=== JanC is now known as Guest1334 === JanC_ is now known as JanC === aaron is now known as Guest77720 [06:56] mzanetti: hi! can you update on https://requests.ci-train.ubuntu.com/#/ticket/1411 (which is updating just to have it in sync with overlay)? I see in the bug report Pat had a problem with hang with the test app, but no comment from you? === maclin1 is now known as maclin === maclin1 is now known as maclin [09:09] Mirv: welcome back?¿ [09:12] tsdgeos: that too :) === olli_ is now known as olli [12:15] Mirv: when you have some time have a look at silo 001 [12:15] it has a small patch that'd be nice to land [12:20] tsdgeos: I have noticed it [12:21] seems pretty safe and long in 5.5 and 5.6 [12:25] yep [12:25] i was trying to get sil2100 to land it but i guess he's been busy [12:25] if you can take over that'd be nice [12:41] tsdgeos, Mirv: it's in QA's hands since Friday [12:41] Nothing more can be done with it [12:41] sil2100: ah, awesome [12:42] sil2100: if this is the blocked silo with your name on then it is blocked :P [12:42] tsdgeos: sorry it took so long, I was actually almost sure I switched it to Approved looong way ago [12:42] sil2100: no worreis [12:43] tsdgeos: could you just comment on the trello card in my stead? ;) You know better which UITK test failures it fixes [12:43] sil2100: url? [12:44] tsdgeos: https://trello.com/c/M75g1XTv/3428-1630-ubuntu-landing-001-qtdeclarative-opensource-src-qtdeclarative-opensource-src-gles-sil2100-tsdgeos [12:49] i have stupid trello trying to format things [12:49] and not having any kind of "what kind of stuppid formatting we use" link [13:20] sil2100: tsdgeos: right! there's then another qtdeclarative upload to be prepared after that one has landed first, for dandrader [14:06] dednick, is this snap decision dialog doable today without changes on unity8 side? https://wiki.ubuntu.com/Networking#Inserting_a_new_SIM [14:06] dednick, i.e. can a user specify button color to be red and such? [14:12] mterry: give me a sec [14:12] I see we can set x-canonical-private-rejection-tint... [14:13] dednick: ermmmm. [14:13] i'll have to look at the dialog code. [14:13] dednick, But can't seem to control which button gets the tint? [14:13] It seems strictly based on order [14:13] but should be fairly easy to ascertain [14:13] mterry, dednick: I should know [14:14] ltinkl, oh hey ^ [14:14] :) [14:14] mterry: i think the order is fixed. [14:14] as in which one is accept action, which one is reject action [14:14] mterry, dednick: the order is defined, 0 accept, 1 reject, then everything else [14:14] "as designed" [14:15] sounds entirely straight forward ;) [14:15] mterry, dednick: look in Notification.qml, around line 480 [14:15] ltinkl, got it... [14:16] ltinkl, dednick: so... what should we do for this dialog? :) [14:16] "tint" here means the button gets green/red, nothing else [14:16] otherwise it's a regular button [14:16] ltinkl, dednick: x-canonical-private-reverse-tints? :-/ [14:16] mterry: well, the red one is the reject one. [14:16] I wish we could associate a tint with a particular action index [14:16] which is button 1 [14:17] dednick, that looks like button 0? [14:17] which is "restart" in the picture no? [14:17] mterry, not quite sure what you _want_ to achieve :) [14:17] dednick, yeah, red == reject, number 1 [14:17] 0 would be accept [14:17] ltinkl, in that picture, is the Restart Now button index 0 or index 1? [14:17] i would seriously hope the one on the right is 1. [14:18] otherwise we ave a very strange indexing system. [14:18] mterry, dednick: the red (restart) is 0 [14:18] yeah [14:18] as in "accept" [14:18] do it now :) [14:18] dednick, buttons go reverse in dialogs [14:18] yup, they get filled from the right side [14:18] of course they do! [14:19] dednick, as designed :p [14:19] dednick, well there's some thought behind it -- your eye ends up in the bottom right by default -- so we want the "preferred" action there [14:19] GNOME does it too [14:19] mterry: i guess the restart needs to be green then. [14:19] ltinkl, so yeah, thoughts on best way to get that dialog in the spec? [14:20] yeah, Gnome and MacOS, the default/accept/OK button is always the rightmost one [14:20] dednick, except it's a destructive action [14:20] mterry, just emit a notification [14:20] mterry, I suppose from indicator-network? [14:20] ltinkl, right... but to get the index==0 button to be red... [14:21] dednick, red is for "negative and irreversible actions" [14:21] It just also happens to be the "preferred" action here [14:22] mterry, ah... hmm, no better suggestion than patching Notification.qml [14:22] should probably have an x-canonical-private-affirmative-index / x-canonical-private-rejection-index. [14:23] ltinkl, dednick: rejection-tint seems bizarrely specific (in assuming index==1 -- like it was coded only for one dialog). What about if we supported x-canonical-private-rejection-tint-XXX where XXX is an action index. And if there isn't an XXX, we assume 1? [14:23] dednick, mterry: hmm no... let me come up with a simple patch [14:23] shouldn't be too hard [14:23] dednick, ltinkl: oh right, a hint can have a value [14:23] ltinkl, dednick: I like -index [14:24] then we can deprecate the "-tint" [14:25] mterry, dednick: https://pastebin.kde.org/p1x46wiug [14:25] this should work for both buttons [14:26] ah, small correction [14:27] while we're in there somebody should clean that button loader up... why are there 2 repeaters? [14:27] oh, a row and cloumn. [14:28] ltinkl, ah but we don't want both buttons red [14:28] yeah, a sec [14:28] ltinkl, dednick: so if we were to implement rection-index, is there anything we need to modify besides unity8's Notification.qml? LIke any spec or approval we need from someone? [14:29] i have no idea if there is a spec around somewhere [14:29] should be. there are loads of hints. [14:30] dednick, ltinkl: looks like we should expose it in the capabilities in unity-notification [14:31] mterry, dednick: yeah, if you add new hints, you should add them also to the unity-notifications backend [14:31] shouldn't be a big prob [14:31] dednick, ltinkl: OR.... we could keep going with the bad current hint and assume that if ONLY rejection or affirmative are specified, it should be for index==0... [14:31] That makes some sort of sense to me, without the bother of adding new hints [14:32] mterry, then it's break as soon as the other button should be colorized too :) [14:32] ltinkl, yes... but maybe by then we have proper dialogs :-/ [14:32] mterry, I wouldn't bet my socks on that ;) [14:33] ltinkl, well fair enough. We can do the right thing. So dednick, ltinkl: any objections to me adding rejection-index and affirmative-index? [14:33] mterry, if we can't come up with anything better, sure [14:38] ltinkl, do you mean to say you don't like the -index idea? [14:38] mterry, I mean I'm fine with it [14:39] mterry, can't see any other (better) way atn [14:39] atm [14:39] ltinkl, dednick: from a client point of view, do they name actions or index them? [14:39] like if I'm writing code for an app, and I add actions to a notification [14:39] mterry, it's an ordered list [14:39] Do I give it a name or just know the order? [14:40] ltinkl, yeah, but do I also name it? Like does any part of the api refer to indexes already or is it all names? [14:40] mterry, just the order; the format is sth like "actionId:Visible text" [14:40] mterry: it's in a model [14:40] mterry, let me verify [14:41] I think they name them... so having them specify name might be nicer [14:41] mterry, dednick: the model is a plain string list (with pairs), so every odd number is a label, every even number is the action id [14:41] like it's an ordered list of action names [14:42] sorry, the opposite [14:42] id+label [14:42] i guess maybe use action name rather than index then [14:42] ltinkl, dednick: yeah. So I'm thinking "x-canonical-private-affirmative-action=actionId" [14:42] mterry, dednick: see MockActionModel.cpp [14:42] x-canonical-private-affirmative-action=plop [14:43] mterry: yup [14:43] So this means you couldn't specify more than one positive / negative button... [14:43] mterry, dednick: yup, better identify them by the actionId [14:43] x-canonical-private-affirmative-actionId=true ? [14:43] that's also exported to the QML side [14:43] Or maybe that restriction is good...? [14:43] ltinkl: no, i think better not muddle with dynamic hint names [14:44] unless it comes from desigh [14:44] dednick, so you suggest x-canonical-private-affirmative-action=plop, where plop is the actionId or index? [14:44] ltinkl: ya [14:44] actionId [14:44] OK, I can go with that. Designs aren't likely to use multiple affirmatives... [14:44] dednick, right, that's what I meant :) [14:45] OK. x-canonical-private-affirmative-action=actionId [14:45] ltinkl: oh, it was directed at mterry; not you :) [14:45] sorry, you're both showing up as highlight color :) [14:46] dednick, mterry: ok then the Notification.qml code should be easy, lines 555-589 [14:46] dednick, ltinkl: we could do x-canonical-private-affirmative-actions=actionId1,actionId2,etc [14:46] could be a list yeah, more flexible [14:47] mterry: fyi; there's a could of spots using the tint i think [14:47] and you need to make sure the old way works still [14:47] yeah that's fine [14:47] we can support old way [14:48] OK, will prepare some branches. This is all in service of bug 1332306, btw [14:48] bug 1332306 in unity8 (Ubuntu) "Hot-swapped SIMs should be detected and prompt user to reboot" [High,Triaged] https://launchpad.net/bugs/1332306 [14:48] Which I inherited from Mirco! :) [14:49] :) [14:51] dednick, ltinkl: should we worry about affirmative-actions=... sounding a lot like https://en.wikipedia.org/wiki/Affirmative_action ? [14:51] lol. [14:51] I'm not worried, but still. Feels weird looking at it in source code [14:52] well i dont think it's trademarked or copywrite; so i think it'll be fine. === Guest77720 is now known as ahoneybun === dandrader is now known as dandrader|afk [15:27] dednick, mterry: regarding the most recent comment... why do we need this dialog in the first place? the SIM should be usable without a reboot [15:28] ltinkl: https://bugs.launchpad.net/ubuntu/+source/ofono/+bug/1332306 [15:28] Launchpad bug 1332306 in unity8 (Ubuntu) "Hot-swapped SIMs should be detected and prompt user to reboot" [High,Triaged] [15:28] ltinkl, I don't know the full technical details. But bug 1332306 gets into it a bit. I think only some platforms support it or something? [15:28] dednick, mterry: yeah, check the last comment [15:29] I wouldn't spend much time on the dialog now :) [15:29] ltinkl, ah cool. OK [15:29] thx :) [15:30] mterry, dednick: I mean, it must be possible without a reboot, some android phones do it (unlike the stock android) [15:34] some network operators require a reboot after picking up new data settings i think. [15:34] but meh. === dandrader|afk is now known as dandrader [17:47] ltinkl, on my dogfood phone, I'm seeing "Mute" in indicator-sound instead of "Silent Mode" -- is that expected? [18:33] bregma, is there an easy way to port config over to a libertine container? Like, my xchat-gnome config and such [18:33] bregma, oh I found the rootfs [18:34] yeah, thye're bind-mounted in [18:34] bregma, oh, what's bind-mounted in? [18:35] * mterry wants a way to have libertine just point at my rootfs [18:35] a 'fake' container [18:36] mterry, the root of each container is available in the host OS at ~/.local/share/libertine-container/user-data and bind-mounted into the container [18:36] bregma, got it [18:37] if you;re trying to break out of the container and have your host filesystem used as a 'fake' container you're going to have to work pretty hard [18:37] breaking containment isn't a supported option [18:37] bregma, just would make for easier dogfooding of u8 -- if all my apps didn't lose config. I know it's not in-scope for libertine. I actually just want a "native apps" scope [18:38] to hell with confinement for this use case [18:38] that would miss the whole point of containment, then, wouldn't it? [18:38] bregma, yes [18:38] bregma, but I'm actually only interested in dogfooding u8, not confinement right this second [18:38] what you want is to use the snap scope and just use snapcraft to repackage all your apps [18:39] it's just that simple [18:39] bregma, sure, the future is bright. :) But dogfooding on the way to that future is not [18:39] dogfooding Unity 8 means following the containment story [18:39] if you;re not using containment and Unity 8, you;re not dogfooding Unity 8 [18:40] (you can always hand-craft .desktop files if you really want to bypass containment, but there's no X11 server for you then) === dandrader is now known as dandrader|afk [18:56] josharenson, I tried to test chooser-gui on Friday, but the branch hadn't been updated? Did you not push to LP? [18:57] mterry: let me check [18:58] mterry: try pulling now? [18:59] josharenson, got updates! OK, will retest [18:59] mterry: cool, sorry about that [19:07] josharenson, also, consider updating from trunk? get nice focus fixes that way too [19:11] mterry: will do, gimmie a few min to finish lunch [19:11] josharenson, no rush :) === alan_g is now known as alan_g|EOD === dandrader|afk is now known as dandrader [21:56] dandrader, still around? https://code.launchpad.net/~dandrader/unity8/mirSurfaceInputBounds/+merge/298780 has a conflict? [21:58] mterry, let me check... [22:01] mterry, no. didn't find any conflict with trunk [22:02] dandrader, LP shows the conflict in debian/changelog [22:03] mterry, LP web diff can get confused. I tried it locally and it merges fine [22:03] dandrader, why the version bump anyway? [22:03] mterry, can you reproduce the conflict? [22:03] mterry, because of ubuntu-keyboard dependencies [22:03] iinm [22:04] dandrader, I can't reproduce the conflict. so good I guess