=== jzheng_afk is now known as jzheng === marcusto_ is now known as marcustomlinson [09:55] pstolowski, cimi, (how) can we test fixes for bug #1549056, bug #1424720 ? [09:55] bug 1549056 in unity8 (Ubuntu) "content sharing cannot work in scopes" [Undecided,In progress] https://launchpad.net/bugs/1549056 [09:55] bug 1424720 in unity8 (Ubuntu) "Add emblem to PreviewHeader widget" [Medium,In progress] https://launchpad.net/bugs/1424720 [09:55] first you can't really [09:55] you'd need a scope that uses those [09:55] which only the guy that reported the bug seems to have one [09:56] we could learn chinese too :) [09:56] second same-ish [09:56] but tryPreviewHeader will show the difference [09:56] on the phone itself i don't think there's scope using it either [09:58] note that the chinese guy said my branch works for him [09:58] i could quickly create a test scope which makes a picture 'sharable' [09:59] i already have a test scope project (Books scope) i use to test various new stuff ;) [10:00] pstolowski, that'd be great to share with QA indeed [10:01] Saviq, ok, will do it, is this needed right now or is later/Monday ok? [10:02] pstolowski, asap would be good, or I don't land filters for you next week ;P [10:02] pstolowski, it's for silo 64 testing ;) [10:02] Saviq, you can be very convincing ;) [10:08] cimi, do i need a local image or http:// image uri to test https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1549056 proper? [10:08] Launchpad bug 1549056 in unity8 (Ubuntu) "content sharing cannot work in scopes" [Undecided,In progress] [10:10] Saviq, do you know^ ? [10:10] pstolowski, both *should* work, no? [10:11] Saviq, sure, just not sure if you want to test a specific case [10:11] pstolowski, no, any would do [10:12] pstolowski, any should work [10:13] pstolowski, if is external, it will be treated as a URL iirc [10:13] * cimi doesn't fully know contenthub [10:13] cimi, Saviq yes, we have that other bug about that, some apps don't deal with it [10:25] Saviq: (proxy to mterry) the next UITK landing is already in testing phase in silo50 ( again :) ) [10:25] zsombi, yeah I pointed him at it [10:25] Saviq: ok, thx :) [10:29] cimi: if you have 10 mins can you test my last comment in https://code.launchpad.net/~josharenson/unity8/fix-over-eager-scope-swipe/+merge/288589 to check i'm not doing something horribly wrong? [10:30] tsdgeos, as I said in the mail I am not sure it's the right fix [10:33] cimi: i don't think it's the right fix either [10:34] but it's not that i can even see it being the console.log being printed [10:37] tsdgeos, oh [10:38] tsdgeos, it's in Shell.qml, you need to restart unity8 and grep for unity8.log [10:38] lol stupid [10:38] lol [10:38] :) [10:44] cimi: did i get that mail saying you don't think this is the right fix? [10:58] hi all - is there someone who could look at bug 1535058? It seems the unity 8 session has not been usable for a while, as apps crash upon start [10:58] bug 1535058 in ubuntu-app-launch (Ubuntu) "applications close instantly when launched from the launcher or dash" [High,In progress] https://launchpad.net/bugs/1535058 [11:10] mzanetti, status of that bug? [11:10] dpm, it works fine, just libstall libpam-cgm [11:10] Saviq, doesnt work for me [11:11] I can try again, maybe libpam received an update lately [11:11] libstall... [11:11] dpm, there's a branch even, you might wanna poke ted/alecu to land it finally [11:11] cimi, I think dednick fixed that... although I have seen another case of it happening. not sure if we have a fix for that too [11:11] cimi, it does work, some caveats still to get it going, but it does work [11:12] Saviq, it didnt work last time I followed your steps, and didnt work too in austin when I gave my laptop to ted [11:13] anyway, I will try again later [11:13] cimi, where's the bug that says the fixes we know of are not enough yet? [11:14] hmmm? [11:14] Saviq, well the bug still says is not fixed, [11:15] cimi, but it has a fix, and no one is working on it because it's assumed the fix fixes [11:15] Saviq, ok but "just install libpam-cgm" is not a fix [11:16] doesnt fix for some of us [11:16] Saviq, dpm: thanks, will remind ted to land it when he gets back next week [11:16] Oh, so something else is needed? [11:16] alecu, not that we know of [11:17] cimi, please try again, check whether your cgmanager.service is "active (running)" [11:17] that you have libpam-cgm installed [11:17] Saviq, yeah, if it doesnt I'll try with ted branch [11:17] cimi, ted's branch only adds a libpam-cgm dep [11:17] Saviq, last time I tried, cgmanager.service was "active (running)" but still apps were not running [11:17] ah ok [11:18] CRAP :) [11:18] trying later, need to work now [11:18] oh wait... that's the other one... [11:19] yeah, I think that was only some missing deps [11:29] Saviq, I had installedd libpam-cgm already [11:29] I tried several things in the comments, but none helped [11:30] Saviq, I was told to try with libpam-cgm and with libpam-cgfs (which I currently have installed). None of them fixed it [11:31] Saviq, as you're saying, the branch only adds the dep, but that does not seem to be the fix [11:32] dpm, you on xenial? can you check if cgmanager.service is running? I don't know of any other issues with running apps (browser was dying, but that's gone now) [11:33] Saviq, I am on xenial. Last time I tried it was running, but I can double-check [11:34] hm that does not look good? http://pastebin.ubuntu.com/15346777 [11:41] cimi, here is a scope you can use to test your fixes lp:~stolowski/+junk/scope-content_sharing (the binary is in click/ dir), can you give it a shot? i don't see emblem in the preview header though, perhaps i'm doing something wrong? [11:41] Saviq, ^ [11:41] dpm, nope it doesn't [11:41] pstolowski, let me read the code [11:41] dpm, sudo apt install cgmanager [11:42] pstolowski, you got silo 64 installed? [11:42] Saviq, I went through that a few weeks ago. The cmanager package does not exist [11:42] Saviq, yes [11:42] dpm, http://pastebin.ubuntu.com/15346801/ [11:42] * dpm reinstalls libpam-cgm instead of -cgfs [11:43] dpm, c*g*manager [11:43] aha! [11:43] thanks [11:43] hm... "cgmanager is already the newest version (0.39-2ubuntu5)." [11:44] http://pastebin.ubuntu.com/15346808/ [11:44] pstolowski, first of all, do you have the branch installed? [11:49] cimi, i've the silo 64 installed [11:49] pstolowski, also, is sc::Variant("thumbs-up") enough to get the icon? [11:50] what does sc::Variant() do? [11:50] I might miss sth [11:51] cimi, it's just a string efectively. i asked you recently about icons wrt to social-actions, and you told me it was enough to get the icon from the theme ;) [11:54] pstolowski, so the string should be "image://theme/thumbs-up" [11:54] cimi, ok... so maybe social-actions has the same issue and that's why i didn't see them? [11:55] try with sc::Variant("image://theme/thumbs-up") [11:55] mmm might be [11:55] I will try now locally [11:55] for socvial [11:58] pstolowski, can you please drop the changelog bump in lp:~stolowski/unity-api/filters-iface - we're doing that centrally now to avoid conflicts [11:58] Saviq, ok [12:00] pstolowski, I am not even sure we have thumbs up icon :) [12:00] cimi, we have [12:00] cimi, i hope it's mapped.. i did find svg on the phone [12:01] pstolowski, thumb-up [12:01] pstolowski, not thumbs [12:01] ouch [12:01] :) [12:01] dammit, just rebuild the click :/ [12:05] cimi, ok, it works, i've rebuilt the scope. the placement looks a little bit weird though [12:06] pstolowski, albert did that branch /me approved [12:06] cimi, anyway, please check your content sharing fixes and emblem in previews and you can give this scope to QA [12:06] https://code.launchpad.net/~aacid/unity8/emblemInPreviews/+merge/286038 [12:07] Saviq, pushed unity-api changes [12:21] pstolowski, tx === alan_g is now known as alan_g|lunch [13:10] hey albert, you busy? [13:23] cimi: me? [13:24] tsdgeos, ahah yeah :D [13:24] cimi: what do you need? [13:24] tsdgeos, I struggle in debugging what is wrong with pawel scope for social actions [13:24] tsdgeos, I am worried it might be a translation of social-actions to socialActions or something similar [13:24] I am confused [13:26] tsdgeos, can you help me with that? https://code.launchpad.net/~stolowski/+junk/scope-social_actions [13:26] cimi: so what do i need exactly? [13:26] tsdgeos, understand what is wrong :) [13:27] tsdgeos, http://paste.ubuntu.com/15347293/ [13:27] tsdgeos, this adds some debug on top of my branch [13:27] cimi, i need to add image://.. to that scope [13:27] ~cimi/unity8/card-social [13:27] pstolowski, is not only that [13:27] tsdgeos, we are not in a rush, but I am confused... [13:27] cimi, ah, ok [13:27] cimi: i mean which branches, scope,s etc [13:28] tsdgeos, lp:~cimi/unity8/card-social [13:28] cimi: or is there a silo? [13:28] tsdgeos, silo 71 [13:29] tsdgeos, plus lp:~stolowski/+junk/scope-social_actions to hack on the test scope, and lp:~cimi/unity8/card-social for unity8 branch [13:29] tsdgeos, I compile the scope, then I run with tools/unity-scope-tool ../deps/scope-social_actions/src/socialtest.pawelstolowski_socialtest.ini [13:29] (I have a deps folder inside my unity8 branch with his test scope) [13:30] cimi, i've just pushed updated icon names for that scope [13:31] pstolowski, it looks like the model I am getting, cardData, is empty [13:31] I get there is a social-actions defined (social-actions and not socialActions too) [13:32] tsdgeos, so first of all we dont seem to have components["socialActions"] but "social-actions" [13:33] tsdgeos, second, cardData seems empty (and I don't know how we can easily print what's inside that in JS) [13:33] JSON.stringify [13:33] is your friend [13:33] cardData["socialActions"] that is [13:34] cimi: well if socialActions haas nothing and social-actions has, that is the problem, no? [13:35] tsdgeos, one of the problems [13:35] tsdgeos, not sure is just that... [13:36] tsdgeos, it's one of the first scopes I touch that's why I am a bit confused and asked you if you can see what I am doing wrong... I might be restarting unity8-dash instead unity8, I might be doing something silly on a friday :)) [13:38] so questions are, is unity8 or the scopes wrong? should be socialActions or social-actions on our side? [13:43] cimi, it's consistent with quick-preview-data (which works), and the implementation is identical in the shell-plugin [13:48] pstolowski, looks like components["socialActions"] does not exist === dandrader is now known as dandrader|afk [13:53] cimi, re-flashing my phone to do some debug [13:53] pstolowski, i do it from the pc [13:54] cimi, components is something tou create right? because you'd need to query the socialActions role from the results model to get actions [13:54] i mean the 'components' dictionary [13:56] pstolowski, I think the scope defines it [13:57] cimi, yes, in the category json definition there is a components mapping. but that's something you don't access directly [13:57] I think we parse that [13:57] cimi, you access this stuff trough resultsmodel roles, such as quickPreviewData or socialActions [13:58] then we create a sample card [13:58] right you need to parse that [14:01] I dont think I understand totally how this works, I hope albert will clarify with a patch I can review :) [14:02] cimi, grepping for Albert's quickPreviewData, I only see it referencing cardData[...], not components [14:02] Saviq, https://bugs.launchpad.net/canonical-pocket-desktop/+bug/1543344/comments/7 [14:02] Launchpad bug 1543344 in unity8 (Ubuntu) "scroll wheel too fast" [Undecided,Triaged] [14:03] pstolowski, yeah can we see a scope using it? [14:03] cimi, sure [14:03] cimi, My Music [14:04] yeah [14:04] the template part is with - [14:04] ./Dash/CardCreator.js:457: var isAudio = template["quick-preview-type"] === "audio"; [14:04] so components is the same [14:05] the camelCasedOne is because of [14:05] ./unity/shell/scopes/ResultsModelInterface.h:88: roles[RoleQuickPreviewData] = "quickPreviewData"; [14:05] so one thing is accessing the card data [14:06] the other is accessing the json [14:06] do i make sense? [14:06] mmm yeah? [14:06] tsdgeos, so we need to have components[social-actions] in cardcreator AND cardtool? [14:06] if you access components yes [14:06] it's always "social-actions" [14:07] if you access cardData it depends on the mappoing that ResultsModelInterface has [14:07] if it has any [14:07] anyway /me lunch [14:08] tsdgeos, if it takes you 5 mins to write a working patch, I understand better, otherwise I will try myself later no problem [14:08] components mapping from category renderer is parsed by shell plugin, so shell plugin knows what to remap if neccessary. i don't think unity8 is interested in this [14:08] quick-preview-type on the other hand is part of the template definition, that's why unity8 reads it [14:13] i think it's overcomplex === alan_g|lunch is now known as alan_g [14:19] ltinkl, what's the new u8 keyboard branch that needed a review? [14:20] mterry, https://code.launchpad.net/~unity-team/unity8/keymapSwitching/+merge/288764 [14:23] ltinkl, I thought switching was already in your other branches [14:24] mterry, yeah... but it was dependent on the indicator, this one is not [14:24] ah ok [14:25] ltinkl, is silo 41 useful for testing this branch? [14:27] mterry, yeah should be (if it builds) [14:28] ltinkl, doesn't build now anyway (some weirdness with its clicks [14:28] yeah [14:28] ltinkl, I don't want to edit that silo :) [14:28] ltinkl, should I just manually build then? I need silo 64 + the three kbdLayout branches? [14:29] * mterry really wishes 64 would land :) [14:29] mterry, you basically need https://code.launchpad.net/~unity-team/unity8/keymapSwitching/+merge/288764 and the branches listed in description [14:30] mterry, oh and probably the unity-api stuff in: https://code.launchpad.net/~unity-team/unity-api/packaging/+merge/286145 [14:30] ltinkl, OK, but keymapSwitching is based off of silo 64's u8 branch. Which probably depends on everything else in silo 64 [14:30] mterry, yeah... === dandrader|afk is now known as dandrader === sdk_ is now known as sdk [14:52] ltinkl, why keep track of a separate activeKeymapIndex variable in AbstractStage.qml? [14:56] mterry, to get the keymap index of the currently focused window, and to be able to switch the keymap at a central place (not depending on the loaded stage) [14:57] ltinkl, but why do we need the keymap index of the currently focused window? We overwrite it when it becomes focused with currentKeymapIndex [14:57] mterry, to calculate the next/previous index [14:58] ltinkl, but why not base that off of currentKeymapIndex? [14:58] ltinkl, I understand the argument about switching stages though... But the current code doesn't look like it pulls from window index ever (unless user presses meta+space, but by that time the window index will have been overridden) [15:00] mterry, it does here: keymaps.indexOf(mainAppWindow.activeKeymap) [15:00] ltinkl, right when setting activeKeymapIndex [15:00] ltinkl, but we don't use activeKeymapIndex except when user presses meta+space [15:00] ltinkl, and when a window becomes focused, we override its keymap with the currentKeymapIndex [15:01] So it seems to me like we can do without activeKeymapIndex [15:01] * ltinkl thinks again [15:01] ltinkl, although I think you're right that we *want* something like it to save keymap across stage reloads [15:02] mterry, yes that and... the main problem here really is, there's no method to _get_ the keymap of a surface [15:02] mterry, Mir has only a setter [15:03] ltinkl, so maybe we should save the index outside of the stage [15:03] ltinkl, added some inline comments in the meantime [15:04] mterry, we save the keymap (not the index tho) in the surface, so that switching stages should (and does) indeed work [15:04] mterry, some of that logic might change when we get surface based WM [15:05] ltinkl, but when switching to a new stage, currentKeymapIndex never gets updated. It's at 0. So because when we switch windows, we use currentKeymapIndex, it won't be saved properly [15:06] mterry, current gets updated by the shortcuts [15:06] ltinkl, and onMainAppWindowChanged might even fire right after stage construction? not sure [15:06] mterry, initially it's at 0 [15:06] ltinkl, right. But let's say I switch stages. Then I switch windows (before pressing a shortcut) [15:06] ltinkl, or if onMainAppWindowChanged fires during stage construction (possible, haven't tested), then it immediately gets overriden [15:06] mterry, 0 meaning the user's first choice [15:07] ltinkl, right right. But let's say I'm in a desktop stage. I've toggled to index 3. [15:07] ltinkl, I unplug my phone, now I'm in phone stage [15:07] ltinkl, currentKeymapIndex is now 0 [15:07] ltinkl, and if I switch windows, the new window will get keymap 0 [15:08] ltinkl, you might be able to fix this with a simple default binding of "currentKeymapIndex: activeKeymapIndex" that gets overridden when using meta+space shortcuts [15:09] mterry, yeah ok, I'll test this shortly, after I get it to build again :) [15:11] pstolowski: is that silo usable at all? [15:11] isn't giving me unity8 from the silo [15:11] but from the stable ppa [15:11] how old is it? [15:11] tsdgeos, needs rebuilding due to unity8 landings [15:12] just force the version from silo for now [15:12] tsdgeos, you're talking about silo 71 right? [15:13] yes [15:13] mzanetti: question what is the name of the element in notes where you actually create the note and can it be added to the copy paste system in unity8 so you can paste stuff into a note? [15:13] rebuilding now.. i didn't rebuilt since we knew something is still broken in it [15:15] yeah but it's kind of annoying to try to find what's broken when just getting to tes tit is hard [15:15] because you don't get the things [15:15] davmor2, I've implemented copy/paste for notes half a year ago... but for various reasons that release never passed QA so far [15:15] i've lost like 45 min trying to help you guys and i've still have nothing [15:15] davmor2, so yea... build trunk and you'll have copy/paste support [15:15] since i can't get the things to even run against eachother [15:15] tsdgeos, fair enpugh [15:16] mzanetti: oh so it was an app issue not copy/paste issue then [15:16] davmor2, well, it wasn't possible with uitk < 1.3 [15:17] ah okay [15:17] mzanetti: awesome will look forward to that it will make it easier to test copy paste from the browser then :) [15:19] davmor2, well... I kinda lost motivation on it tbh... so not sure what's going to happen with it [15:19] pstolowski: the click is outdated too [15:19] * davmor2 hugs mzanetti till he likes it again [15:20] it kept on bouncing back from QA for really stupid reasons. after 3 months of trying I gave up [15:20] tsdgeos, my scope? [15:20] yes [15:20] mzanetti: :( [15:20] pstolowski: http://paste.ubuntu.com/15348018/ [15:20] no social-cards [15:20] in the compiled .so [15:21] even the code has them [15:21] ssocial-actions i mean [15:21] -s [15:21] that also made me lose quite a bit of time [15:22] mterry, addressed your comments, will clean up the debug once I've verified it still works :) [15:23] tsdgeos, sorry... updating [15:27] tsdgeos, I compile both the scope and unity8 locally, then I run the scope with unity-scope-tool [15:28] cimi: you need the scope shell code too otherwise it won't work [15:28] ltinkl, you might want to add a qmltest that switches stages, for that code path we talked about [15:28] tsdgeos, yeah... I installed them too [15:29] (I think) [15:29] * cimi checks [15:30] mterry, yup, will likely move the test from desktop stage to shell [15:31] tsdgeos, updated click pkg in lp:~stolowski/+junk/scope-social_actions [15:47] cimi: so i can give you some hints [15:48] but this needs really fixes to make it work [15:48] tsdgeos, tx [15:48] since you're using ListItems that you can't really use there === dandrader is now known as dandrader|lunch [16:06] tsdgeos, I'm all ears [16:06] or eyes [16:06] :) [16:06] cimi: it's on the MR [16:07] tsdgeos, oh great! some of the things were already in the patch I shown you [16:07] maybe [16:07] tsdgeos, some are what I guessed but didn't want to :) [16:07] i didn't check the patch since you said it was only to add debugging [16:08] the fields in CardTool [16:08] I thought we needed two indeed [16:08] let me try but thanks for now [16:08] tsdgeos, scope works fine with this? [16:09] cimi: no it doesn't since as said the qml component can't be created [16:09] because of the use of ThinDivider in CardSocialActions [16:09] i guess i can try removing that [16:10] then it complains [16:10] image:://theme/thumb-down [16:11] pstolowski: that is on you ↑ ? [16:11] can you fix it? [16:11] uh, yeah [16:14] Mirv: ping [16:23] pstolowski, cimi, where are we with the emblem/sharing? got a scope for me? :) [16:23] Saviq, we do have scopes yes :) [16:24] lp:~stolowski/+junk/scope-content_sharing [16:24] need to test the sharing too [16:26] ltinkl, ~unity-team/unity-api/kbdLayout doesn't actually bump its version number to 108, despite ~unity-team/qtmir/kbdLayout needing that version [16:26] mterry, yup, that's in https://code.launchpad.net/~unity-team/unity-api/packaging/+merge/286145 [16:26] mterry, part of the silo [16:27] ltinkl, guh, ok... I was going off the list in https://code.launchpad.net/~unity-team/unity8/keymapSwitching/+merge/288764 [16:27] ltinkl, anything else? [16:27] mterry, nope, don't think so [16:28] Saviq, I use your frankenstein silo? [16:29] cimi, not ready yet [16:29] Saviq, oki [16:29] cimi, oh wait what for? [16:29] Saviq, content sharing [16:29] cimi, not frankenstein, under QA already :P https://requests.ci-train.ubuntu.com/#/ticket/993 [16:29] there's a new frankenstein https://requests.ci-train.ubuntu.com/#/ticket/1105 [16:30] so many nice gifs in telegram [16:30] we need to switch :) [16:30] and don't you tell me it doesn't work :P [16:30] I wanted a scope to show that it works, not one to show that it's broken :P [16:32] Saviq, so let me try :) [16:32] pstolowski, did you try the sharing scope with silo 64? [16:33] cimi, no, sorry [16:40] Saviq, on the desktop, the content hub appears [16:40] Saviq, however I dont have handlers :) [16:40] Saviq, it shows though blue colors everywhere, looks like is broken too with uitk 1.3 [16:42] cimi, yeah, there's a bug [17:37] josharenson, so the greeter branches that need reviewing are all in that silo? [17:43] mterry, yes === alan_g is now known as alan_g|EOW === dandrader is now known as dandrader|afk === dandrader|afk is now known as dandrader === JanC_ is now known as JanC