=== mhall119 is now known as mhall119__ === mhall119_ is now known as mhall119 [00:48] thumper: you still around? [00:48] hi mhall119 [00:48] I am... kinda [00:48] heh, me too [00:49] can we chat on IRC for a bit? [00:49] sure === dyams|away is now known as dyams [08:16] Saviq: you there? [08:56] JohnLea: question, unity-2d and unity-3d, behave different if you press the windows key and then alt+f1 (unity-2d closes the dash while unity-3d no) can you confirm that unity-3d behaviour is the wanted one? If not any idea who to ask? [08:57] I think there is a bug for that [08:57] I think alt+f1 should close the dash [08:57] * thumper waits for JohnLea :) [08:57] thumper; have you got the bug #? I updated the bug for this a week or so ago [08:58] JohnLea: not in front of me [08:58] thumper; me neither, but there is one ;-) if you hold on a sec I'll dig it out === dyams is now known as dyams|lunch === JohnLea_ is now known as 17SAAGUZ7 === dyams|lunch is now known as dyams [09:24] thumper: https://bugs.launchpad.net/unity-2d/+bug/919209 right? [09:32] greyback: so want me to do the MR to unify the WIDTH declaration somewhere? [09:33] tsdgeos: sure. What do you think is good, a file in tests/misc? [09:33] definitions.rb ? [09:34] sound good to me [09:35] * tsdgeos goes for it [09:36] greyback: add all the definitions or only the ones we have at the moment? [09:39] tsdgeos: hmm, what can we be certain of. launcher width, pane height. Dash dimensions should be correct. [10:10] saviq: ping [10:16] tsdgeos: do you how to verify that the application menu, say Menu->File is displayed or not? [10:17] dyams: haven't tried it, but i'd hope the menu to be introspectable too? [10:19] tsdgeos: unfortunately not. panel-service draws the menus, we've no access to that via testability [10:20] well, it draws them [10:20] but the QActions are still on our side [10:20] no? [10:21] yeah, we reveice them & tell the panel-service to open the menu [10:22] otoh maybe we can actually query the panel-service ? [10:23] greyback: https://code.launchpad.net/~aacid/unity-2d/unity-2d_test_common_definitions/+merge/90251 i only added LAUNCHER_WIDTH and PANEL_HEIGHT since it's the only ones we use, could not find the dash values either [10:23] tsdgeos: that's fine [10:24] tsdgeos: yes I'd like to be able to ask the panel-service if a menu is open, what is highlighted, etc. I'm unsure if it exports such info, but if it does, a little code in the panel to give testability access to that info would be nice [10:24] tsdgeos: it's the biggest weakness in the testability stuff IMO [10:28] dyams, hey, what's up? [10:28] sorry guys, in a cafe again, no power at home >:[ [10:28] doh :-/ [10:29] saviq: about multi monitor shell stuff [10:30] Saviq: you need to stop relying on solar power ;) [10:34] saviq: nerochiaro: you were discussing about the QDeclarativeItem re-parenting crash issue [10:44] greyback, smarty pants :P [10:44] Saviq: how did you know my high-school nickname? :) [10:45] greyback, guess what I was called :P [10:45] Saviq: I doubt I'd be able to pronounce it [10:47] dyams: it looks like we managed to find a way to work around the crash. we're checking with Kaleo the performance [10:48] nerochiaro: ok [10:49] nerochiaro: is it by bypassing the QDeclarativeView and using the QGraphicsView directly? [10:49] dyams: yes [10:50] nerochiaro: ah [10:52] nerochiaro, oh, so sounds like we might use a single QGraphicsScene in the end? [10:54] Saviq: it looked like that would require much more work, so the most promising direction for now is one view per screen and only one with a dash, then move the dash around between them as needed [10:54] nerochiaro, ok, cool [10:55] greyback, I have to miss the standup again today, I'm in no environment to do it now, can you take over? [10:56] my input would be: I'm reviewing, testing, ensuring test coverage, still not doing anything useful ;P [10:57] Saviq: No prob [10:57] and being an a$$ to people around [10:57] greyback, thanks [10:57] any tips? [10:57] greyback, one thing I talked about with Kaleo the other day [10:58] what we discussed with Ugo earlier [10:58] i.e. ensuring tests do test [10:58] a simple `bzr revert ` might be enough to do that in most cases [10:59] Saviq: *nod* [10:59] saviq: is that for me you are saying ^^ [10:59] dyams, no, greyback [11:00] saviq: ok [11:00] greyback, so it might not be worth it to devise no evil plans === chaoticuk_ is now known as chaoticuk [12:39] nerochiaro: there? [12:40] tsdgeos: yes [12:40] nerochiaro: all my tasks are in review stage, so i'm idling until they get reviewed and potential issues are found, want me to take "Fix spread behavior" from you in the kanban? [12:41] tsdgeos: please do, i'm still tied up with multimonitor tests and a few other tasks in review i need to fix [12:41] oki [12:42] thanks [12:42] sure [12:42] better than do nothing and be bored ;-) [12:43] tsdgeos, you know we still have some other tasks to take there in the backlog? ;) [12:43] tsdgeos, but yeah, you can grab them from Ugo until you run out of them [12:43] Saviq: i do, but though that the ones that were "inside" the board had more priority than the ones outside [12:43] Saviq: it's fine, better he takes stuff off my plate, since i'm won't be able to work on them after this week [12:44] nerochiaro, yeah I know [12:44] just making sure [12:44] nerochiaro, me too, just making sure tsdgeos knows we have more :) [12:44] :) [12:49] tsdgeos: I've something else for you to do if you're bored :) [12:49] greyback: sorry, too late :D [12:56] tsdgeos: Still like me to dig into what's breaking in unity-2d_test_dash_launcher_interactions? Or was that the test you fixed? [12:56] tsdgeos: and why do you object to using the TIMEOUT variable. It was partly your idea :) [12:57] s/variable/variable almost everywhere/ [13:02] is there a supported way to change the layout of virtual desktops? (e.g. move from the default 2x2 to 1x6) === jackyalcine is now known as JackyAlcine [13:11] jml, ccsm [13:12] seb128: I was under the impression using ccsm was strongly discouraged. === greyback is now known as greyback|lunch [13:13] jml, it is, I can reply "don't do that" to your question if you prefer ;-) [13:13] jml, it's ccsm or don't do it basically [13:13] seb128: heh, ok :) [13:14] seb128: do you know if it's something that we would like to allow users to configure? [13:14] but just haven't got around to it yet because ccsm is good enough for the daring few [13:15] jml, dunno, it's not in the options design looked at for the system settings at least [13:15] jml, didrocks did an upload today to precise which adds things like the launcher autohiden and icon sizes to the system settings ui [13:16] seb128: that's pretty cool :) [13:16] jml, but the number of workspace is not part of the things which went into that spec [13:16] jml, see https://docs.google.com/a/canonical.com/document/d/1ILTJDiDCd25Npt2AmgzF8aOnZZECxTfM0hvsbWT2BxA/edit?ndplr=1 section 2.2 [13:19] seb128: that's a good spec, thanks. [13:20] JohnLea: any plans to allow users to configure the number of workspaces or their layout? [13:24] JohnLea: I'm more interested in whether you think it's acceptable in principle than if there's a schedule for including such an option. [13:48] huh [13:49] and the system fonts are *not* configurable and are not Ubuntu Mono for the monospace font? [13:51] greyback|lunch: because if i use the TIMEOUT in the verify_not, the tests are looooooooooooooong === greyback|lunch is now known as greyback [13:57] tsdgeos: true :) Ok I'm happy with 2 seconds then [13:57] :-) [14:05] tsdgeos: You'll need to update ~aacid/unity-2d/unity-2d_test_dash_launcher_interactions to reflect the definitions.rb stuff [14:06] greyback: yes [14:06] dyams: ping [14:17] greyback: updated [14:17] tsdgeos: thanks [14:18] greyback: do you know who created the task that syas "Open Dash and Invoke Spread. Toggle spread with Super+S. After 3-4 attempts Dash will stay open and stop responding to Super Or Escape etc." [14:18] ? [14:18] tsdgeos: nope [14:19] i thnk it was dyams, wonder if it's late in the day for him already [14:19] tsdgeos: he's 4 or 5 hours ahead of us [14:19] yeah it is [14:19] and it's a national holiday for him too [14:19] doh :-/ [14:19] tsdgeos: looking for something to do? [14:20] :) [14:20] i'm not sure i understand what is the sequence he makes to find the problems [14:20] greyback: i do now since i don't know what to make of that task [14:20] tsdgeos: nor am I. Did he link a bug? [14:20] no, because it's -shell only [14:21] Invoking spread makes dash hide, I don't see how dash will re-appear [14:21] yeah [14:21] well, let's ignore it until tomorrow [14:21] greyback: what did you want me to do? (i have a few test more in mind though) [14:43] greyback: damnit, i repushed a merge to https://code.launchpad.net/~aacid/unity-2d/unity-2d_test_super_s_does_not_type_s/+merge/90068 you'll have to reapprove it :-/ [14:43] tsdgeos: ok [16:26] Hi all. [16:26] The global app menu bar was messed up in a new update to precise [16:26] where can i report it? [16:27] elvisd: you using unity2d by any chance? [16:28] no unity, [16:28] the menu bar appears and works until i switch application [16:30] example: 1. open nautilus --> menu works; create a file, change folder, ... --> menu works. 2. Focus Xchat window 3. Re-turn to nautilus --> Menu doesn't work [16:31] elvisd: hmm, sounds like dbus-menu problem [16:33] ok. wher can i report it. only here to you, mailing or launchpad? [16:33] elvisd: report it on launchpad please [16:34] elvisd: http://bugs.launchpad.net/unity/ [16:34] Thank you for the link ;) [16:34] I'm not 100% certain where the error lies, so report it there, and the bug triager will figure it out [16:37] elvisd, that should be fixed now [16:37] Saviq: regarding https://code.launchpad.net/~aacid/unity-2d/unity-2d-shell_fix_dash_launcher_interactions/+merge/90156 i think we can merge it and after try to find out while the whole test fails, since i don't see anything wrong there [16:37] that was a bug introduced in the libindicator upload yesterday [16:37] tsdgeos, ok then [16:37] Ah ok.When should be available the fix? [16:38] should already be available [16:38] kenvandine, I stop filing the bug then. [16:38] maybe it hasn't hit your mirror... what versions of libindicator3-7 and indicator-appmenu do you have? [16:39] kenvandine, how to check please? [16:39] greyback: unity 2d legacytray filters out skype. I think it should be white listed [16:39] apt-cache policy indicator-appmenu [16:39] kenvandine, Installed: 0.3.2+hud5-0ubuntu1~ppa2 [16:40] kenvandine, apt-get dist-upgrade doesn't propose any update [16:40] ah, i have hud installed... [16:40] elvisd, oh! [16:40] that is why [16:40] hamax: please log a bug about it. I whitelisted it myself manually [16:40] i'll upload the fix to the hud ppa too :) [16:41] kenvandine, ah ok :) [16:42] so i can continue testing hud [16:49] tedg: FF history isn't opening from the hud, ie hud displays history>url clicking on it doesn't open the url in FF [16:52] davmor2, bug 921231 [16:53] tedg: cool as long as it's known :) [16:59] davidcalle: ping [17:00] mhall119, unknown host. [17:00] mhall119, what's up :) [17:00] I know where you're hosted [17:01] have you had a chance to look over the lens guidelines ? [17:02] No time for it, but I will tonight. [17:02] ok, thanks, I'd like to get that posted somewhere soon, as I've seen a couple of people asking questions that it could answer [17:02] mhall119, ok [17:03] kamstrup: ping [17:13] seb128, who do we bug about getting ppa builds re-scored? [17:14] kenvandine, usually ask #ubuntu-devel, at this time infinity or stgraber can be good bets [17:15] kenvandine, hey, did you ever notice that clicking the close button on gwibber takes ages to close the window? [17:15] kenvandine, http://paste.ubuntu.com/817868/ [17:15] mhr3, no.. [17:16] kenvandine, then i'm either very demanding, or you have very fast computer :) [17:16] anyway, ^^ makes it instant(tm) [17:16] sure [17:17] i am not sure what it can be doing though [17:17] kenvandine, it's vala, it's unreffing all the other objects [17:17] mhr3, did you see my response about the lens filters? [17:17] and destroying them just takes a while [17:17] thanks for the patch :) [17:17] kenvandine, oh yea, it'll be some hickup in unity, i need to digg into it [17:18] the good news is if i get it from the lens the changed signal is emitted :) [17:18] hi all.. tried the HUD ppa this am -- love the ideas behind HUD... but ... [17:18] kenvandine: when will gwibber get a multi-feed view again? [17:18] I've had to switch to the tweetdeck chromium app [17:19] mhall119, i dunno... i really wanted to do that for precise [17:19] ok, so maybe *had* is a bit strong [17:19] but haven't had time to hack on gwibber [17:19] I'd help, but you switched it to vala :( [17:19] i found dbusservice was taking 70% of the cpu [17:19] kenvandine, when i was testing it was mostly working (changes visible in unity), but if i restarted the lens without restarting unity, it no longer cared about changes in the filter model [17:20] is that a known issue? [17:20] mhall119, i think when we get more renderers available in unity the lens might be a better fit [17:20] tedg, see swfiua question [17:21] i think tedg might have just uploaded a fix for that to the PPA [17:21] oh excellent.. [17:21] tedg, it's building now btw [17:21] dbusservice ? [17:21] No, I dont' think anyone got the package with the infinite loop. [17:21] kenvandine, oh and btw any ideas why i dont see the gwibber lens if i run latest unity? [17:22] swfiua: i saw something similar... where it was really crunching and results took several seconds to update [17:22] btw any chance of an option to configure the hot-key that fires up HUD? eg the menu key would make sense for me --- ALT is used a lot by emacs [17:22] tedg: sorry, think it was dbus-daemon [17:22] wondering if it was a first run issue.. [17:23] mhr3, define latest? [17:23] kenvandine, trunk [17:23] i see it in the hud build and the precise version [17:23] swfiua, Hmm, shouldn't be really... I mean it gets the menu cache at first, but that shouldn't peg the daemon. [17:23] swfiua, But it would make HUD slow [17:23] kenvandine, yea, precise version works fine [17:23] damn! [17:23] tedg: oh it made it slow alright ;) [17:23] swfiua, Key is configurable in CCSM [17:23] i wonder if we screwed something [17:24] mhr3, so i would say unity is broken :) [17:24] of course, might be something else i'm running that is causing grief [17:24] tedg: tks for the ccsm tip [17:24] ccsm needs to die [17:24] tedg, alt doesn't really work though because it shows the menus [17:24] kenvandine: not before the useful options in there are available elsewhere [17:24] not sure why that works to raise the hud for anyone [17:24] kenvandine, oh btw did you latest dee addition for text processing? it's magic [17:24] swfiua, i know... that is why it isn't dead [17:25] mhr3, no... [17:25] will it speed up my text searches? [17:25] probably slow them down... [17:25] but magic :) [17:25] i hate having to iterate over the model and do string comparisons [17:25] http://bazaar.launchpad.net/~kamstrup/dee/icu-transliterators/view/head:/tests/test-icu.c#L45 [17:26] kenvandine, why do you have to do that? [17:26] is there a better way? :) [17:26] oh... i could add a column filter [17:26] that doesn't help [17:26] and what are you trying to do? [17:27] when you search in the lens, i do substring matches on a few columns [17:27] to add the results to the results model [17:28] kenvandine, oh, just use one of the Index classes [17:28] which when i first did that, i was told the only way was to iterate over the model [17:28] TreeIndex probably [17:28] cool [17:28] if you want it fast [17:28] would be nice to speed it up :) [17:28] it should work in O(log n) so... ;) [17:28] any of the other lenses using that? [17:29] dont think so, it was added fairly recently [17:32] mhr3, the icu stuff is cool [17:32] Dee is just awesome :) [17:32] kenvandine, congratulate kamstrup on that ;) [17:32] * kenvandine high fives kamstrup [17:32] kamstrup, now go fix ModelIter in GI [17:32] :) [17:32] it's keeping me from adding more gwibber tests :) === agateau_ is now known as agateau [17:34] kenvandine, it works fine in vala :) [17:34] yeah, but i have a rule to keep the vala and python tests == [17:34] and rick wants to use it too [17:35] yea... i'll need to look at it at some point [17:35] mhall119, what do you think about the unity lens filling for folks that want multi-column [17:35] mhr3, bug 917761 [17:37] now i to fix it then [17:37] i need* [17:37] woot [17:37] :) [17:37] mhr3, it has basically never worked [17:38] and it makes models pretty much useless in python... [17:38] sure you can iterate over them and stuff [17:38] give me a sec :P [17:38] haha [17:43] mhr3, merged your patch, thanks! [17:51] kenvandine: I don't think that'd work out [17:52] browsing a chronological list of messages in the dash would be awkward at best [17:52] i want vertical columns in unity [17:52] and activating a message would show more info on it on the right side [17:53] similar to the Ubuntu TV demo shows movie info [17:53] * kenvandine has no buy in on this from design though :) [17:53] kenvandine, you should talk to john about that :) [17:53] i think it would be cool [17:54] i'll get sent to the naughty step if i just make more custom renderers without their input ;) [17:54] gord, has there been any discussions on columns? [17:54] and the renderer would need to be done twice, for both unity and unity-2d [17:55] kenvandine, i've talked to them about it a few times, but not seen anything come out of it - seriously, just ask John lee [17:55] i will [17:55] he knows i want to focus more on richer unity integration [17:55] i can say it would be nice to have this and that, but what we need is lens people saying "we need this and that to get good results from this lens" [17:55] mhall119, as a user that likes multi-column, do you think something like that would fit your need? [17:56] personally i would like to spend less time focusing on the gwibber client and more time working on making the unity integration awesome :) [17:58] gord: has anyone from the TV team talked to you about making a renderer for their TV guide? [17:59] mhall119, that is kind of a problem... the TV team is using unity-2d which doesn't share renderers with unity [17:59] afaik at least [17:59] kenvandine: well right now the EPG isn't a lense at all [17:59] partly because they didn't have something to do the layout right [18:00] and also because they need more metadata passed to the guide than just an icon and URI [18:00] right [18:00] when is another question, is there a way to pass arbitrary metadata between a scope and a lense other than URI encoding? [18:00] gord, I could use a coverflow renderer. [18:01] gord, renderers between 2d and 3d don't share any code right? [18:01] the video lense looked to have something like coverflow [18:01] mhall119, nope [18:01] mhall119, there still is the second text field. [18:01] kenvandine, nope [18:01] davidcalle, would be nice, not for this cycle [18:02] * gord looks at feature freeze date [18:02] gord: can I send one of the TV guys your way to see if an EPG renderer is possible? [18:03] mhall119, i'm in contact with them enough :) its possible, we control our stack, we can do anything really [18:04] ok, by possible I really meant "Not so much effort that it wouldn't be worth it" [18:06] kenvandine, so, how does this look http://paste.ubuntu.com/817933/ ? [18:07] mhr3, no way... are you trying to show up kamstrup? [18:07] :) [18:08] kenvandine, yea, he did too much cool stuff for today, i need to catch up [18:09] haha [18:09] mhr3, awesome... thanks! [18:11] mhr3, test it in javascript too :) [18:11] or push a branch somewhere and i can [18:12] yea, i'll prepare a mp [18:12] cool [18:24] kenvandine, https://code.launchpad.net/~mhr3/dee/gi-modeliter/+merge/90306 [18:24] cool [18:24] thx [18:32] mhr3, kenvandine: as I noted on the Dee.ModelIter in GI bug, I am afraid it'll require a huge chunk of work in pygi [18:32] kamstrup, ahem ^^^ [18:33] mhr3: ? [18:33] https://code.launchpad.net/~mhr3/dee/gi-modeliter/+merge/90306 [18:33] that's the huge chunk of work :P [18:33] wtf now way :-) [18:34] you know I have to reject the branch now don't you [18:34] to preserve my pride [18:34] ok, i'll resubmit it tomorrow [18:34] haha [18:34] kamstrup, and it works :) [18:39] mhr3, kamstrup: i can use the iter in javascript too :) [18:39] woot [18:39] i love easy bugs :P [18:39] i haven't figured out how to connect to signals in javascript yet [18:39] mhr3, now you are just trying to hurt kamstrup [18:39] :) [18:40] kenvandine, oh come on, you can't say that out loud :P [18:40] mhr3, kenvandine: sorry, to save face I had to mark it needs-fixing [18:40] ;-) [18:40] hehe [18:42] kamstrup, damn you :) [18:43] kamstrup: ping [18:43] mhall119: hey there [18:43] kamstrup: hey, you linked me to a blog about changes tothe lens api a while back, but I lost the url, mind giving it again? [18:43] mhall119: grillbar.org [18:44] kamstrup: thanks [18:45] and real quick question, can a lens be written in python or vala that will work in both Unity 4 and 5? [18:47] mhall119: in theory yes, but it'll require a bit of fiddling [18:50] mhr3: I added the benchmarks. There is no significant change [18:50] ok, so it's not that one API is a superset of the other [18:51] kamstrup, i knew that, that's why i didn't add them :) [18:52] kamstrup, i'm not sure what you mean by the second point though [18:53] mhr3: call copy() and free() on a Dee.ModelIter and verify that copy returns a pointer == the orig, and maybe that free is a no-op... call it N times and check you don't crash ;-) [18:54] kamstrup, but those are hidden [18:54] mhr3: g_assert (g_boxed_copy(DEE_TYPE_MODEL_ITER, iter) == iter); [18:55] i never knew if you're supposed to pass the raw struct there, or some gtypeinstance or something [18:55] kenvandine, So how much do you love me? [18:55] int i; for (i= 0; i < 100; i++, g_boxed_free(DEE_MODEL_ITER, iter); [18:56] kenvandine, I want to apply this, but it'll make everything using dbusmenu-gtk FTBFS.... how should we version that? https://code.launchpad.net/~ballogy/dbusmenu/diff-dir-for-gtk3/+merge/82839 [18:56] mhr3: just the pointer, that's why you need to pass in a gtype as first arg [18:56] makes sense [18:57] mhr3: got you! you just said "makes sense" to my for-loop that uses an embedded ,-operator to do the assertion! ;-) [18:57] wow I am a geek [18:58] or maybe mhr3 has really read enough vala-generated C code to start liking te ,-operator... like /me [18:58] kamstrup, oh you're sooo evil, i'm gonna tell njpatel [18:58] noooooo [18:59] he'll spank me good [18:59] kamstrup, like your checkDirty... eeek [18:59] mhr3: exactly! did you see the attention it gathered on g+? :-) [18:59] kamstrup, but i liked it, cause vala generated code like that in the early versions [19:00] kamstrup, so congrats you work like a codegen now ;) [19:00] mhr3: what? it doesn't anymore? [19:00] tedg, UGH! [19:00] it was my favorite thing! [19:00] kamstrup, right?! but no, they dropped it [19:00] now I hate vala [19:01] kamstrup, just file a bug, "i want commas in the generated code" [19:01] totally gonna do [19:01] otherwise i'll rewrite everything in mono [19:01] kenvandine, See this is good to do *after* the libindicator change because things don't need to build right away. See how generous I am? [19:01] tedg, that'll make all the indicators FTBFS then [19:04] kenvandine, It'll break: $ apt-cache rdepends libdbusmenu-gtk3-4 [19:04] kenvandine, Which is basically the indicators and appmenu-gtk [19:04] yeah [19:04] who is going to fix them all ? [19:04] :) [19:04] might as well fix all the deprecations too [19:04] kenvandine, I will [19:05] kenvandine, No, I'm still waiting for distro to back out that version of GTK [19:05] :-) [19:05] kenvandine, Do you think we need to change the API version, or can we just do this with a release? [19:05] * kenvandine ignores ted [19:05] * tedg doesn't want to change the pkgconfig files [19:05] it is just the indicators... [19:06] but... what about folks building it outside of ubuntu? [19:06]