[00:00] yeah, too bad I get some other random issues with unity after restarting it... fun times [00:00] mhr3_, thanks you! Yes it does, but getting it to display the pretty pixels is step one :) [00:00] I guess unity is just telling me not to play games, and do work [00:01] (when I restart unity, from time to time, gnome-terminal gets into a weird state where I can't alt-tab to it - i can still click on it to get focus though) [00:02] kklimonda, :(, usually restarting the terminal fixes that [00:02] though i thought we fixed that issue... [00:02] has you fixed it in the last few days? [00:02] have* [00:03] kklimonda, nope that should have been like a year or so ago...so thats a different bug... [00:03] ah, I'll report it then when I get it to happen again [00:03] kklimonda, thank you! [00:06] kklimonda, also the fact that your settings don't change dynamically through ccsm is a bug (as it should be dynamic...) [00:06] bschaefer: I know, I'm really confused on why it's happening [00:07] yeah.. if you go to preferences and see what backend you're using? [00:07] as it should be set to gsettings [00:07] and the intergeneration box should be checked [00:07] (in ccsm) [00:08] where in ccsm? [00:08] kklimonda, bottom left [00:08] ah [00:08] above close, and advanced serach [00:08] search [00:08] yeah its in a strange spot [00:08] yeah, it's GSettings, and the integration box is checked [00:08] sadface [00:09] then im not sure what its not dynamic as i've never ran into that issue... [00:09] kklimonda, what version are you running of unity? [00:09] * bschaefer should have asked this first really... [00:10] bschaefer: the one that comes with 14.04 [00:10] o so you've got the most update to date version hmm [00:10] (thats around for download) [00:10] 7.1.2+14.04.20131106.1-0ubuntu2, I just restarted it [00:11] thats the version i have installed...let me try running that version as im on trunk untiy [00:12] kklimonda, did you upgrade to 14.04 or clean install? [00:12] it's a clean install [00:12] as it still seems to work here hmm [00:14] kklimonda, ccsm is not a nice application at all [00:14] Trevinho, if you're still around, have you ever seen the problem where CCSM doesn't actually save the changes? [00:15] it isn't, but then i have used it a few times in the past, and never had this particular issue :) [00:15] yeah, its having problems saving the changes it seems... [00:16] kklimonda, could you try install 'dconf-editor' and see if it'll change it this way [00:18] once you've dconf-editor, press Ctrl+f (for find), then search for 'show-hud' [00:18] from there you should see the '' key, if you just change it to '', it should stop the hud from opening right away [00:18] yeah, I just run apt-get dist-upgrade and unity exploded.. [00:18] :( [00:18] sigh, that will teach me ;) [00:19] can't load plugin 'ccp' [00:19] thats not a good error...umm, this is after a dist-upgrade and reboot? [00:19] compiz is having a problem starting (not unity this time!) [00:19] no, during dist-upgrade unity just segfaultet [00:20] segfaulted* [00:20] well, compiz [00:20] and then when I try to run in again, it hangs at this error [00:20] dconf-editor shows that the show-hud command is changed, and it works after I restart unity [00:20] but not right away? [00:20] ie. not dynamic. [00:21] yeah [00:21] that is very very strange... [00:21] I can make a clean user and see if it still happens [00:21] that isn't a bad idea [00:21] once I get my system into a usable state again :) [00:21] yeah :), umm getting back from that depends on the actual seg fault reason [00:22] but hopefully continuing with the upgrade [00:22] then a reboot...should fix it [00:30] ccp is not loading because the new libprotobuf changed ABI.. [00:30] (I had to downgrade some stuff) [00:34] it seems to be working under a new account, so I'll just assume something weird happened on my main [00:34] which is weird, this installation is like a week old.. [00:37] strange... [00:38] sounds like its looking in the wrong place... [00:38] well at lease now its working (though thats never fun to have to make a new account just because of that...) [00:38] kklimonda, if you get some time to file a bug against compiz? [00:40] yeah, I wonder if it's related to that other bug I reported, where ccsm is not saving changes I make (I tried to disable "Initiate Window Move" but it's always being reverted to the default value if I disable it, and when I set a different combination it's just ignored) [00:40] bug 1273690 [00:40] bug 1273690 in unity (Ubuntu) "can't disable Alt+Button1 shortcut for moving windows around" [Undecided,New] https://launchpad.net/bugs/1273690 [00:41] kklimonda, sounds like it, it seems to be running of different files. ie, its reading one file and saving to a different file [00:42] then when you relog it looks at the saved one [00:43] mhm, sounds plausible [00:45] as all the settings are stored in a single compiled file hmm if you would be able to umm [00:45] kklimonda, get a strace off of the ccsm that fails to load dynamically [00:46] that way I can see if its reading/writing to the wrong files [00:46] kklimonda, or just upload that log to the bug [00:46] ok, I'll do it tomorrow when I get up :) [00:46] (it'll be a large file) [00:46] thanks for the help [00:46] yeah [00:46] kklimonda, thanks a lot! [00:46] and np! [08:23] tsdgeos, 'fraid I overridden your review a bit on TextSummary [08:28] Saviq: seen it, no worries, good stuff you found :) [08:30] tsdgeos, did you ever notice weirdness with childrenRect when it's supposed to "shrink"? I had to do lines 11-12 of https://code.launchpad.net/~unity-team/unity8/unity8-card-overlay/+merge/204790 in Card.qml for it to behave properly [08:30] tsdgeos, otherwise it'd grow and stay grown, even if the contents shrunk [08:30] Saviq: i'm going to approve ted's the table is anywhere MR https://code.launchpad.net/~ted/unity8/on-the-screen/+merge/203851 [08:31] tsdgeos, ok ;) [08:31] actually no, needs a pot rebuild [08:32] tsdgeos, the other possibility is there's a binding loop and it just doesn't propagate up (which, btw, Qt complains about), but I found no way to break that loop [08:33] no, never noticed it [08:33] if we have a binding loop it may be that [08:33] and maybe the way to break it is how you did [08:39] tsdgeos, only I hate the solution ;) [08:40] but well :| [08:40] /biab [08:51] Saviq: ok, pushed a branch of the image_preview thing that now is able to merge corrrectly [09:05] mhr3_: Saviq: the actions definition is different in "preview definition example" and in "Preview JSONs" on the document [09:05] which is right? the second one' [09:05] ? [09:16] tsdgeos, yea, second one === mzanetti- is now known as mzanetti_ [09:17] tsdgeos, i mean... both are correct :) [09:20] mhr3_: ? [09:21] ah you changed it [09:30] come on! [09:30] now other tests fail! [09:30] * tsdgeos jumps out of the window === lool- is now known as lool [09:38] uh oh, developer down! [09:44] Saviq, hey man [09:44] Saviq, you asked me to change implicitHeight with height [09:44] but it's implicitHeight also the minimum allowed height? [09:46] sil2100, hey, can we land libunity without the train? [09:46] sil2100, as in the old lib [09:54] mhr3_: Saviq: where do i get the colour of the button for the action? [09:54] i've seen some stuff as green, other as orange [09:54] tsdgeos, design? [09:54] wyluzuj poslady Saviq! [09:55] mhr3_: design wants some green, some orange, so where does that come in the json? [09:56] Saviq, (I don't know what it means, just told my gf you're picky with my reviews this morning :D) [09:56] tsdgeos, i'm not convinced that should be configurable... way to make the previews eeek [09:56] and inconsistent [09:56] mhr3_: talk to design :D [09:59] tsdgeos, pinged rosie [10:42] bregma: heya. I started looking at unity stack a bit yesterday. compiz needed some changelog syncs. it's working nicely (I updated nux + compiz + unity) on my machine, but I filed a bug #1276516 about the autopilot situation on cu2d. [10:42] bug 1276516 in Unity "unity trunk failing autopilot tests (02/2014)" [High,New] https://launchpad.net/bugs/1276516 [10:43] bregma: to approach releasibility, we'd need some solution to those or how we'd validate the release via cu2d results. [10:44] tsdgeos, the green part will be defined by the scope author [10:44] tsdgeos, but don't worry about that yet [10:44] Saviq: ? [10:44] you mean a property [10:44] ? [10:45] Cimi, lol [10:45] tsdgeos, yeah, but later [10:45] ok [10:45] tsdgeos, for now we're only going orange [10:46] Saviq: there's also the width thing [10:46] buttons are "smart" enough to give themselves a proper width [10:46] but the OptionSelector doesn't seem to be able to do that [10:46] so do we force say 30gu of width? [10:46] or we have a property for that? [10:46] Cimi, re: implicitHeight vs. height, it's not "minimal", it's only "default", but in this case the widget won't adapt if you change its height, so it's not really default, it's the height it's meant to be [10:47] or try to fix optionselector to be smart? [10:47] tsdgeos, optionselector? where do we use optionselector? you mean combo button? [10:47] is there a combo button? [10:47] tsdgeos, there isn't indeed [10:47] well the docu says there is [10:47] tsdgeos, there isn't in the SDK [10:47] tsdgeos, hence we just need to go for multiple buttons now [10:47] so i thought it meant option selector [10:48] tsdgeos, nah, it's a different component that's "meant to be" in the SDK [10:48] which confusingly we then have something called valueselector also [10:48] Saviq: so just a row of buttons? or a column? [10:48] tsdgeos, column, yeah [10:49] tsdgeos, actually [10:49] tsdgeos, take what's in Notification.qml [10:49] tsdgeos, it's a poor man's combo button [10:50] hmmm [10:51] it's a column [10:51] isn't it? [10:51] tsdgeos, there's two buttons, and the left one opens the column [10:53] * tsdgeos reads the code more carefully [10:53] Saviq, length can be null? [10:54] just wondering about the function [10:54] also [10:54] in case it's null [10:54] Cimi, no, but foo["foo"] can be null [10:54] Saviq: besides that, you saw that tests failed in mako now?¿ [10:54] Saviq, modelData["length"] [10:54] Saviq, I'm passing this to lengthToString(s) [10:54] Cimi, let me see what I commented about [10:55] Cimi, I commented about modelData["subtitle"] that it could be null [10:56] Saviq, yes [10:56] Saviq, but what happens if length is too? [10:56] Saviq, it goes to that JS [10:57] Cimi, ah well, yeah, you should return "" if it's not a number [10:57] Saviq, I just have !== undefined [10:57] Cimi, if (typeof s !== "number") [10:57] cool [10:57] thx [10:57] I can drop undefined then [10:58] tsdgeos, grr [10:58] Saviq: i wonder how that happened, is because the tests were stopping before because the network ones failed? or? [10:59] tsdgeos, no, a rewrite of the datetime indicator was released, probably changed the name or something, checking [10:59] tsdgeos, or it's a valid failure and the indicator was broken [11:04] Cimi, and well, me being picky, is that a new thing? [11:05] Cimi, wanna know what "wyluzuj poślady" means? [11:05] Saviq, that could be good :) [11:05] Cimi, "loosen your cheeks", and it's not about your face ;) [11:05] Saviq, she said at first that all polish people are picky :) [11:06] Saviq, so what does it means? :) [11:07] Cimi, look up ↑↑ [11:07] Saviq, thought it was something related to smiling [11:08] Cimi, in a sense... [11:09] Saviq, she said it's something I can tell you in case we are friends [11:09] Saviq, so I wrote you :) [11:10] Cimi, yup [11:10] Cimi, not offensive at all (or well, as usual - depends on how you say it) [11:12] Saviq, I fixed your things and added test for NAN [11:12] Saviq: can't copy directly Notifications since it expands with left button instead of right, so you meant adapt right? I was thinking to do a base class that both Notifications.qml and this would use, but it'd be a work since behaviour is not exactly the same, so you think it's worth or just copy/adapt the code and wait for the combobox to land in SDK? [11:13] guys, I'll be away for few mins, need to reconnect router with new cat 5e cables [11:17] tsdgeos, we need to expand left button I believe [11:17] tsdgeos, https://docs.google.com/a/canonical.com/file/d/0B-a_7E3tDxOgUUY0MjFTQThBZFE/edit [11:17] why? [11:18] Saviq: that is not "the left button" [11:18] Saviq: each button is it's own widget [11:18] if the expectation is that in that screenshot [11:18] is that left button will do something like notification does [11:18] we can't do it [11:18] they don't know eachother [11:19] or we need to rethink the actions thing [11:19] tsdgeos, well, that's the question - don't they... [11:19] Saviq: i don't see why, it'd just see left one expanding down [11:19] and that's it [11:19] tsdgeos, I think we can make a "smart combo button" i.e. one button - on the right, two buttons - next to each other [11:19] more buttons [11:19] expanding from the elft [11:19] left [11:20] Saviq: i didn't understand at all what you meant [11:21] tsdgeos, think of a "row of buttons" as one widget of type: "actions" [11:21] which is not what how we were thinking of it, right? [11:21] tsdgeos, first one of those actions would be the button on the right, second - the button on the left, any more - expanded from the left one [11:21] tsdgeos, not before it was redesigned, no ;) [11:22] tsdgeos, but I think it might've been bad thinking [11:22] ETOOMUCHTHINKING [11:22] but you don't want the left one to do expansion [11:22] you want it to be a button and also expand [11:23] otherwise how do you actually click "Add to queue" [11:23] ? [11:23] tsdgeos, oh yeah, well, that's the missing combo button [11:23] tsdgeos, but we should be able to do something of the sort the Notification has [11:24] well that is what i was saying, put two buttons together and let it be similar to a combobox [11:24] i'll try to code it and we can talk over it later [11:24] Saviq: what is actually interesting if you see https://docs.google.com/a/canonical.com/file/d/0B-a_7E3tDxOgUUY0MjFTQThBZFE/edit [11:25] is why are the buttons one next to another [11:25] and not on top of the other [11:25] like all the other preview components [11:25] who decides that? [11:25] tsdgeos, that's the thing - I think that's a single widget [11:25] +meant to be [11:25] /should be [11:26] agreed [11:27] tsdgeos, so yeah, let's implement it with that in mind [11:27] need a new json for it [11:27] tsdgeos, no we don't [11:27] ? [11:27] tsdgeos, type: "actions" takes one or more actions [11:27] how come [11:27] yes [11:27] and more actions [11:27] is a combobox [11:28] tsdgeos, well, that's the thing [11:28] tsdgeos, not any moer [11:28] moar [11:28] but how do i put a button and a combo box in one actions [11:28] with that json? [11:28] tsdgeos, put more than 2 actions [11:28] and i have 3 buttons [11:28] tsdgeos, no [11:28] tsdgeos, 2 buttons, left a combo [11:28] doesn't make much sense [11:29] what if i am in a tablet [11:29] tsdgeos, you get two columnt [11:29] columns [11:29] why shouldn't i get to use all my width [11:29] tsdgeos, the widget shouldn't change whether on tablet or phone at all [11:29] tsdgeos, because that's wasting the width [11:29] instead of a button and a combo? [11:29] tsdgeos, that's output from Cape Town [11:29] ok [11:29] tsdgeos, no button to span the whole width of a phone [11:29] tsdgeos, as that wastes space [11:30] sure, i'm not saying that [11:30] i'm saying 2 buttons (one a combo) when i could have 3 is a bit weird [11:30] but ok [11:30] if you want that [11:30] i'll code that [11:31] tsdgeos, if you want three, that's when you put two widgets - one with two actions, one with one action [11:32] ok [11:32] man... I got a headache today [11:33] tsdgeos, I'll add a comment to the dash spec, 'cause it doesn't deal with that, but I think it's the best route [11:35] tsdgeos, found the issue with the tests - datetime takes longer to load now it seems, and the test didn't wait for it [11:35] insteresting [11:36] Saviq: so just to confirm, with this new definition on how the json works, you can never have just a combobox, if you want a combobox, you'll always hace a button on it's right too, correct? [11:36] tsdgeos, yes [11:36] ok [11:37] tsdgeos, you get one button, two buttons, or a button and a combobox [11:41] tsdgeos, if you disagree with that, comment please on https://docs.google.com/a/canonical.com/document/d/1n880Fih5KyGPcoP5chidnHDG_8TxXUgSuij7f4rHpuk/edit?pli=1&disco=AAAAAH2p9p4 [11:41] Saviq: i think it limits what people can do [11:41] but i'm not the designer so i'm ok [11:41] tsdgeos, we're doing that all over the place don't we :) [11:41] tsdgeos, what you disagree? [11:42] tsdgeos, "I'm not a designer so I'm ok" is not the answer I want to hear from anyone working on my team :) [11:42] tsdgeos, in some areas you have more experience than designers, so share your issue with me :) [11:43] cimi: i'm not disagreeing with designers, i'm disagreeing with Saviq :D [11:43] tsdgeos, amazing! [11:43] lol [11:43] i do think that in some case it may make sense to just have a combobox [11:44] but since i am not designing the interface i don't know if that makes sense in the previews or not [11:44] and if it does, we can add it later, is not that as if the code i'm writing now won't even change [11:44] tsdgeos, we could add a prop to actions [11:44] tsdgeos, that would say "I want all to go into combo" [11:45] tsdgeos, but yeah, we can add that later if needed [11:47] tsdgeos, https://code.launchpad.net/~saviq/unity8/fix-indicator-tests/+merge/204889 [11:48] * tsdgeos runs the autopilot tests [11:48] * tsdgeos dies of old age [11:49] tsdgeos, try to run them on emulator ;D [11:49] he'll be dead before the emulator has booted :P [11:50] Saviq: you miss the guidelines thing [11:50] tsdgeos, refresh [11:50] ah, not anymore [11:54] cimi, 56 if (s <= 0 || typeof(s) !== "number") return ""; [11:54] cimi, I think those should be reversed [11:54] cimi, we don't want to even look at whether s <= if it's NaN [11:55] cimi, text: (modelData["subtitle"] && modelData["subtitle"] !== undefined) ? modelData["subtitle"] : "" [11:55] cimi, text: modelData["subtitle"] || "" [12:01] tsdgeos, just a small tweak to the docs: https://code.launchpad.net/~aacid/unity8/image_gallery_preview_widget/+merge/204712/comments/478927 === MacSlow is now known as MacSlow|lunch [12:08] Saviq: do you like the new styling i just pushed more? [12:09] tsdgeos, +1 [12:09] tsdgeos, I'll put * everywhere after the next release [12:30] Saviq: so on the "combo - button" scenario, i'm still unsure of how we want the thing to work [12:30] obviously pressing the right button is pressing it, that's easy [12:30] but do we want the combo to have two sensitive areas, one for activate and one for "open"? [12:30] and when the combo opens, do we want the first button to hide like in notificationss? [12:31] and when the combo is opened, does clicking on one of the opened options activate it? or just select it and close the combo and then you have to click on the combo? [12:31] in the non "open" area to activate it [12:33] tsdgeos, *we* do the simplest thing that makes sense [12:33] tsdgeos, I think the combo button is desgined... somewhere... [12:37] tsdgeos, Jouni should have them, he should be online in an hour or two (as should SDK folks) [12:37] ok [12:46] Saviq: no worky https://jenkins.qa.ubuntu.com/job/generic-mediumtests-runner-mako/5078/#showFailuresLink [12:47] or maybe it did fix 1 [12:56] tsdgeos, that's another issue... [12:56] tsdgeos, they broke app launching in the latest image... [12:56] tsdgeos, we'll have to wait for the next one === alan_g is now known as alan_g|lunch [13:13] is launchpad slow today or is it my connection? [13:14] hmm, getting several launchpad "Oops!" pages now... [13:22] o/ === MacSlow|lunch is now known as MacSlow [13:33] dandrader, oops from launchpad suggest LP issues, your connection shouldn't have any bearing, no? [13:33] karni, hey [13:33] Saviq, yes, I got those Oops after I notices the overall launchpad slowness === mardy_ is now known as mardy [13:33] noticed [13:33] hi Saviq :) [13:34] dandrader, so yeah, it seems pretty slow here [13:40] mhr3_, so, scope::preview() only gives out a single PreviewModel now, is there a multi-column version anywhere yet? === alan_g|lunch is now known as alan_g [13:47] saviq, not finished yet [13:51] aaagh stupid QtCreator... every single time I create a folder with the name of the file I want to create... [13:56] Saviq: so arrpove yours that indeed seems to decrease the number of failing tests? [13:56] tsdgeos, well, it fixes that one failure [13:56] tsdgeos, so yeah [13:57] tsdgeos, and we wait for improvement on the other side [14:01] Saviq: duude [14:01] mzanetti_, what up? [14:01] I constantly get: mzanetti Nickname already in use, try a different one. [14:01] Saviq: do a whois mzanetti [14:02] mzanetti_, I am! [14:02] * mzanetti_ wonders why you can hold that one even though I am identified @ NickServ [14:02] * tedg wonders if IRC explodes if you try to whois yourself [14:03] whois mzanetti === dandrader is now known as dandrader|afk [14:03] well, with the / [14:03] :P [14:03] yeah yeah :P [14:03] need to log in to disconnect... [14:04] mzanetti_, here === mzanetti_ is now known as mzanetti [14:05] Saviq: cheers [14:05] still wondering how this is possible [14:05] but I never really got irc authentication anyways [14:09] mzanetti, you have to explicitly ask NickServ to get your nick back [14:09] ghost him! [14:10] tedg: seen the comment i made in your anywhere MR? [14:10] tsdgeos, Yup, just did. [14:10] ahhh [14:10] tedg: that's going to conflict [14:11] K, I'll deal. [14:11] with https://code.launchpad.net/~unity-team/unity8/unity8.previews_textSummary/+merge/204177 [14:11] but that's not approved for the moment [14:11] so do yours and see if we can sneak it in [14:12] this delayed landing is a pain for merges [14:12] because you have to think about billions of branches that may or may not conflict with yours [14:12] Hah, you're telling me! :-) [14:14] I'm starting to loose track of which branches I've told other people to run so they get the functionality they need. === sdeziel is now known as sdeziel_ [14:17] tsdgeos, http://pastebin.ubuntu.com/6879267/ ? [14:17] that is what i have yes [14:18] kind of [14:18] your docu is better === dandrader|afk is now known as dandrader [14:19] tsdgeos, So I ran it, then I deleted the changes that weren't related to my change. [14:19] tsdgeos, That should reduce the conflict [14:19] maybe [14:19] or not [14:19] who knows :D [14:20] Well, yes, it should reduce the likelihood of conflict ;-) [14:24] Saviq: re: your comment, I did that for title only because that was critical fix for us. I'd be happy to continue work in anothe branch, if that's okay with you. https://code.launchpad.net/~unity-team/unity8/new-scopes-title-alignment/+merge/204677 [14:24] karni, ok [14:27] tsdgeos, PreviewActions.qml? [14:27] ignore that [14:27] Saviq: you're complaining on the "s" ? [14:27] tsdgeos, lol [14:27] tsdgeos, no, asking whether that's the name of the widget ;D [14:28] * Saviq should be less picky in reviews [14:28] i had PreviewActions [14:28] moved to PreviewAction [14:28] but i guess i have to move it to PreviewActions again [14:28] given the new meaning we gave the json [14:28] tsdgeos, yeah, plural :) [14:28] tsdgeos, well, the json was always "actions" for one-or-more actions :) [14:28] i know [14:29] but didn't like it ^_^ [14:29] tsdgeos, saviq, what's the conclusion on the colors? do we have new prop? [14:29] what [14:29] mhr3_, not yet [14:29] mumble [14:29] mumble: symbol lookup error: mumble: undefined symbol: _ZN6google8protobuf18GoogleOnceInitImplEPlPNS0_7ClosureE [14:29] :S [14:29] yay [14:29] tsdgeos, yup, no mumble today [14:30] i can trying using the android client on the Z10 [14:30] let's see how it flies :D [14:30] tsdgeos, I can't :P [14:30] @unity: hangout today, mumble doesn't load with ↑↑ [14:31] https://plus.google.com/hangouts/_/72cpiqua3mnkr7dti836vn0ui0 [14:31] Saviq, will take me a while to jump on ghangout (have to turn on my laptop, etc) [14:31] dandrader, sure [14:31] SU notes: attending sprint sessions, letting Saviq kick ass and save the world before mwc, ha ;) reviewing code, etc [14:31] karni, you'll have to put it in yourself ;P [14:32] oh there's a new protobuf deb comin in dist-upgrade [14:32] maybe fixes the thing [14:32] Saviq: come again :)? [14:32] I'm trying to join too, but need to install the plugin. [14:32] Hi all [14:32] I'm running Trusty and installed some updates (one of them was lightdm) and rebooted. Since then, Unity doesn't autostart so I get a bare desktop. Any idea where to look for debugging? [14:32] karni, in the doc I mean ;) [14:32] Saviq: haha, I will [14:33] guys, mind throwing the SU doc at me? [14:34] found it [14:36] kgunn: -mumble +hangout [14:36] kgunn: https://plus.google.com/hangouts/_/72cpiqua3mnkr7dti836vn0ui0 [14:36] Saviq, hm... I'm not allowed to connect to that hangout... [14:36] MacSlow, https://plus.google.com/hangouts/_/72cpiqua3mnkr7dti836vn0ui0?authuser=1 [14:38] elopio, ↑↑ we're doing hangout today as mumble's dead [14:39] Saviq, I'm in, thanks. [14:39] elopio, d'oh [14:45] mterry, ping. Do you have time now? [14:46] elopio, sure! Let me dig into how system-settings is called over dbus [14:48] tsdgeos: so when would that qt release actually hit ? (...thinking of risk & mwc being right around the corner) [14:48] any idea on what's wrong with Unity not starting after login? [14:48] altho...maybe not so big deal since Saviq says we're _virtually_ on that rel anyway [14:49] kgunn, wait wait wait [14:49] kgunn, we're not switching for mwc [14:49] kgunn, we have 5.2 (and some of 5.2.1) in qt5-beta2 ppa [14:49] kgunn, so there's little risk between that and full 5.2.1 [14:50] ok, np.... [14:50] kgunn, but we still have bugs to fix for it (although apparently not unity8 ones!) [14:51] right, is that phonedations that makes the call for when to pull in a new Qt rel ? [14:51] kgunn, QA rather [14:51] mmm [14:52] kgunn, basically when we get green on 5.2 and no bugs can be identified - that's when [14:53] of course.... [14:53] was wondering if there was even a "conformance test suite" in addition to that... [14:53] guessing not [14:59] karni, https://code.launchpad.net/~unity-team/unity8/new-scopes-title-alignment/+merge/204677/comments/479092 [15:03] AAARGH [15:16] mhr3_, action activation will happen on the preview object? [15:17] yes [15:17] mhr3_, any name for the method yet? [15:18] saviq, didn't pick yet, feel free to [15:18] mhr3_, we're using triggered(string, string, QVariantMap), how's that? [15:19] fine with me [15:19] mhr3_, I'm starting to feel like we need a "separator" widget... [15:20] mhr3_, as I can't find a rule in the visuals for where to put it and where not [15:20] :/ [15:21] i can see scopes specifying margins in px soon [15:21] mhr3_, let me see if katie has an idea [15:21] <-- katie has quit (Ping timeout) FAIL [15:22] ok actually found a rule... [15:22] maybe [15:31] mhr3_, btw, we discussed roleNames a while ago, and decided there's no point in keeping them in m_roles, as they will only ever get called once, so we're usually creating them in roleNames directly and letting it drop when going out of scope [15:32] saviq, hm, not like it used up insane amounts of memory [15:32] mhr3_, of course, just saying [15:32] yea, will keep in mind === alan_g is now known as alan_g|tea [15:39] saviq [15:39] Saviq, is this fine now? https://code.launchpad.net/~unity-team/unity8/unity8.previews_audioPlayer/+merge/203387 [15:39] "card-size": "medium", // "small", "medium", "large", 12 ÷ 38 (in grid-units) when "category-layout": "vertical" [15:40] should have been card-layout, right? [15:40] mhr3_, of course [15:40] i was confused if it wasn't supposed to be vertical-journal [15:42] actually [15:42] wasn't it supposed to be "category-layout": "vertical-journal"? === alan_g|tea is now known as alan_g [15:48] saviq, is there max for xollapsed-rows? [15:48] collapsed* [15:49] mhr3_, man, I should probably read what you write [15:49] you should :P [15:49] mhr3_, it should be journal [15:49] mhr3_, in vertical size is s, m, l - only in journal you can say what height every row is [15:50] mhr3_, spec defines collapsed-rows per category layout [15:51] saviq, i wonder if we have a maximum value there [15:51] 1024? :) [15:51] mhr3_, no, like... 2 [15:51] so, 1 or 2... lots of options [15:52] mhr3_, indeed [15:52] collapsed-rows Number of result rows displayed while the category is collapsed; possible values: any integer between 1 and 2 [15:52] doesn't sound right :) [15:52] lol [15:52] mhr3_, it only affects grid and journal, too ;) [15:52] mhr3_, sounds like we should make it bool [15:52] saviq, i'm starting to think you should have been writing this ;) [15:53] mhr3_, collapsed-2-rows: true ;) [15:53] heh [15:59] Saviq: can get lp:~aacid/unity8/action_preview_widget and run "make tryPreviewActions" [15:59] and tell me what you think? [15:59] I think it's not great, but unless I should be coding the combobox (which will take some time) I don't see anything better we can do [16:05] tsdgeos, hmm yeah that looks fine, we'll only have two buttons at the most in our previews I think [16:06] ook [16:20] mhr3_, will I still have "id" and "type" in preview widget properties, or are you stripping them? [16:21] saviq, they're separate roles, not in the big variantmap [16:21] mhr3_, bleh :P [16:25] you should be glad! [16:25] it's guaranteed to always be there [16:25] and always a string [16:40] Saviq, ping? maybe I missed your answer [16:40] Saviq, how is the preview widget now? [16:41] Cimi, you didn't, just ACK'ed it [16:41] Saviq, maybe tsdgeos could review it [16:42] Cimi, he did already, didn't he [16:43] Cimi, if you get him to do it - be my guest ;) [16:43] Saviq, now again [16:47] do what? [16:49] tsdgeos, do a last review of audioPlayer (probably just the last commits would be enough) [16:50] mhr3_, properties^Wattributes... widgetId, type... yay for consistency ;P [16:51] consistency is relative :) [16:51] Saviq: looks ok to me === dandrader is now known as dandrader|lunch [17:03] sil2100, hey, any idea why https://code.launchpad.net/~marcustomlinson/unity-scopes-api/smartscopesproxy_fixes/+merge/204827 hasn't been merged yet? [17:03] hmmm [17:04] console.log(spy.signalArguments[0][0]); [17:04] should give me the first paramter of the first signal that was spied, right? [17:23] Saviq: I'm still waiting for 2 branches of unity-mir to land. Did they fall off the CI train? [17:24] greyback, http://i.dailymail.co.uk/i/pix/2009/05/25/article-1187119-051592F9000005DC-459_634x421.jpg [17:24] Saviq: in that case, can I squeeze this in: https://code.launchpad.net/~gerboland/unity-mir/fix-readme/+merge/204572 [17:24] greyback, sure [17:24] Saviq: it needs a quick review plz [17:25] * greyback appalled Saviq reads the daily mail [17:25] greyback, you should stop supporting in-source builds [17:26] patches welcome :) [17:26] greyback, why the -DCMAKE_INSTALL_PREFIX=/usr ? [17:27] Saviq: otherwise installs in /usr/local [17:27] greyback, well, yeah, you shouldn't "make install" it anyway, should you.. [17:29] greyback, but yeah, ok [17:29] Saviq: it's fast & dirty, just wanted to update the wrong info, tis all [17:29] greyback, got it [17:30] greyback, ACK [17:31] mhall119, hey, so we've started generating docs from unity8 (doxygen style), think it'd make sense to publish them somewhere? [18:02] mhr3_, hrmpf "Error: Unknown method return type: scopes_ng::PreviewModel*" [18:02] but Q_DECLARE_METATYPE(scopes_ng::PreviewModel*)!! [18:02] mzanetti, ↑ you solved that error for yourself recently with exactly that, right? [18:03] ah hmm [18:03] not exactly this [18:03] it was a parameter, not the return type [18:03] but yeah. I guess its the same [18:03] Saviq: I'll address the rest of the comments on card-overlay branch [18:04] mzanetti, I think in this case it's because it's not wrapped in QVariant maybe [18:04] karni, cool, I got previews *almost* working [18:04] Saviq: fantastic :) [18:04] huh... no. its a pointer. don't wrap that in QVariants === dandrader|lunch is now known as dandrader [18:08] mzanetti, yeah, don't listen to me [18:08] mzanetti, was thinking of different place in the code [18:08] :) no worries [18:08] ah, it's not qmlRegistered [18:10] Saviq: why doxygen? [18:10] mhall119, as opposed to? [18:11] QDoc === alan_g is now known as alan_g|EOD [18:17] Saviq: I think that unity8-card-overlay willl have to be against new-scopes. CardCarousel.qml is not present in unity8 (only contains DashCarousel, while it is CardCarousel.qml that should override whether overlay is shown on current item) [18:17] whatcha think? [18:18] s/against/rebased on new-scopes [18:18] karni, https://code.launchpad.net/~unity-team/unity8/newscopes-card-overlay/+merge/204797 [18:18] haha [18:18] * karni marks WIP [18:18] Saviq: shall I just continue work in that branch? [18:21] karni, yup, is why I pushed it under unity-team [18:21] thanks bug [18:22] *bud :D [18:26] Saviq: I ask about qdoc because that's what all of the SDK uses already, and I have an importer for qdoc into the API website for publishing [18:26] for doxygen somebody would have to write an parser/importer to get it published to the API website [18:27] mhall119, oh well, I thought doxygen was meant to be used... [18:27] mhall119, fortunately they're compatible [18:28] in that case I can set you up on developer.ubuntu.com/api/unity// for qdoc generated stuff [18:28] developer.u.c/api is just a reverse proxy too, so we can set up unity.u.c/api/ to point to it too [18:29] ...maybe [18:29] well, ok, on second thought that would take some work [18:29] mostly because of the website theme [18:31] mhall119, if you can try and see how this will work out, that'd be nice [18:33] Saviq: worst case scenario, we deploy a separate instance of the API website just for Unity [18:33] it's all Juju charmed now, so that shouldn't be hard [18:33] (famous last words) [18:34] :) [18:34] mhr3_, http://paste.ubuntu.com/6880650/ === boiko_ is now known as boiko [18:44] Saviq: I assume the shader should fade in together with CardHeader (as if they were one thing) [18:45] I'll do it for text first. Once that works, I'll include the shade effect. [18:49] karni, yeah, fade in text, too === jono is now known as Guest56083 [19:11] Saviq: another one of my cmake hacks :D https://code.launchpad.net/~mzanetti/unity8/make-single-testfunction/+merge/205035 [19:12] mzanetti, huh, is $(FUNCTION) something cmake knows? [19:12] no [19:12] Saviq: well, sort of yes [19:12] its make [19:12] $(FOO) can be used to pass arguments to make [19:12] but not real arguments [19:12] only vial FOO=blabla [19:13] mzanetti, will it work with ninja? [19:13] * mzanetti tries [19:13] works with nmake according to stackoverflow [19:14] https://bugs.launchpad.net/unity8/+bug/1276747 [19:14] Launchpad bug 1276747 in autopilot (Ubuntu) "after starting a new app cannot get proxy object for unity8" [Undecided,New] [19:14] saviq, eh, yea, sorry [19:14] om26er, yeah, saw that - affected autopilot, too, we're not doing anything special around this [19:14] saviq, will add to my previewstack branch [19:14] mhr3_, nw, it excepts anyway [19:15] ;) [19:15] sucks :P [19:16] Saviq, thanks for that [19:17] Saviq: too bad... ninja doesn't like the $ [19:17] wants $$ [19:18] but then it breaks [19:23] mterry, https://code.launchpad.net/~elopio/unity8/url-dispatcher_test/+merge/205037 [19:33] ok, finishing early today ;P [19:46] Saviq: early you say :D 8:30 haha. If you happen to drop around, the isCurrentItem attached property doesn't seem to work for me :| http://paste.ubuntu.com/6881024/ [19:47] btw your overlay is lovely [19:47] font color: http://paste.ubuntu.com/6881030/ [20:04] karni, "=== true" can probably be dropped [20:04] ack [20:05] just pushed the version that doesn't work with attached property isCurrentItem [20:05] left a comment, don't know why it doesn't work. [20:05] karni, http://developer.ubuntu.com/api/qml/sdk-1.0/Ubuntu.Components.Themes.PaletteValues/ use these for the default color [20:05] if you put true/false, it toggles overlay on all items, so the attached property should work [20:05] Saviq: I asked katie for those, she just said "white" in overlay xD thanks! [20:06] karni, attached properties are only available on the direct delegate item [20:06] the Loader in that case.. [20:06] karni, yeah, for overlay white, but for the default - use one foregroundColor [20:06] * karni nods [20:06] karni, not Loader, Card? [20:07] karni, ah I know what you mean [20:07] that in that case, that'd be showHeader: isCurrentItem [20:08] and that doesn't work either [20:08] actually, it says ReferenceError: isCurrentItem is not defined [20:08] we use Loader as the delegate in Carousel [20:08] maybe that's the reason [20:08] karni, no, you're right - ListView.isCurrentItem would only be available in the Loader [20:08] parent.ListView.isCurrentItem works [20:09] within Card { } in CardCarousel [20:09] just found out by trial and error ;P [20:09] karni, tricky ;) [20:09] makes sense, as Loader would be parent of Card, but the Card becomes the delegate so.. not very intuitive [20:10] hahahah, when you click - isCurrentItem changes. when you scroll the carousel, it doesn't :D [20:10] challenge! [20:10] let my try that on my own, tho! xD [20:11] actually, got a session in 5 minutes, but I'll try to set current item from within Carousel component [20:11] karni, showHeader: explicitlyScaled [20:11] smart xD [20:12] karni, that's what it's there for (but the name is tricky) Cimi, you and your prop names :P [20:12] SLICK [20:12] hahah [20:12] yeah, naming is one of two worse problems in programming. the other is cache invalidation ;P [20:12] *proper naming [20:18] karni, to only use a single animation, you can go: opacity: header.opacity * 0.6 for the overlay [20:19] yeah, I thought it'd be better to use one (or pull Shader under CardHeader with z: -1 ) [20:23] karni, not sure what that has to do ;) [20:23] karni, opacity is inherited, but z-order does not change parentship [20:23] Saviq: I thought I could use one "Behavior" declaration, but I guess I can't (like, one animator, if that even matters) [20:24] ok, now I get it heh, sorry :) [20:24] karni, you could, though, set parent: of the header to the overlay, when in overlay mode [20:24] :D [20:24] karni, and sure, you could use a single Animation object [20:24] karni, but that's effectively what opacity: header.opacity * 0.6 does [20:25] * karni nods, fixed [20:25] karni, with less code, 'cause you don't have to explicitly configure the Animation [20:25] Correct, sir. Pushed [20:26] Shall I leave it WIP to center align title within overlay (whish works in vertical mode, so Left aligned) ? [20:26] but seriously, you should go enjoy your evening pal :) [20:26] karni, not sure it should center in overlay [20:26] Saviq: that's what's in design doc, katie added overlay section [20:27] karni, I'll enjoy my week starting Saturday, when I'm off snowboarding in Austria, but first I need to earn it ;) [20:27] :O wohooo \o/ [20:27] Saviq: leaving for a week? you've certainly earned it [20:28] karni, yeah, I just want to leave you guys in a state where you can mostly apply tweaks to the whole system, rather than have to implement huge portions of it [20:28] :)) [20:29] Appreciated [20:29] Very much [20:29] Saviq: booo...leaving [20:29] enjoy [20:30] hope there's snow [20:30] karni, you should run stuff before pushing ;) [20:30] opacity: opacity: header.opacity * 0.6 [20:30] d'uh [20:30] that's what happens when trying to work in middle of session, my bad :/ [20:30] was sure I :s 'ed ;P [20:30] kgunn, hold your horses, I'm still here :D [20:31] fixed, sorry about that :// [20:31] Saviq: oh right...8 not 6 [20:34] karni, re: margins, we should probably push them down into CardHeader, no need for them to be in two places [20:34] karni, /me does [20:34] :D was about to say "consider it done" [20:34] ok. /me listens to the session [20:41] karni, height: header.height is not good, we don't want to make overlays of different height in case there's only one line of title or missing subtitle [20:41] karni, CardTool needs to give up headerHeight [20:43] Saviq: ack [20:43] I'll figure it out [20:44] Saviq: I assume it'd be either 2 lines (for title) or 3 lines (for title+subtitle, if both are mapped) [20:45] That sucks, cose we'll have 2 lines even if no title wraps to two lines [20:45] :( [20:46] but I guess that's better than have different height, yes [20:46] karni, it sucks less than overlay going up and down [20:46] * karni agrees [20:46] karni, or well, maybe it depends [20:46] karni, as in Carousel you only ever show a single overlay [20:46] maybe I should ask katie? [20:46] right [20:46] karni, but in Grid you show them all [20:46] good point [20:46] actually, very good observation [20:46] karni, so yeah, asking for design input would be good [20:46] I'll do that [20:48] karni, I moved most of your changes into unity8-card-overlay [20:49] karni, and merged+pushed into newscopes-card-overlay [20:50] * karni looks [20:51] Looks great, Saviq, thank you [20:53] * Saviq really likes it how we managed to keep the diff between old and new scopes really small [20:56] Yea, you did it fantastic === shiznix_ is now known as shiznix [21:45] tedg, I'm having trouble launching apps through upstart-app-launch, are there any cunning debug tricks? [21:46] bregma, Which branch are you using? [21:46] trunk ... in a Unity8 session on Mir on the desktop [21:47] bregma, K [21:47] bregma, So I'd set the glib debug flag [21:47] bregma, initctl set-env --global G_MESSSAGES_DEBUG=all [21:47] bregma, That'll put more interesting stuff in ~/.cache/upstart/application* [21:48] sounds like a good start [21:48] bregma, Wait, are you using trunk or main? [21:48] Like what's released in archive [21:48] whatever's in the release archive [21:48] K, that should be your next step then. [22:01] tedg, does upstart-app-launch/desktop-exec reset the environment or something? [22:02] bregma, It configures it, upstart creates a clean environment each time. [22:03] bregma, You can get it with initctl get-env [22:04] well, I'm setting the QT_QPA_PLATFORM explitcitly with initctl set-env (and env directives) but the app crashlog does not show anything but a few basic variables [22:05] set-env --global ? [22:05] What is generally best is to put that in an upstart job that is "start on starting dbus" so then it gets into the dbus env as well. [22:06] bregma, This is my ~/.config/upstart/glib-debug.conf as an example: http://paste.ubuntu.com/6881793/ [22:51] mterry: are you still here? [22:51] elopio, yup [22:51] I don't understand why url-dispatcher stops working when we do the initctl stop unity8 [22:51] elopio, I'm looking at things now too [22:51] is the unity process we launch on the tests missing something? [22:51] elopio, does autopilot know the right DBUS_SESSION_BUS_ADDRESS? [22:53] mterry: I bet not, but you said that comes from an environment variable, right? [22:53] elopio, you might be thinking of me saying about MIR_SOCKET [22:53] elopio, I'm talking about url-dispatcher being able to see dbus [22:53] elopio, I'm testing this idea now [22:54] elopio, no, that's fine [22:55] my previous tests restart unity8 and still can talk to the session dbus. But I barely know what I'm talking about :) [22:56] my guess is that initctl start unity8 does something we are not doing on the test's launch unity. [22:56] was just a though [22:56] hm [22:59] elopio, if I enter a bogus url or whatever to dispatcher, I get an error [22:59] I don't see that here [23:01] mterry: could the camera app be behind unity? [23:01] elopio, camera-app isn't running [23:02] mterry: it is here. [23:02] oh , let me try again [23:03] elopio, right you are! [23:04] How does AP launch unity8? I thought it was via upstart too [23:05] good :) [23:05] sadly, that leaves me as clueless as before, jaja. [23:05] mterry: it does initctl start unity8, but it also does a lot of other things I'm looking for the first time [23:05] os.unlink( [23:05] os.getenv('MIR_SOCKET', [23:05] os.path.join(os.getenv('XDG_RUNTIME_DIR', "/tmp"), [23:05] "mir_socket"))) [23:05] I don't know what's that about. [23:14] elopio, that's just cleaning up from previous runs [23:14] elopio, shouldn't be anything to worry about [23:19] mterry: you got your wish...0.1.4 mir is landing now...its actually in devel-proposed image now [23:19] kgunn, sorta. I want 0.1.5 though! [23:19] elopio, odd results [23:19] and it has the orientation argument in the configure_output [23:19] mterry: it took so long.... [23:20] i just remerged devel into the MP :) [23:20] kgunn, oh good! [23:20] duflu set it up perfect so the so bumps were in devel...but the deb bump is only in trunk [23:20] so i didn't have to wait [23:20] enjoy! [23:20] kgunn, now we just have to get the new AP tests sorted. Neither one is working even against current image right now [23:20] eegads [23:20] kgunn, tests are hard [23:20] lol [23:21] mterry: no kidding i keep telling people this [23:21] theyu [23:21] they're like hey...just get a guy to do a test for 2 weeks... [23:21] elopio, so when I manually run camera-app (sudo -u phablet -i camera-app) I get different results between the two runs (normal and AP) [23:21] i'm like...uh, no...it takes maintenance [23:21] elopio, normal gives me "QUbuntu: Could not create application instance" [23:21] ok, gonna drop for abit...later [23:22] elopio, AP gives me http://paste.ubuntu.com/6882127/ [23:22] kgunn, bye [23:22] elopio, why would we get those results? [23:23] elopio, neither brings up camera-app [23:24] mterry: I have no idea. [23:24] veebers should know more about this code, but I think new zeland is on holidays. [23:26] mterry: at what point of the autopilot test are you trying to launch the camera? [23:27] elopio, I'm doing my current testing in a separate adb terminal and running the camera-app manually either during AP test or during normal operation [23:27] elopio, not running it from the test, just during the test [23:28] right. It's weird, the set up of the unitytestcase must be playing tricks with the environment. [23:29] elopio, but it's weird that during AP test, it seems to get further [23:38] mterry: this is how we are starting unity on the tests: [23:38] /sbin/initctl start unity8 BINARY=/usr/bin/unity8 ARGS= XDG_DATA_DIRS=/usr/share/unity8/mocks/data:/usr/share/ubuntu-touch-surfaceflinger:/usr/local/share/:/usr/share/ LD_LIBRARY_PATH=/usr/lib/arm-linux-gnueabihf/unity8/qml/mocks/libusermetrics:/usr/lib/arm-linux-gnueabihf/unity8/qml/mocks/LightDM/single QML2_IMPORT_PATH=/usr/lib/arm-linux-gnueabihf/unity8/qml/mocks QT_LOAD_TESTABILITY=1 [23:38] if I do that manually, and manually launch the camera, it doesn't appear in front. [23:38] ok, good [23:38] so we can start taking away things :) [23:39] oh we have quite a few mocks there [23:43] elopio, /usr/share/unity8/mocks/data in XDG_DATA_DIRS [23:44] hm [23:44] maybe something else too [23:45] elopio, that and dropping QML2_IMPORT_PATH=/usr/lib/arm-linux-gnueabihf/unity8/qml/mocks [23:46] mterry: yes! without those two, it works. [23:49] there is a fake application manager, maybe not updated to work with url dispatcher? [23:50] it seems to be faking too much. [23:51] elopio, why are we faking that much in AP mode? [23:51] * mterry looks at the delta for the mock and real thing [23:52] mterry: I don't know. I'm a new comer. I added one fake, that was to avoid talking to the click servers [23:52] I suppose there are some good reasons for these others. [23:52] * elopio makes a mental note to always add a comment on the reason for using a fake.