trying to poke upstream to commit something..
[16:13] <tjaalton> found out that mesa-dev had a proposed patch a month ago that never got anywhere
[16:13] <tjaalton> replied to the thread, still nothing
[16:13] <tjaalton> anyway, the patch is a oneliner that shouldn't break anything..
[16:13] <tjaalton> the new one that is
[16:15] <seb128> just upload! ;-)
[16:17] <tjaalton> yeah..
[16:19] <tjaalton> done
[16:19] <seb128> tjaalton, thanks
[16:20] <seb128> tjaalton, is https://bugs.launchpad.net/oem-priority/+bug/1287341 something you would be interested to look at by any chance? ;-)
[16:20] <ubot2> Launchpad bug 1287341 in unity-settings-daemon (Ubuntu) "Touchscreen controls both screens in dual-monitor setup" [High,Triaged]
[16:20]  * seb128 wonders if we have anyone who knows about the input system/touch enough to be able to do the work required there easily
[16:21] <seb128> mlankhorst doesn't feel comfortable looking at the settings daemon
[16:21] <seb128> others don't know much about the "input matrix"
[16:21] <tjaalton> ugh :)
[16:23] <mlankhorst> actually I mostly lack the hardware :p
[16:24] <mlankhorst> seb128: my guess is the transform multiplies (x, y, 1) with the matrix to get the output coordinates
[16:25] <mlankhorst> can see it when you do xinput, then see something like this: ↳ Saitek Cyborg R.A.T.9 Wireless Mouse      id=9    [slave  pointer  (2)]
[16:25] <mlankhorst>  xinput list-props 9
[16:25] <mlankhorst> Coordinate Transformation Matrix (143): 1.000000, 0.000000, 0.000000, 0.000000, 1.000000, 0.000000, 0.000000, 0.000000, 1.000000
[16:25] <seb128> mlankhorst, https://bugzilla.gnome.org/show_bug.cgi?id=698284 wrote
[16:25] <ubot2> Gnome bug 698284 in xrandr "Missing feature causes weird behavior when using touchscreen when two (or more) monitors are present" [Normal,Unconfirmed]
[16:25] <seb128> "we don't map touchscreens to an output, we map it to a portion of the desktop."
[16:25] <seb128> " don't know how to fix this other than a database in the style of libwacom"
[16:26] <tjaalton> that doesn't sound too promising..
[16:26] <mlankhorst> probably needs to happen :/
[16:27] <bregma> big problem with a generalized solution to that touchscreen problem is that there is no association between the touch device and the display
[16:27] <mlankhorst> yeah..
[16:28] <bregma> not even at the kernel level
[16:29] <mlankhorst> but for now assuming touchscreen maps to the eDP or integrated screen would be best bet :/
[16:31] <bregma> it would makea reasonable starting default but isn't anything like a good solution since there are lots of external touchscreens out there
[16:32] <seb128> yeah
[16:32] <mlankhorst> how are they handled on other platforms?
[16:32] <seb128> it would also mean that if I dock my 3 years non-touch laptop on an external touch monitor things would be the wrong way around
[16:32] <tjaalton> but confirmed..
[16:33] <tjaalton> *bug
[16:33] <mlankhorst> tjaalton: which?
[16:33] <tjaalton> the touch one
[16:33] <mlankhorst> ah
[16:33] <tjaalton> i've got this lovely harris beach sdp..
[16:33] <bregma> mlankhorst, how are they handled on other platforms? poorly, just like high-DPI
[16:34] <tjaalton> now excuse me while it dist-upgrades trusty->trusty for the next two hours..
[16:34] <tjaalton> hmm I have win8 on this too
[16:36] <mlankhorst> could we make it a ui thing perhaps?
[16:37] <mlankhorst> let the user suffer for next few years until things standardize :/
[16:37] <mlankhorst> anyway I really g2g!!
[16:37] <mlankhorst> bb
[16:39] <tjaalton> looks like works just fine on win8
[16:39] <tjaalton> touch is duplicated when the screen is duplicated
[16:40] <tjaalton> but not when it's extended
[16:40] <tjaalton> and when only the external screen is used, touch events go there
[16:45] <Laney> https://www.youtube.com/watch?v=I9mnUl9v1wI#t=17
[16:45] <Laney> look at that sexy application
[16:45] <CrazyLemon> hey guys... i'm getting "weird" behavior from nautilus. If i open trash, minimize it there is a arrow on the left side of nautilus launcher icon showing open window. When i click on that icon ANOTHER window opens. Is this expected behavior? I'd prefer to open already opened window (trash in this case)
[16:46] <Laney> launches suspiciously fast :-)
[16:46] <bschaefer> CrazyLemon, https://launchpad.net/bugs/1163041
[16:46] <ubot2> Launchpad bug 1163041 in unity (Ubuntu) "Inconsistent behaviour with minimized Trash windows" [Low,Triaged]
[16:47] <bregma> seb128, can you upload gnome-screensaver with https://code.launchpad.net/~3v1n0/gnome-screensaver/not-autostart-in-unity/+merge/214655 once I get a silo for line 65?
[16:47] <Laney> also IRC isn't a good medium for reporting bugs
[16:47] <CrazyLemon> bschaefer cool.. thanks for pointing that out :)
[16:47] <bschaefer> CrazyLemon, yup, i happened to have been poked about it yesterday, so i had it on hand
[16:47] <bschaefer> CrazyLemon, we are trying to get a fix out for it, but final freeze is soon :(
[16:48] <seb128> bregma, yes
[17:13] <bregma> seb128, landing-006 if you please
[17:13] <seb128> bregma, on it
[17:33] <beidl> CrazyLemon: ha! nice, I thought I was the only one noticing that bug :)
[17:33] <CrazyLemon> beidl you are not alone! :>
[17:34] <beidl> CrazyLemon, I assumed I could fix that bug myself, but man that code base is big/complicated :D
[17:35] <seb128> bregma, what, it seems like I can't upload to that ppa either
[17:35] <bregma> seb128, how does it feel to be weak and helpless again?
[17:35] <seb128> bregma, that can't be happening to me!!!
[17:39] <seb128> bregma, sorting it one on #ubuntu-ci-eng
[17:42] <seb128> tedg, ogra_: https://launchpad.net/~ci-train-ppa-service/+archive/landing-003/ if you want to test the indicator-messages update (include upstart job)
[17:45]  * tedg adds repo
[17:45] <tedg> Is there a way to list the packages that no longer have a repo on your system?
[17:45] <tedg> Trying to purge a PPA that I deleted.
[17:48] <seb128> tedg, not that I know about
[17:54] <mdeslaur> tedg: apt-show-versions | grep "No available"
[18:01] <jose> jono: ping, I see some tech probs on the livestream?
[18:02] <jose> whoops, wrong channel
[18:02] <tedg> mdeslaur, Neat, now I'm kinda scared about how big the list is…
[18:03] <mdeslaur> tedg: you've been upgrading since dapper, haven't you? :)
[18:03] <tedg> mdeslaur, Heh, no, this machine came with 12.04 though
[18:03]  * mdeslaur has a bunch of Ubuntu gOne stuff in there
[18:17] <ogra_> seb128, yep, thanks
[18:21] <seb128> ogra_, yep works fine, or yep you are going to try it? ;-)
[18:21] <ogra_> seb128, on it (on the phone atm)
[18:21] <seb128> ogra_, k, I tested the desktop so phone is enough
[18:22] <ogra_> great
[18:22]  * ogra_ twiddles thumbs watching apt-get update ...
[18:25] <ogra_> seb128, looks fine, starts, shows an SMS is sent to the phone
[18:25] <seb128> ogra_, great, thanks
[18:25] <seb128> let me publish then
[18:25] <ogra_> \o/
[18:25] <ogra_> thanks for taking that
[18:25] <seb128> yw!
[18:36] <seb128> bregma, unity failed to build in landing 006
[18:37] <seb128> "Settings schema 'org.gnome.desktop.a11y.applications' does not contain a key named 'screen-reader-enabled'"
[18:37] <seb128> Trevinho, ^
[18:55] <beidl> if someone is interested in my findings regarding the launcher just suddenly appearing instead of sliding in using the 4-finger swipe gesture:
[18:56] <beidl> in Launcher.cpp, the events OnDragStart/Update/Finish get called correctly, and the drag_out_delta_x_ gets a usable value assigned with every update.
[18:57] <beidl> but it seems that RenderArgs() is not getting called, so I assume that (if I understand this correctly) the drawing doesn't get queued up when swiping the launcher in. only when swiping the launcher out.
[18:57] <seb128> Trevinho, bregma, bschaefer: ^ that might interest you
[18:57] <bschaefer> seb128, thanks!
[18:58] <bschaefer> beidl, that sounds like what would happen
[18:58] <bschaefer> beidl, the Launcher.cpp is another monster class, that was written before we really were using nux (most of the logic bits)
[18:58] <bschaefer> soo its very disconnected vs a lot of other classes we have
[18:58] <bschaefer> beidl, hopefully Trevinho can check that, i don't have  a way to do 4-swipe
[18:59] <beidl> disconnection isn't bad depending on the definition, I guess :)
[19:00] <bschaefer> its in a bad way :)
[19:00] <bschaefer> re-inventing bits of the wheel that nux solves
[19:00] <beidl> I guess most of you can't wait to re-start that whole thing with unity8 :D
[19:01] <beidl> I've just basically checked this by adding a hand full of couts in the code. events get recognized correctly, (probably) right values are assigned, etc. But QueueDraw only seems to actually do correct drawing when pushing the launcher away.
[19:02] <bschaefer> beidl, are we detecting the opposite?
[19:02] <bschaefer> if so, can you force the renderering?
[19:02] <bschaefer> or force a queue draw
[19:02] <bschaefer> beidl, also just to check, the bug you're running is swiping the launcher in, does not work?
[19:02] <bschaefer> but out does?
[19:04] <beidl> to rephrase myself: when dragging the launcher in using 4 fingers, there is no slide in animation. to be more precise:
[19:05] <beidl> 1) either I drag the fingers slowly until the drag_delta_x reaches its maximum of 300 and the launcher suddenly pops up, no slide effect
[19:05] <bschaefer> i wonder if our edge barrier code messed all that fun stuff up... hmm
[19:06] <beidl> 2) or I swipe with a certain speed and if a certain threshold is reached, the slide animation is there, for the last few pixels before the max of 300 is reached
[19:07] <bschaefer> beidl, is there a bug about this?
[19:07] <bschaefer> as thats definitely a bug :)
[19:07] <beidl> bug 1303141
[19:07] <ubot2> Launchpad bug 1303141 in unity (Ubuntu) "[Regression] Unity Gestures: Launcher doesn't slide in" [Undecided,New] https://launchpad.net/bugs/1303141
[19:08] <beidl> shouldn't edge barrier detection only trigger if the mouse pointer is in actually at the edge of the screen? ^^
[19:08] <bschaefer> beidl, was this a recent regression?
[19:08] <bschaefer> beidl, yes
[19:08] <bschaefer> beidl, i don't know much about the gestures tbh :)
[19:08] <beidl> I can't tell, I only installed 14.04 3 days ago
[19:08] <beidl> but on 12.04.4 it worked
[19:08]  * bschaefer doesn't even know who did the gestures work for the launcher
[19:09] <beidl> I could try it with 13.10
[19:09] <bschaefer> that would be nice, as we can possibly pin point the regression if it was recent
[19:09] <bschaefer> if its to far back, then it'll be a bit more tricky
[19:10] <bschaefer> beidl, but the real issue, is no animation?
[19:10] <bschaefer> as if its popping
[19:10] <beidl> yup, no animation
[19:10] <beidl> https://www.youtube.com/watch?v=GY4YiZJ4Lrg
[19:11] <bschaefer> fun .. hmm
[19:11] <beidl> so the effect when dragging the launcher to the left works if the launcher is set to be dragged out
[19:12] <beidl> but dragging in has no effect, only up to a certain threshold and then again only for the last few pixels until the launcher is fully revealed
[19:13] <beidl> depending on how fast I swipe in and when I let go of the touchpad
[19:15] <bschaefer> beidl, so it seems to have a problem going from hiding, to slowly showing it self...
[19:15] <bschaefer> beidl, possibly check what the state hide machine things its state is, when updating
[19:15] <bschaefer> attempting to drag the window out
[19:15] <bschaefer> it could be getting stuck in the hidden state, until it pops
[19:16] <bschaefer> as in the update function... i don't see it tell the state machine anything
[19:16] <bschaefer> and on start, if its hidden it doesn't change the state machine either
[19:16] <beidl> should I just check it by checking all possible quirks?
[19:17] <bschaefer> beidl, do you see the DragOutProgress() function?
[19:17] <beidl> yup
[19:18] <beidl> that one never gets called when dragging in
[19:18] <bschaefer> try printing out the progress
[19:18] <bschaefer> o really?
[19:18] <beidl> it only gets called by RenderArgs
[19:18] <bschaefer> is RenderArgs getting called, or is this cond failing?
[19:18] <bschaefer>   if (options()->hide_mode != LAUNCHER_HIDE_NEVER || hide_machine_.GetQuirk(LauncherHideMachine::LOCK_HIDE))
[19:18] <beidl> RenderArgs doesn't get called either
[19:19] <beidl> so I assume drawing is blocked
[19:19] <bschaefer> possibly its getting called be returning early?
[19:19] <bschaefer> do you have a print statement in DrawContent?
[19:19] <bschaefer> as that should be very very noisy
[19:19]  * bschaefer doesn't see a return in render args
[19:19] <bschaefer> as we do a QueueDraw() on each DragUpdate
[19:19] <beidl> I'll add some here and there :)
[19:19] <bschaefer> :)
[19:20] <bschaefer> beidl, also check: nux::GestureDeliveryRequest Launcher::GestureEvent(const nux::GestureEvent &event)
[19:20] <bschaefer> possibly, we aren't getting an update event?
[19:20] <bschaefer> when dragging in?
[19:21] <beidl> update event is there as it should be
[19:21] <bschaefer> beidl, its getting called as you drag in?
[19:21] <beidl> yup
[19:21] <bschaefer> which should then call QueueDraw()
[19:21] <bschaefer> soo DrawContent should be getting called...
[19:21] <bschaefer> which called RenderArgs()
[19:23]  * bschaefer wishes his touchpad supported 4 touch input
[19:27] <bregma> bschaefer, you know we can capture touch input and replay it
[19:28] <bschaefer> bregma, i think you've mentioned this before, like with the test from libgrib
[19:30]  * bregma tries to restore that knowledge form offline backup
[19:36] <jdstrand> seb128: hey, does this look right to you: http://paste.ubuntu.com/7223201/
[19:37] <seb128> jdstrand, yes?
[19:37] <seb128> jdstrand, what is "loginctl" telling you?
[19:38] <jdstrand> seb128: so, 13.10 and 12.04 didn't have lightdm owning 5 /dev/pts/* files
[19:38] <bregma> bschaefer, sudo apt-get install evemu-tools, then dpwnload this into a file called 'elan.device': http://paste.ubuntu.com/7223202/
[19:38] <jdstrand> seb128: I'm not logged in
[19:38] <bschaefer> bregma, awesome, thanks!
[19:38] <seb128> jdstrand, sorry, the "yes?" was because it might be a known bug
[19:39] <seb128> jdstrand, can you get me the loginctl output?
[19:39] <bregma> bschaefer, then download this into a file called 'elan-4-swipe,events': http://paste.ubuntu.com/7223204/
[19:40] <bregma> bschaefer, then run 'sudo evemu-device elan.device &' and remember the device it prints out (something like /dev/input/event17)
[19:40] <jdstrand> seb128: oh, if I login and back out, it picked up a 6th pts device
[19:40] <jdstrand> seb128: so, just loginctl as a normal user?
[19:40] <seb128> jdstrand, yeah, I bet loginctl lists lightdm sessions?
[19:40] <bregma> bschaefer, then run 'sudo evemu-play /dev/input/event17 < elan-4-swipe.events' to replay the gesture
[19:40] <seb128> jdstrand, yes
[19:40] <bregma> bschaefer, when you're done, kill the evemu-device process
[19:41] <jdstrand> http://paste.ubuntu.com/7223225/
[19:41] <jdstrand> seb128: ^
[19:41] <bschaefer> bregma, sweet, ill give that a go
[19:41] <seb128> jdstrand, loginctl session-status c3?
[19:42] <bschaefer> bregma, didn't print out a device :(
[19:42] <jdstrand> seb128: http://paste.ubuntu.com/7223231/
[19:42] <bschaefer> bregma, nevermind
[19:42]  * bschaefer was in the wrong dir... duh
[19:43]  * bregma slaps bschaefer upside the head
[19:43] <seb128> jdstrand, I guess that greeter is running? what about c1?
[19:43] <jdstrand> the greeter is running. I am logged in via ssh
[19:43] <bschaefer> bregma, fun part, the events happen so fast, the launcher pops open then get stuck
[19:44] <bregma> bschaefer, it's supposed to stay open so you can use it, are you sure it's stuck?
[19:44] <jdstrand> seb128: http://paste.ubuntu.com/7223239/
[19:44] <bschaefer> bregma, well i can drag it back away
[19:44] <bschaefer> since i've pulled it open
[19:44] <seb128> jdstrand, ok, thanks
[19:44] <jdstrand> seb128: note, this is with a fresh install in a VM (I'm performing install audits)
[19:44] <seb128> jdstrand, that looks like the issue that today's lightdm is supposed to fix
[19:44] <bschaefer> bregma, but yeah it works, and it snaps open, just like beidl bug
[19:44] <jdstrand> ah cool
[19:44] <bregma> bschaefer, you mean you can't unshow it?  I'll just capture the close swipe too....
[19:45] <Laney> yeah I see them, they're opened by indicators
[19:45] <bschaefer> bregma, that'll be good, and yeah
[19:45] <bschaefer> bregma, not stuck in a buggy way
[19:45] <bschaefer> just i cant push it close my self
[19:45] <seb128> jdstrand, but I just tried and it looks like it might not be fixed
[19:45] <bschaefer> bregma, have to restart unity
[19:45] <seb128> where is robert_ancell? ;-)
[19:45] <beidl> bregma, I tried it as well, I don't know how fast you did that swipe, but it was pretty instant
[19:45] <jdstrand> should I file a new bug or will you?
[19:45] <bschaefer> beidl, yeah, if we get the close swipe, we can tell :)
[19:46] <bregma> bschaefer, http://paste.ubuntu.com/7223243/
[19:46] <bschaefer> cool thanks!
[19:46] <bregma> bschaefer, beidl, I went at the regular speed, it definitely just pops open
[19:46] <seb128> jdstrand, speaking of robert_ancell
[19:46] <seb128> jdstrand, open a bug please
[19:46] <robert_ancell> er...
[19:47] <seb128> robert_ancell, hye
[19:47] <seb128> hey even
[19:47] <jdstrand> ah
[19:47] <jdstrand> robert_ancell: hi!
[19:47] <bschaefer> bregma, cool, soo I can reproduce it then :).
[19:47] <bschaefer> bregma, smooooth on the close, snap on the open
[19:47] <seb128> robert_ancell, jdstrand was asking about left-open session from lightdm, seems like we still have open session with indicators even after today's update (which looked liked from the changelog it should fix the issue)
[19:48] <jdstrand> robert_ancell: basically, I saw this: http://paste.ubuntu.com/7223201/. then seb128 asked for: http://paste.ubuntu.com/7223225/, http://paste.ubuntu.com/7223231/, http://paste.ubuntu.com/7223239/
[19:48] <jdstrand> robert_ancell: note, this is with a fresh install in a VM (I'm performing install audits)
[19:48] <jdstrand> seb128, robert_ancell: I don't have 1.10.0-0ubuntu1  yet
[19:48] <jdstrand> I still have 1.9.14-0ubuntu2
[19:48] <seb128> jdstrand, robert_ancell: I'm running today's update (including lightdm 1.10) and I've rebooted the machine a bit earlier, and opening/closing a guest session still leads to leftover sessions with indicators
[19:48] <bschaefer> bregma, thats kind of awesome that its pretty easy to output that info and push it back through a device
[19:48] <robert_ancell> jdstrand, it should be fixed in unity-greeter, not lightdm
[19:49] <robert_ancell> we do still have a logind session open because of pulseaudio
[19:49] <seb128> robert_ancell, I've the new unity-greeter and it's not fixed
[19:49] <seb128>                   ├─8094 init --user --startup-event indicator-services-start...
[19:49] <seb128>                   ├─8156 /usr/bin/pulseaudio --start --log-target=syslog
[19:49] <seb128>                   ├─8174 /usr/lib/pulseaudio/pulse/gconf-helper
[19:49] <seb128>                   ├─8207 /usr/lib/i386-linux-gnu/indicator-bluetooth/indicato...
[19:49] <seb128>                   ├─8219 /usr/lib/i386-linux-gnu/indicator-sound/indicator-so..
[19:49] <robert_ancell> seb128, 14.04.09?
[19:49] <seb128> ii  unity-greeter                                         14.04.9-0ubuntu1                                          i386         Unity Greeter
[19:50] <robert_ancell> does ps show those as running?
[19:50] <seb128> ii  lightdm                                               1.10.0-0ubuntu1                                           i386         Display Manager
[19:50] <seb128> lightdm   8219  0.0  0.1 119552  4548 ?        Ssl  21:42   0:00 /usr/lib/i386-linux-gnu/indicator-sound/indicator-sound-service
[19:50] <robert_ancell> seb128, jdstrand, are these triggered on a normal login or a user switch?
[19:51] <seb128> robert_ancell, I did "boot; log in my normal user; go to session indicator, pick guest; log out from guest; log back in to normal user"
[19:51] <robert_ancell> seb128, so does it occur after first login or only after guest login?
[19:51] <seb128> guest login
[19:52] <seb128> I had none after my user login
[19:52] <jdstrand> robert_ancell: let me upgrade. for me, I only logged in via ssh (ie, not via lightdm) and lightdm owned 5 /dev/pts/* devices. I logged in and out and then lightdm picked up a 6th
[19:52] <robert_ancell> ok
[19:52] <seb128> jdstrand, you don't have today's updates though, so part of the issue is probably what got resolved with today's updates
[19:52] <beidl> bschaefer, it might be a wild guess, but is it possible that queued draw calls could be skipped entirely and stay there without being unqueued? maybe it's a nux bug.
[19:52] <seb128> jdstrand, can you upgrade lightdm/unity-greeter and try again?
[19:53] <jdstrand> seb128: yes, that is why I said 'let me upgrade' :)
[19:53]  * jdstrand is doing so now
[19:53] <seb128> jdstrand, sorry, I skipped that, I was typing ;-)
[19:53] <beidl> bschaefer, because I've noticed that the performance of unity degrades on my Intel HD 4000
[19:54] <robert_ancell> seb128, I still get nothing after logging in as guest
[19:54] <seb128> robert_ancell, did you log out/back to your session from the greeter?
[19:55] <robert_ancell> seb128, yes
[19:55] <robert_ancell> seb128, I'll do an upgrade to
[19:55] <robert_ancell> toop
[19:55] <robert_ancell> too
[19:57] <seb128> robert_ancell, let me reboot as well, just making sure I had the updates before my restart (I think I did but better to make sure)
[19:58] <bschaefer> beidl, could be... though they usually aren't skipped...
[19:58] <bschaefer> beidl, nux is another monster as well
[20:00] <jdstrand> robert_ancell, seb128: http://paste.ubuntu.com/7223299/
[20:00] <jdstrand> robert_ancell, seb128: that is without logging in via lightdm
[20:00] <ochosi> robert_ancell: thanks for the hint with the exit-signal, will check that!
[20:00] <robert_ancell> ochosi, np
[20:00] <mdeslaur> seb128: are you able to create a folder on your desktop?
[20:00] <robert_ancell> jdstrand, i.e. that is boot and waiting at greeter?
[20:01] <seb128> mdeslaur, with today's compiz update yes, update and restart ;-)
[20:01] <mdeslaur> seb128: ah, ok, thanks
[20:01] <mdeslaur> sorry for the noise
[20:01] <seb128> mdeslaur, no worry ;-)
[20:01] <jdstrand> robert_ancell: yes
[20:01] <seb128> mdeslaur, that happens only if you have a window on screen btw, you can "show the desktop" as a workaround
[20:02] <robert_ancell> jdstrand, ok, so that all looks normal
[20:02] <beidl> bschaefer, compiz RAM usage definitely gets higher and higher when repeatedly swiping the launcher in and out... I smell leaks
[20:02] <seb128> jdstrand, robert_ancell: ok, I can't reproduce the leftover indicators issue after a reboot, so I guess jdstrand's issue is a different one
[20:02] <bschaefer> beidl, i would guess leaks as well... theres a bug of stuff in nux dealing with gestures, plus what we do with the launcher :(
[20:02] <seb128> robert_ancell, good job fixing that one btw ;-)
[20:03] <robert_ancell> seb128, yeah, I just randomly noticed it. Has it always been broken or did glib change its signal handling behaviour?
[20:03] <robert_ancell> I can't find anywhere in the docs that actually says "if you return false the default signal handler will be run"
[20:04] <seb128> I don't know, it noticed this cycle/pinged ted about it some days ago
[20:04] <seb128> robert_ancell, blame desrt!
[20:04] <robert_ancell> seb128, was there a bug? I couldn't find one
[20:04] <seb128> robert_ancell, no, I asked ted if that was an issue with upstart/indicators being ported to use upstart this cycle
[20:04] <seb128> robert_ancell, he said he would check with you
[20:04] <beidl> bschaefer, what about valgrinding that sucker? probably way too slow?
[20:04] <seb128> I guess you got to it before he managed to do that
[20:05] <jdstrand> robert_ancell: so, I'm not saying it is wrong, but in prior releases lightdm didn't own any of the /dev/pts/* devices while sitting at the greeter. in 14.04 it owns 5. this is intentional?
[20:05] <bschaefer> beidl, it is slow and we have before, but I've not done it wiht gestures
[20:05] <bschaefer> soo i would think it needs some love
[20:06] <bschaefer> beidl, how i usually start valgrind + unity
[20:06] <bschaefer> G_SLICE=always-malloc G_DEBUG=gc-friendly valgrind --tool=memcheck --num-callers=50 --leak-check=full --track-origins=yes compiz --replace 2>&1 | tee ~/unity-valgrind.log
[20:06] <bschaefer> it is very very VERY slow though
[20:07] <jdstrand> robert_ancell: if it is intended, I can add to our install audit notes that this is by design
[20:07] <robert_ancell> jdstrand, I can't think what would be opening them
[20:07] <robert_ancell> fuser doesn't show any of them being owned by any processes
[20:08] <bschaefer> beidl, bregma also getting these fun warnings: http://paste.ubuntu.com/7223335/
[20:08] <bschaefer> beidl, but i seem to be getting to DrawContent/RenderArgs()
[20:09] <robert_ancell> jdstrand, please file a bug and I'll investigate
[20:10] <bschaefer> beidl, the problem im seeing, is the progress when opening is either 0.0, then it jumps to 1.0
[20:10] <bschaefer> theres no animation going on...
[20:10] <jdstrand> robert_ancell: fyi, http://paste.ubuntu.com/7223345/
[20:10] <jdstrand> robert_ancell: I'll file a bug
[20:10] <bschaefer> which.... could be because it doesn't get there *while* dragging...
[20:11] <robert_ancell> jdstrand, ah, so it's probably the indicators/dbus/dconf doing it
[20:11] <robert_ancell> desrt, ^ do you know what these open pts's are from?
[20:13] <jdstrand> fyi, bug #1304640
[20:13] <ubot2> Launchpad bug 1304640 in lightdm (Ubuntu) "lightdm owns several /dev/pts/* devices in 14.04, none in 13.10" [Undecided,New] https://launchpad.net/bugs/1304640
[20:13] <beidl> bschaefer, most probably. anyways I'm done with valgrinding
[20:13] <bschaefer> haha
[20:13] <bschaefer> beidl, yeah valgrind is a hassle
[20:14] <beidl> makes every hardware look dated :D
[20:14] <beidl> bschaefer, http://paste.ubuntu.com/7223364/
[20:15] <bschaefer> beidl, i think this is one thing that is happening, when you manually do it as well:
[20:15] <bschaefer> WARN  2014-04-08 22:11:39 nux.geisadapter GeisAdapter.cpp:157 Failed to dispatch events! Status: 20
[20:15] <bschaefer> beidl, we must be losing events somewhere...
[20:15] <bschaefer> but hmm
[20:15] <bschaefer> beidl, also thats a small log compared to some we use to get :)
[20:16] <beidl> bschaefer, but that's strange, since the OnDragUpdate gets called successfully
[20:16]  * bschaefer remembers having to parse 100k logs
[20:16] <bschaefer> beidl, it is, and im getting progress updates
[20:16] <bschaefer> for both drag in/out
[20:16] <bschaefer> and both don't call draw inbetween the update
[20:17] <bschaefer> beidl, the open: http://paste.ubuntu.com/7223371/
[20:18] <bschaefer> beidl, the close: http://paste.ubuntu.com/7223375/
[20:18] <desrt> robert_ancell: that looks interesting
[20:18] <bschaefer> beidl, but DragOutProgress() only starts getting called after all that... i think theres just a better animation when closing?
[20:20] <bschaefer> well now im a bit confused :)
[20:20] <desrt> robert_ancell: it's pretty clear from the output that all of the processes involved here are simply getting this pts on their stdout/stderr at startup, though
[20:20] <desrt> so i guess if there is a problem then the problem lies what the thing that is spawning them
[20:20] <robert_ancell> upstart?
[20:20] <desrt> i guess?
[20:21] <desrt> look at the fd numbers... 1/2
[20:21] <beidl> bschaefer, I also thought it could be the launcher staying full transparent until letting go, but setting changing the effect in CCSM doesn't help either
[20:21] <beidl> *effect setting
[20:22] <desrt> more interesting is the fact that there seems to be several copies of dconf running as the lightdm user.....
[20:22] <desrt> that's definitely not good
[20:22] <desrt> i guess that also means that you have multiple session busses up
[20:22] <bschaefer> beidl, it could be, ill check as well
[20:26] <robert_ancell> desrt, weird
[20:29] <bregma> bschaefer, that warning at GeisAdapter.cpp:157 is bogus, a value of GEIS_STATUS_CONTINUE is not an error
[20:30] <bschaefer> bregma, i think it could be caused by where i position my mouse
[20:30] <bschaefer> before i run your events
[20:30] <bschaefer> nope, but idk what that STATUS_CONTINUE means
[20:33] <bregma> bschaefer, GEIS_STATUS_CONTINUE means there are event remaining in the queue, they'll be processed the next time the main even loop gets executed
[20:33] <bregma> which will be right away, because there are still events in the queue
[20:34] <bschaefer> beidl, weirdly, (1.0f - DragOutProgress()) returns 0 when dragging open, and 1 when dragging close
[20:34] <bschaefer> beidl, i don't think DragOutProgress() does anything, and I think its all about: auto_hide_animation_.GetCurrentValue()
[20:34] <bschaefer> bregma, haha
[20:34] <bschaefer> bregma, hmm i wonder why they print errors then?
[20:34] <bschaefer> why *nux* does at lease
[20:35] <bregma> bschaefer, that sounds like some of the old Nux animation code that was changed back in 13.10 to use new animation
[20:35] <bschaefer> bregma, yeah, which i think is causing the bug
[20:35] <bschaefer> as its returning 0
 * 0 :)
[20:35] <bregma> I suspect it may have been ported incorrectly
[20:35] <bschaefer> yeah which is expected if you ever look at that launcher code haha
[20:35] <bschaefer> beidl, try this! (Seems to work for me)
[20:36] <bschaefer> beidl,     float autohide_progress = auto_hide_animation_.GetCurrentValue();// * (1.0f - DragOutProgress());
[20:36] <bschaefer> just use the auto_hide_animation_GetCurrentValue() under RenderArgs();
[20:36] <bschaefer> and its all smooth for me :)
[20:36] <bschaefer> that looks soooo much better...
[20:37] <beidl> building
[20:37] <bschaefer> beidl, cool, yeah we switched to a new animation frame work,and we updated a bunch of the animations
[20:37] <bschaefer> seems like we missed this one :)
[20:37] <bschaefer> Trevinho, would have known right away though haha
[20:37] <bschaefer> as he did the launcher animation re-work :)
[20:39] <beidl> bschaefer, the good news: yes it animates! :) the bad news: it doesn't follow finger movement like in the video
[20:39] <bschaefer> :(
[20:39] <bschaefer> beidl, so now it just animates?
[20:40] <bschaefer> only when it gets above 300?
[20:40] <beidl> when dragging in, yes. when dragging out it's almost instant
[20:41] <beidl> when dragging out by a few imaginary pixels, it hides, then letting go makes it slide in again
[20:41] <bschaefer> welllll shoots
[20:43] <bschaefer> beidl, in == hide? out == open?
[20:43] <beidl> sry, in == reveal, out == hide
[20:44] <bschaefer> beidl, soo reveal is still broken? As it doesn't follow the fingers?
[20:44]  * bschaefer is surprised dragging to hide works...
[20:44] <beidl> yes, it only animates when above 300 and letting go
[20:45] <bschaefer> beidl, yeah you're right, theres no drawing going on until after you let go (i think?)
[20:45] <bschaefer> let me double check that
[20:45] <bregma> beidl, there's really no way to guarantee the Launcher drag will follow the finger movements smoothly because of the relative sizes and pixel and point densities of the various input and output devices involved
[20:46] <beidl> bregma, well the delta_drag_x changes respectively to the finger movement
[20:47] <beidl> and on 12.04 it worked quite nicely
[20:48] <bschaefer> beidl, i don't think we are getting any gesture events until it gets past the 300...?
[20:49] <bschaefer> as when I use that event bregma gave me it either works, animates and gets above 300
[20:49] <bschaefer> or does nothing
[20:49] <bschaefer> and complains about failing to dispatch an event
[20:50] <bschaefer> bregma, if you've some time to send me an event that drags the launcher attempting to open it, but doesn't quite get there?
[20:50] <bschaefer> as it use to drag out a little, then you let go, its not within the threshold, and hides again
[20:50] <bschaefer> IIRC
[20:53] <beidl> std:: cout << drag_out_delta_x_ << std::endl; right before QueueDraw() in OnDragUpdate() gives me this when dragging in-out-in-out
[20:53] <beidl> http://paste.ubuntu.com/7223506/
[20:54] <bregma> bschaefer, I can't get it to do that
[20:55] <bschaefer> bregma, o,well shoot
[20:55] <bschaefer> beidl, hmm are you able to draging towards revel, then back down?
[20:55] <bschaefer> does it update correctly like that?
[20:56]  * bschaefer needs a touch pad
[20:56] <bschaefer> Trevinho, do you have one?
[20:56] <beidl> bschaefer, yes
[20:56] <bschaefer> beidl, and its not getting to DrawUpdate/RenderArg when dragging right?
[20:57] <beidl> bschaefer, right
[20:57] <bschaefer> beidl, which is the problem...but it doesn't seem to get to draw when dragging close either?
[20:57] <beidl> even though QueueDraw returns
[20:57] <bschaefer> beidl, does it render when you're closing it? (the test events i have show it only renders until the dragging is done, but down during updates)
[20:59] <beidl> you mean with your autohide_progress change?
[20:59] <bschaefer> beidl, well, in general
[21:00] <beidl> bschaefer, generally, when closing, yes, it renders
[21:00] <bschaefer> beidl, when i was using the close event, it wasn't getting to DrawUpdate() when the delta x was updating
[21:00] <bschaefer> hmm
[21:00] <bschaefer> it all happens so fast on my end haha
[21:02] <beidl> bschaefer, I guess a real-time-record would be way too big for a paste :D
[21:02] <bschaefer> well, i think a real device would make this a bit easier :)
[21:03]  * bschaefer keeps trying to touch his screen
[21:03]  * beidl closes his laptop
[21:03] <beidl> can you reach it? haha
[21:03] <bschaefer> no haha
[21:04] <bschaefer> well, at this point we can make the animation no instant
[21:04] <bschaefer> not*