[08:35] Saviq, hey, do you know what's the status of https://bugs.launchpad.net/canonical-devices-system-image/+bug/1393438? is the fix on that bug supposed to address the issue? where is the translatable string in that case? [08:35] Launchpad bug 1393438 in unity8 (Ubuntu) ""Enable location data" string displays untranslated" [Undecided,Confirmed] [08:36] seb128, digging... [08:37] seb128, right, so it's missing some work on our (unity8) side [08:37] Saviq, k, that explains, thanks [08:38] Saviq, any chance that get assigned to somebody? the string is quite visible in the ui [08:38] seb128, just did [08:38] thanks [09:21] Saviq, sorry to have reported that dialer bug against unity8, I was unsure what component to pick, it's a bit of an app management issue ... could be the job from whoever started it to close it, or could be up to the dialer to exit after call [09:21] seb128, no worries, I left a unity8 task just in case [09:22] as it's indeed not clear dialer can do all it needs to do [11:04] tsdgeos, testing the temp scope to temp scope is a bit tricky: the only scope calling activate is the click scope, but the click scope cant be made unfavourite [11:06] Cimi: so what's the real world scenario we are fixing? [11:07] it's store as temp, and then a preview in store calling another temp, no? [11:10] Cimi, the store scope calls activate too [11:10] if that's what we're having [11:10] Cimi, you go to store, install a scope, press Search [11:10] we should try to replicate something similar [11:11] it doesn't need to be "the store" in our mocks [11:11] but something similar to what we're fixing [11:11] so the test is at least the real world case [11:11] we're fixing temp scope from tempo scope [11:12] Cimi: via what? preview or just activation of a result? [11:13] tsdgeos, for now preview, but I guess we need to see other possible scenarios [11:13] sure [11:13] in the test I am manually calling activate [11:13] yeah [11:13] from a temp scope [11:13] don't do that [11:13] let's please make a test that does the same the user does [11:14] but how do I create a ubuntu store scope with previews? [11:14] it doesn't need to be a ubuntu scope [11:14] it can be any of the mock scopes [11:14] and they have previews already [11:15] you just need to make one of those previews be special and do the same the store scope does to us when pressed [11:15] yeah, don't know what it does [11:16] have to dig [11:19] where is the code of the store? [11:19] you don't need the code for the store [11:20] just add some console.log in the preview code [11:20] and see how what it does [11:21] triggeredAction looks like a good place to add a console log [11:21] OK [11:22] /food [11:45] tsdgeos, it might be that is the preview model launching actions? [11:47] I see in Preview.qml that the buttons call previewModel.triggered() [11:47] (with action open_click [object Object] with the latter probably a scope) [11:50] Cimi: and how does that come back to us? in gotoScope/openScope of Dash.qml/DashContent.qml ? [11:52] tsdgeos, don't know [11:54] Cimi: console.log! === MacSlow is now known as MacSlow|lunch [12:00] tsdgeos, the action is called by the backend [12:01] of course [12:01] i'm just trying for you to understand the call path [12:01] so you can make our mock do something similar === alan_g is now known as alan_g|lunch === MacSlow|lunch is now known as MacSlow === alan_g|lunch is now known as alan_g [14:20] mterry: hey [14:20] mzanetti, hello! [14:20] mterry: I'm trying to fix the as launcher model [14:20] mterry: can I know if the session is running? [14:20] in order to ignore the non pinned ones [14:23] mzanetti, umm.. I just checked and I actually can't see that in the greeter API that LightDM exposes... [14:23] mzanetti, LightDM surely knows [14:23] mzanetti, but... oh wait. [14:23] mzanetti, no the greeter does know [14:24] mzanetti, so the greeter can know via liblightdm [14:24] mzanetti, but your branch doesn't care about that side of things, right? [14:24] mzanetti, your branch doesn't actually *use* the AS launcher model? [14:24] mterry: not really [14:25] mterry: it's just preparing it for you, so you can take it into use without having to mess with the launcher too much [14:25] mterry: I've added the pinned information to the sync process now [14:27] mterry: so what do you propose? [14:28] mzanetti, let me see if you can do anything with that information, given the current state of the LightDM plugin... hold on [14:29] I could add a "hideUnpinned" property to the launcher api. then when you use that in the split greeter, set that to true somehow when the session is not running [14:29] mzanetti, so it is exported via the "loggedIn" data property on the LightDM.UsersModel object [14:30] mzanetti, so that should be simple to hook up to your hideUnpinned property, yeah [14:30] ack === marcusto_ is now known as marcustomlinson [14:51] Anyone have an opinion on QQmlListProperty versus QAbstractListModel? I preferred QQmlListProperty for simple lists of objects, but now I think about it, it just has a "list changed" signal - it doesn't inform the view more detail about what changed like QAbstractItemList does (dataChanged, itemAdded/Removed) [14:52] greyback: yeah... QQmlListItem is pretty useless unless you want to create/change the list from qml [14:53] err. QQmlListProperty [14:54] mzanetti: so I'm seeing. But I never really liked creating a QAbstractListModel with a single roleName either, since the role is kinda pointless [14:54] yeah... [14:54] I for one so far always ended up converting my QQmlListProperties into real models at some point [14:56] * greyback takes note [15:15] Saviq: Cimi: how do you guys see https://code.launchpad.net/~aacid/unity8/properVRangesCurrentScope/+merge/246465 ? [15:16] tsdgeos, "see" in what way? it's a Needs-review MP? [15:16] Saviq: yeah [15:16] and i need someone not me [15:16] to check if we actually want that :D [15:16] or not [15:16] ok, I'll check it out [15:17] Saviq, thx [15:18] Cimi, hey! can we land https://code.launchpad.net/~stolowski/unity-scopes-shell/fix-temp-scopes/+merge/248235 ? did you have a chance to test it with your fix? [15:19] tsdgeos, today I noticed that when we open or switch to a scope, we load all results then we do collapse [15:20] pstolowski, seems to work, I am working on unity side [15:20] Cimi: what you mean load all results and then collapse? [15:21] tsdgeos, try with the kickstarter scope from the store [15:21] Cimi, ok, cool [15:21] tsdgeos, you see the categories start populating, then it collapses [15:21] it starts expanded then collapse [15:22] maybe when we load scopes we should load just collapsed results at first, then load the rest later? [15:23] Seems mumble dies again? [15:24] Cimi: hmmm we load everything collapsed [15:24] or we should [15:24] Cimi: can you record a video? [15:25] paulliu, yeah, it looks like [15:26] tsdgeos, ok [15:46] tsdgeos, https://www.youtube.com/watch?v=RTG6xPYmR48 [15:48] also, this "hide manage dash and scroll right to the scope" isn't the best, but it is more a design thing [15:49] Cimi, I assigned you to the bug today, the one about "Enable location" being untranslated in our UI, we need to add a special setting type that will have the strings to translate hardcoded [15:51] Saviq, saw it [15:51] Saviq, I am still trying to do this test for the temp scope fix [15:51] Cimi, yeah sure, just wanted to let you know the desired approach [15:51] Cimi, this looks like we're leaving the first category expanded until there's more categories, which is per design, but I agree the visual glitches are not nice [15:51] Cimi: hmmm, that's weird, we default to non expanded for all categories, does this only happen on that scope or all? [15:52] tsdgeos, don't we have code that expands if there's only one category? [15:52] tsdgeos, noticed only on that scope [15:52] tsdgeos, Cimi, it might be that the scope registers the categories as they come in [15:53] it's tedg fault :P [15:53] instead of early in the reply [15:53] Saviq: hmmmm, can't find anything like that in the code tbh [15:54] Cimi, Me? Temperature? It's nice here, not cloudy London. [15:54] tsdgeos, GSV.qml:236 [15:54] Cimi, Do you even remember what the sun is like anymore? ;-) [15:55] tsdgeos, and then GSV.qml:307-308 [15:55] tsdgeos, one category == not expandable; == expanded [15:55] Saviq: but it has to match expandedCategoryId [15:56] expandedCategoryId is only set when clicking on see all [15:56] tedg, morning. [15:57] tsdgeos, not really, height is selected based on bool expanded, which is always true if bool expandable is false [15:57] tedg, I forgot :( [15:57] Cimi, You need to come visit! [15:57] ah right right [15:58] Saviq: not expandable means not collapsable either [15:58] tsdgeos, yup, it's just full-height always [15:58] tsdgeos, which means that if we're just getting one category, all results, then another category [15:58] yeah i think the code expects the categories not to change on runtime [15:58] tsdgeos, we load them all just to collapse them again [15:58] yeah [15:59] yeah, the scope API I'm afraid allows that, if not encourages [16:01] tedg, will come to see you and evan... [16:04] tedg: Saviq: so on https://code.launchpad.net/~ted/unity8/delay-indicator-start/+merge/241124 ? discard? [16:05] I'm still +1 on it. I think it makes more sense. Doesn't fix that bug, but it should help. [16:06] No reason to make Unity compete for resources. [16:06] mzanetti: any idea why daniel didn't top approve https://code.launchpad.net/~mzanetti/unity8/launcher-fixes/+merge/248761 ? [16:07] tsdgeos: he probably just forgot... was sitting next to him when he told me he's ok with it [16:07] is he on holidays this week? [16:08] don't think so... I guess having a swap day for travelling across the sea [16:08] ok, let's wait for him tomorrow then [16:10] mzanetti: can you do https://code.launchpad.net/~dandrader/unity8/improveLauncherTests/+merge/248919 alternatively? [16:10] tsdgeos: yep, on my list [16:10] lemme claim it [16:10] cool [16:16] MacSlow: so where are we on https://code.launchpad.net/~macslow/unity8/swipe-dismiss-snap-decisions/+merge/233347 can you reproduce the errors Saviq mentions? [16:16] tsdgeos, that seems fixed already, but he's got some other failures [16:16] i see [16:16] actually a hash sum mismatch for qmluitests, /me rebuilds [16:17] MacSlow, the ap test for lifecycle is sometimes flaky, don't worry about that [16:17] Saviq, I assumed that after a few tries on my machines here [16:18] MacSlow, just start with kicking a jenkins rebuild next time [17:25] mterry: ok. pushed the branch [17:26] mterry: didn't change the appid/url thing for the reason of consistency. We use appId all over to identify items... not sure I want to do something else here [17:27] mzanetti, OK makes sense. Like I said, I was just half-remembering a conversation about that [17:27] mzanetti, will review again, thanks! [17:37] mterry: ah, I agree that we should collect split greeter based plugins in the Greeter dir. also that it should be another MP :D [17:38] mzanetti, yeah it's grown beyond a thin layer on top of lightdm [17:39] indeed === alan_g is now known as alan_g|EOD