[08:35] <Saviq> mzanetti, that's a duplicate - bug #1243889 - right? if you can remember the other one - please mark dupe
[08:36] <mzanetti> Saviq: yes.
[08:37] <mzanetti> narf... I *always* click on the bug number and join weird channels
[08:40] <Saviq> lol
[09:12] <tsdgeos> hmmmm
[09:12] <tsdgeos> does anybody know what causes the "up" animation in the preview?
[09:12] <tsdgeos> is it OpenEffect?
[09:12] <tsdgeos> Saviq: ↑↑↑ ?
[09:13] <Saviq> tsdgeos, "causes"?
[09:13] <Saviq> tsdgeos, OpenEffect splits the dash view, yes
[09:13] <tsdgeos> that's unfortunate :/
[09:13] <tsdgeos> doesn't take my new floating tabbar into account
[09:13] <Saviq> tsdgeos, right...
[09:14] <Saviq> tsdgeos, arguably, the preview could be moved up the stack
[09:14] <Saviq> tsdgeos, so that it's on top of the whole dash
[09:14] <tsdgeos> or at least the OpenEffect
[09:14] <Saviq> but then that'd mean we're splitting all of the dash
[09:14] <Saviq> not just the current view
[09:14] <Saviq> which means moar pixels :/
[09:18] <tsdgeos> Saviq: ?
[09:19] <tsdgeos> isn't "all of the dash" == "the current view"?
[09:19] <tsdgeos> ah no
[09:19] <Saviq> tsdgeos, no, I mean the other scopes, too
[09:19] <tsdgeos> we load everything
[09:19] <tsdgeos>         cacheBuffer: 2147483647
[09:19] <tsdgeos> but..
[09:19] <tsdgeos> the OpenEffect works over sourceItem.width, sourceItem.height
[09:19] <tsdgeos> not sure we'd be moving more pixels than that
[09:20] <Saviq> true
[09:20] <Saviq> if the effect is only as big as the view - it won't sample beyond that
[09:20] <Saviq> or affect anything beyond that
[09:20] <Saviq> tsdgeos, it feels like you should rebase on top of mzanetti's switching-previews before you start that
[09:21] <Saviq> which I should re-review btw :/
[09:21] <tsdgeos> he he
[09:21] <Saviq> mzanetti, btw https://bugs.launchpad.net/unity8/+bug/1243824/comments/2
[09:21] <mzanetti> tsdgeos: lp:~unity-team/unity8/switching-previews
[09:21] <tsdgeos> mzanetti: Saviq: i think i can "easily" fix it how it's now, let me get to almost-feature-finished
[09:21] <tsdgeos> and then i'll rebase
[09:22] <Saviq> tsdgeos, ok
[09:22] <mzanetti> Saviq: mhm. yeah. that matches with my suspicion
[09:24] <Saviq> mzanetti, the other movement in jhodapp's video (movement of the app itself) I'm not sure what it is
[09:24] <Saviq> mzanetti, it might be the screenshot, which doesn't cater for fullscreen apps...
[09:24] <mzanetti> Saviq: I think that's the panel
[09:25] <mzanetti> Saviq: ah no. the app's size seems to change
[09:25] <mzanetti> but that might well be related to the panel show/hide
[09:26] <Saviq> mzanetti, don't think so, as it doesn't happen in portrait
[09:26] <mzanetti> hmm... interesting
[09:26] <Saviq> I mean
[09:26] <Saviq> in portrait when you touch the app-panel edge
[09:26] <Saviq> but does happen on right edge
[09:27] <Saviq> and in gallery, too - but only if it's the only app
[09:27] <mzanetti> Saviq: because it's the right edge only
[09:27] <mzanetti> would be interesting to see if happens the other way round too. with the launcher's edge
[09:27] <Saviq> mzanetti, yeah but I mean it's unrelated to the panel
[09:27] <mzanetti> ok, yes
[09:28] <mzanetti> wait... I think we're confusing panels :D
[09:28] <Saviq> mzanetti, toolbar
[09:28] <Saviq> mzanetti, unrelated to toolbar
[09:28] <mzanetti> right, yes. agreed
[09:28] <Saviq> mzanetti, it's our hinting thingy for when there's only one app
[09:28] <mzanetti> oh!
[09:28] <mzanetti> yeah, sure
[09:29] <Saviq> mzanetti, so you'll fix it with your right edge rework ;)
[09:29] <mzanetti> Saviq: ok
[09:29] <Saviq> mzanetti, was kinda kidding, but yeah - it may very well be the output of your stuff
[09:30] <mzanetti> Saviq: btw. I managed to catch another one while testing with edges yesterday: https://code.launchpad.net/~mzanetti/unity8/dont-tease-while-moving/+merge/192366
[09:30] <Saviq> mzanetti, yeah, saw that
[09:31] <Cimi> fixed? https://bugs.launchpad.net/unity8/+bug/1235190
[09:43] <Saviq> Cimi, invalid - not our bug
[09:43] <Saviq> Cimi, btw, can you dupe bug #1234921 to the one you were fixing, please
[09:47] <Saviq> mzanetti, how about bug #1232260 - seen it anywhere again?
[09:48] <mzanetti> Saviq: hmm... Not knowingly. I did have some crashes lately, but didn't eagerly ready each and every stack trace any more in the last week
[09:48] <mzanetti> I'll watch out for it
[09:50] <Saviq> mzanetti, I'll make incomplete then
[09:51] <mzanetti> ok
[09:57] <mzanetti> Saviq: I'm afraid the right edge thing will require to rethink all the stages stuff
[09:57] <mzanetti> from both perspectives: Design and implementation
[09:59] <Saviq> mzanetti, of course it will
[09:59] <mzanetti> ok
[10:06] <Saviq> mzanetti, there's no bug for the right edge prototype you're doing now is there?
[10:06] <Saviq> bug #1228733 fyi
[10:09] <mzanetti> Saviq: cheers
[10:10] <Saviq> mzanetti, you got the gdbus call that you used to reset the launcher?
[10:10]  * Saviq needs to turn the demo back on
[10:10] <mzanetti> gdbus call --system --dest org.freedesktop.Accounts --object-path /org/freedesktop/Accounts/User32011 --method org.freedesktop.DBus.Properties.Set com.canonical.unity.AccountsService launcher-items "<[{'defaults' : <true>}]>"
[10:11] <mzanetti> Saviq: ^
[10:11] <Saviq> mzanetti, thanks
[10:11] <Saviq> jeez can we please fix the console over adb? kthxbai
[10:13] <mzanetti> +1 :D
[10:20] <Saviq> greyback, is there a bug about parenting surfaces in mir/unity-mir? bug #1240611 could use a relation there
[10:21] <greyback> Saviq: https://bugs.launchpad.net/unity-mir/+bug/1230091
[10:21] <Saviq> greyback, thanks
[10:28] <Saviq> yay for bug triaging...
[10:28] <Saviq> /food
[10:30] <Saviq> mzanetti, hmm so you weren't shown all the video prototypes?
[10:30] <mzanetti> Saviq: if you ask like this the answer is probably: no
[10:31] <Saviq> mzanetti, for main + side stage, right-edge, the most probable solution would be to show the side and main stage apps next to each other, on separate stacks
[10:31] <Saviq> mzanetti, obviously this does conflict with dragging the side stage in
[10:32] <Saviq> mzanetti, and I don't think they have a ready-made answer
[10:32] <mzanetti> ok. so the mail is still valid
[10:32] <Saviq> mzanetti, yeah of course it is
[10:32] <Saviq> mzanetti, please bug them so that they show you everything they have
[10:33] <Saviq> mzanetti, and at least nudge in the direction they're leaning towards
[10:34] <mzanetti> ok
[10:36] <greyback> Saviq: mzanetti: can you send me link to those video prototypes please? I've lost it
[10:36] <Saviq> greyback, we never had them ;)
[10:36] <Saviq> greyback, Oren showed it from his laptop - he'll probably steal them now and not let us see! ;)
[10:36] <mzanetti> greyback: Saviq: there is a qml prototype in the willow-team repository
[10:37] <Saviq> mzanetti, yeah, that's old
[10:37] <greyback> Saviq: he didn't share??
[10:37] <Saviq> somewhat
[10:37] <Saviq> greyback, AFAIR he shared the screen
[10:37] <greyback> pah
[10:37] <mzanetti> it's quite close to the video that vesar showed me through the hangout
[10:37] <Saviq> greyback, so that he could control
[10:37] <Saviq> mzanetti, yeah, problem is they're looking at some other similar approaches
[10:37] <Saviq> mzanetti, and you should see them all
[10:38] <mzanetti> true
[10:38] <Saviq> especially when we're not yet sure which way to go
[10:38] <mzanetti> vesar: ^^ *hint*
[10:38] <Saviq> if only for you to think of it in a way that would let us go between the different solutions a little bit easier
[10:38] <greyback> hmm, I've got the phone video
[10:38] <greyback> might be old tho
[10:39] <mzanetti> I guess that's the one I saw in the meeting yesterday
[10:39] <Saviq> greyback, yeah, it probably is the first one - and Oren showed us like 5 different ones then
[10:39]  * Saviq really food
[10:41]  * greyback away
[10:50] <nic-doffay> Saviq, how can I slow down the desktop gallery previews again? I recall you telling me this a while ago and forgot to make a note of it...
[10:58] <Saviq> nic-doffay, qmlscene --help
[12:12] <dednick> mpt: ping
[12:51] <sil2100> Trevinho: ping!
[12:51] <sil2100> :)
[12:53] <sil2100> bregma: hi! Do you know when Trevinho will be around today?
[12:54] <bregma> sil2100, he's in California, he might be around Monday
[12:54] <sil2100> Ouch
[12:54] <bregma> evidently we allowed him a day or two of vacation
[12:54] <bregma> won;t happen again
[12:55] <bregma> :)
[12:55] <mzanetti> dednick: I got assigned to do a reevaluation of things that need to be tested etc. what happened to the indicator autopilot tests?
[12:56] <Saviq> mzanetti, there weren't any ;)
[12:56] <Saviq> mzanetti, but alesage is working on new ones
[12:56] <Saviq> mzanetti, https://code.launchpad.net/~allanlesage/unity8/indicator-stubs/+merge/192059
[12:59] <mzanetti> ok, thanks
[13:03] <Saviq> dednick, add a TODO on https://code.launchpad.net/~nick-dedekind/unity8/indicator-segfault-lp1243146/+merge/192121 to mention there *will* be a QTBUG please
[13:04] <Saviq> /we need to do another round of TODO / FIXME review
[13:12] <greyback> Saviq: to add to your list of TODOs, it would be great if we could get cross-building with pbuilder chroots working reliably.
[13:12] <Saviq> greyback, oh yeah, that's there already
[13:12] <Saviq> greyback, btw, just found a potentially nice trick
[13:12] <greyback> Saviq: oO, please share :)
[13:12] <Saviq> greyback, system-image-cli --filter full
[13:12] <Saviq> greyback, will revert your write-enabled phone
[13:13] <Saviq> to the latest pristine image
[13:13] <Saviq> without touching any other data
[13:13] <greyback> Saviq: cool
[13:13] <Saviq> with --build 0 will replace the current image
[13:14] <Saviq> no way yet to downgrade - bug #1244208 and bug #1244211
[13:15] <Saviq> but that at least means we can easily revert to a clean image without losing $HOME and such
[13:15] <Saviq> which means I could start dog-fooding the phone after all
[13:29] <dednick> Saviq: ok
[13:32] <dednick> Saviq: done.
[13:35] <greyback> tsdgeos: updated my 5.2 build, now unity8 builds correctly
[13:35] <greyback> but doesn't run :)
[13:35] <tsdgeos> he he
[13:42] <mhr3> Saviq, btw were you able to get stacktrace from the carousel crash you mentioned?
[13:44] <mterry> Saviq, looks like I missed the meeting, sorry.  Had to run to a coffee shop because my internet was down
[13:44] <mterry> Saviq, will add notes myself
[13:45] <Saviq> mterry, sure, nw
[13:59] <Cimi> Saviq, cannot reproduce this https://bugs.launchpad.net/unity8/+bug/1234108
[14:00] <Saviq> Cimi, https://bugs.launchpad.net/unity8/+bug/1234108/comments/4
[14:00] <Cimi> Saviq, I'm on sf
[14:00] <Saviq> Cimi, fine, then - mark incomplete if you can't reproduce, although it'd be good to find out why
[14:01] <Cimi> Saviq, I'll keep debugging
[14:01] <Cimi> Saviq, one thing is sure
[14:01] <Cimi> Saviq, if I swipe immediately when the app starts
[14:01] <Cimi> Saviq, I go back to dash
[14:01] <Cimi> Saviq, but when the app finishes loading
[14:01] <Cimi> Saviq, it switches to the app
[14:02] <Saviq> Cimi, different bug
[14:02] <Saviq> Cimi, it's filed already - and happens on Mir, too
[14:03] <Cimi> Saviq, can be fixed on the connection with the application manager
[14:04] <Cimi> Saviq, that might mean that applicationManager.mainStageFocusedApplication is true
[14:05] <Cimi> while should be false when you swipe back to dash
[14:10] <Saviq> mzanetti, https://code.launchpad.net/~unity-team/unity8/fix-autopilot-touch can be removed, right?
[14:11] <mzanetti> yeah
[14:11] <Saviq> Cimi, it's not that, the app gets focused when it's ready - it switches to focused, but it shouldn't, if you swiped away
[14:11] <Cimi> Saviq, that's what I mean
[14:12] <Cimi> Saviq, it should be false
[14:13] <Saviq> Cimi, it's about it being null, for that matter, but even so - it *is* null when you swipe to dash, but then the app manager signals that the app wants to get the focus - and that's when we're still obeying, even though we should not
[14:13] <Saviq> Cimi, we knew that already ;)
[14:14] <Saviq> greyback, how about https://code.launchpad.net/~unity-team/unity8/refactor-wm-and-test/+merge/172583
[14:14] <Saviq> greyback, abandoned or still development?
[14:14] <greyback> Saviq: working on that now
[14:14] <Saviq> greyback, oh cool
[14:15] <Saviq> greyback, https://code.launchpad.net/~unity-team/unity8/u8m-osk can be dropped, though?
[14:15] <Saviq> same with https://code.launchpad.net/~unity-team/unity8/unity8-integrate-mir
[14:15] <Saviq> and https://code.launchpad.net/~unity-team/unity8/unity8-run-mir-plusLinkerFun
[14:15] <greyback> Saviq: done
[14:15] <Saviq> greyback, cheers
[14:16] <Saviq> greyback, https://code.launchpad.net/~unity-team/unity8/unity8-integrate-mir still there?
[14:16] <greyback> Saviq: yep, there are other branches with MRs up to merge into it, and a build recipe
[14:16] <greyback> Saviq: going through them
[14:16] <Saviq> greyback, okies
[14:20] <Saviq> mfisch, hey, could https://code.launchpad.net/~mfisch/unity8/unity8-lp1215951 be deleted maybe?
[14:21] <Saviq> ricmm, hey, could https://code.launchpad.net/~ricmm/unity8/use-volume-key-signals be deleted maybe?
[14:21] <Saviq> or marked abandoned?
[14:21] <mfisch> Saviq: deleted
[14:21] <Saviq> sil2100, how about https://code.launchpad.net/~sil2100/unity8/test - delete / mark abandoned?
[14:21] <Saviq> mfisch, thanks!
[14:21] <ricmm> Saviq: yes
[14:21] <ricmm> deleted
[14:22] <Saviq> ricmm, thanks
[14:22]  * Saviq likes Autumn clean-ups ;D
[14:22] <Saviq> ssweeny, could https://code.launchpad.net/~ssweeny/unity8/unity8-lp1215951 be deleted?
[14:23] <ssweeny> Saviq, yes it's been superseded
[14:23] <Saviq> ssweeny, cool, could you please delete it?
[14:23] <ssweeny> Saviq, sure
[14:23] <Saviq> ssweeny, thanks
[14:24] <ssweeny> Saviq, no worries
[14:24] <Saviq> pstolowski, hey, https://code.launchpad.net/~stolowski/unity8/monitor-network should probably stay, but https://code.launchpad.net/~stolowski/unity8/search-progress-prop ? could be deleted / marked abandoned?
[14:25] <pstolowski> Saviq, right, going to remove the latter
[14:25] <Saviq> pstolowski, thanks!
[14:25] <sil2100> Saviq: I don't even remember what's this about, so delete I guess
[14:26] <Saviq> sil2100, please do :)
[14:26] <Saviq> thanks!
[14:27] <Cimi> mzanetti, we don't have a qml test for that https://code.launchpad.net/~cimi/unity8/fix-1236286/+merge/192399
[14:27] <Saviq> Cimi, great time to add one! ;)
[14:27] <Cimi> hah
[14:27] <mzanetti> Cimi: I think there is one in Shell.qml. but if I'm wrong (which might well be) then yeah, what Saviq said
[14:29] <Saviq> tsdgeos, the search entry gets confused with the tabs if you run wide screen (./run -- --fullscreen)
[14:31] <Saviq> somone's back from AMD :D
[14:31] <Saviq> tsdgeos, and clipping when overshooting in phone form factor, too
[14:31] <Cimi> mzanetti, it's a simple one
[14:32] <Cimi> mzanetti, and it still works with 26 because it's like 30 movement in the test
[14:32] <Cimi> (I checked if there was room for a test of it)
[14:32] <Saviq> tsdgeos, or maybe just the "fat bar" is confused
[14:32] <tsdgeos> Saviq: what do you mean gets confused?
[14:32] <Saviq> the one between the page header and contents
[14:32] <tsdgeos> that's it's whay of hiding
[14:32] <tsdgeos> first get wide as hell and then hide
[14:32] <Saviq> tsdgeos, the search entry? it shouldn't be hiding
[14:33] <Saviq> tsdgeos, it should only get wide and hide on screens < 60gu
[14:33] <tsdgeos> that's not how the page header code was
[14:33] <tsdgeos> it took into account the length of the label
[14:33] <tsdgeos> and it's what the new code also does
[14:34] <Saviq> tsdgeos, it only took it into account to know if it should hide or not
[14:34] <Saviq> tsdgeos, and it shouldn't take all tab labels into account, just the current one, somehow
[14:34] <Saviq> tsdgeos, on tablet it should never move, basically - just expand when you focus it
[14:34] <tsdgeos> Saviq: and if you expand it then what?
[14:34] <tsdgeos> they'll overlap
[14:34] <Saviq> tsdgeos, it only expands to 60gu or so
[14:34] <Saviq> tsdgeos, never more
[14:35] <tsdgeos> Saviq: i mean the tab bar
[14:35] <tsdgeos> or should it clip ?
[14:35] <Saviq> tsdgeos, the tab bar should be reduced to width - textField.width
[14:35] <tsdgeos> ok
[14:36] <Saviq> tsdgeos, basically, on screens > 60gu, the only movement the search entry should do is expand to 60 gu
[14:37] <Saviq> tsdgeos, or well, yeah if it fits with the label - it should not hide
[14:37] <tsdgeos> ok
[14:37] <Saviq> tsdgeos, which should not take the whole tab bar into account
[14:37] <tsdgeos> Saviq:  can you comment on the review request so i don't forget and whoever reviews next week also has it in mind?
[14:38] <Saviq> tsdgeos, yeah will do
[14:39] <mzanetti> Cimi: add a data() function to it so that it runs twice, once with less than 26, once with more than 26
[14:40] <Cimi> mzanetti, I'll see what's best
[14:43] <mzanetti> Cimi: ok
[14:43] <mzanetti> Cimi: I disapproved one of your other branches :/
[14:44] <mzanetti> Cimi: https://code.launchpad.net/~unity-team/unity8/fix-1238232/+merge/191424
[14:46] <mzanetti> Cimi: but I also marked the bug as invalid. so nothing personal to you :D
[14:48] <Cimi> hah ok
[14:49] <mzanetti> tsdgeos: you won't believe it. I still didn't apply your patch and today I had 2 crashes: one at 15:28 and one ad 16:48
[14:50] <tsdgeos> :D
[14:56]  * tsdgeos leaves for the airport
[14:56] <tsdgeos> see you guys on monday
[15:00] <nic-doffay> Saviq, there's a black bar overlaying everything, any idea how I can disable this?
[15:00] <nic-doffay> (unity8)
[15:01] <Saviq> nic-doffay, is it there in unity8 trunk?
[15:01] <nic-doffay> Saviq, let me confirm quickly...
[15:09] <mzanetti> MacSlow: ping
[15:09] <MacSlow> mzanetti, yup
[15:09] <Cimi> mzanetti, tests updated
[15:09] <mzanetti> MacSlow: why would this require autopilot tests? https://code.launchpad.net/~nicolas-doffay/unity8/dismiss-keyboard/+merge/188599
[15:09] <mzanetti> Cimi: cheers. I'll check it out soon
[15:11] <MacSlow> mzanetti, should that not be checked (protected against regressions in the future)?
[15:12] <mzanetti> MacSlow: this doesn't use any external processes or anything it communicates with
[15:12] <mzanetti> MacSlow: qml test is way to go
[15:13] <MacSlow> mzanetti, but isn't the triggering of the keyboard and the entry-field such a case?
[15:14] <MacSlow> mzanetti, I admit I don't know the implementation details atm
[15:14] <mzanetti> MacSlow: No. I think the OSK should autopilot tests to make sure it comes up with all sorts of textfields
[15:14] <MacSlow> mzanetti, ok then
[15:14] <mzanetti> MacSlow: but we in this particular case should have a wml test that makes sure we unfocus the textfield
[15:15] <mzanetti> wml -> qml
[15:15] <MacSlow> mzanetti, want me to update my comment there?
[15:15] <mzanetti> MacSlow: I added a comment, feel free to add one yourself if you want
[15:17] <mzanetti> MacSlow: added one more comment.
[15:17] <MacSlow> updated
[15:18] <Cimi> mzanetti, ti odio :D
[15:18] <mzanetti> Cimi: piacere
[15:18] <Cimi> hah
[15:18] <Cimi> mzanetti, praticamente sei passato in ogni mia review e hai aggiunto "test please" :D
[15:18] <mzanetti> Cimi: sorry... I was assigned to walk through all bugs and check for tests :/
[15:19] <Cimi> mzanetti, I hoped to get unnoticed
[15:19] <mzanetti> Cimi: If I wouldn't do it, saviq would do it in 2 weeks
[15:19] <mzanetti> Cimi: and its not just you. I even unapproved a already approved branch from dednick
[15:19] <Cimi> mzanetti, I wouldn't test this though https://code.launchpad.net/~cimi/unity8/carousel-music-video/+merge/192118
[15:19] <mzanetti> Cimi: and am accumulating testing todos for myself too
[15:19] <Saviq> Cimi, do you need a testing-is-good talk again? ;P
[15:20] <Cimi> Saviq, I know it's good
[15:20] <Cimi> Saviq, good doesn't mean fun though :)
[15:22] <mzanetti> Cimi: https://code.launchpad.net/~cimi/unity8/carousel-music-video/+merge/192118/comments/443063
[15:24] <mzanetti> Cimi: however, I know that scopes are really badly mocked right now. I would be ok with skipping tests in this one. Given that soon enough we need to do a really big Dash+Scopes Testing+cleanup session
[15:25] <mzanetti> Cimi: can you please just have a look if you think my comment is feasible right now or not
[15:25] <Cimi> mzanetti, I already replied
[15:25] <Cimi> mzanetti, it's on a different branch
[15:26] <mzanetti> Cimi: damn. I confused it with the other one...
[15:26] <mzanetti> sorry about that
[15:26]  * mzanetti has seen too many branches in the last 2 hours
[15:27] <Cimi> mzanetti, even if I hate you now, we're still friends by the way, don't worry :P
[15:27] <Cimi> Saviq, I hate you too don't be jealous :P
[15:27] <mzanetti> Cimi: But you know what I mean... we can write tests that at least load the component and check if it at least compiles
[15:27] <Cimi> yeah
[15:27]  * Cimi is joking
[15:28] <mzanetti> :)
[15:28]  * mzanetti => 10 minute break before going crazy
[15:43] <Saviq> ok, EOW peeps, talk to you next time o/
[15:51] <mzanetti> Saviq: bye
[15:54] <Cimi> you here guys? :)
[15:54] <Cimi> mzanetti, or Saviq :)
[15:54] <mzanetti> I am
[15:54] <mzanetti> Cimi: ^
[15:58] <mzanetti> Cimi: notice something here? https://code.launchpad.net/~om26er/unity8/search_indicator_dash_only/+merge/191012
[15:59] <Cimi> mzanetti, missing tests?
[15:59] <mzanetti> ;)
[15:59] <mzanetti> Cimi: you should ask others for tests too. (including me and Saviq)
[16:01] <mzanetti> om26er: https://code.launchpad.net/~om26er/unity8/search_indicator_dash_only/+merge/191012
[16:02] <om26er> mzanetti, I am on vacation till 5th Nov. but if you give me pointers I'll fully try to write a test for that voluntarily
[16:03] <mzanetti> om26er: ah ok. I'll be away for one week starting on the 5th :)
[16:03] <om26er> mzanetti, can you comment on the MR for what needs to be done. I would like to improve my qml testing skills for sure
[16:04] <mzanetti> om26er: well, it's quite straigt forward: there is a tests/qmltests/tst_shell.qml. Just add a new function in there, and do whatever you'd do in autopilot too. instead of select_single() etc you can use findChild()
[16:04] <mzanetti> om26er: the rest is the same, except it's easier in qml ;)
[16:05] <mzanetti> there is compare() just like Equals() and tryCompare() instead of Eventually(Equals())
[16:05] <mzanetti> om26er: ^
[16:06] <om26er> mzanetti, ok, I am *trying* to fix a bug in the terminal app on the phone, once I am done with that. I'll pick this one
[16:06] <mzanetti> om26er: feel free to ping me for help anytime
[16:06] <om26er> mzanetti, sure will do. Thanks :)
[16:10] <mzanetti> om26er: heh, found this. looks like its not that new for you after all :) https://code.launchpad.net/~om26er/unity8/fix_1238837/+merge/190743
[16:10] <om26er> mzanetti, yep, that was like my first qml test.
[16:11] <mzanetti> om26er: ok. because you said that I read through it and I think I spotted a problem
[16:11] <mzanetti> or better a weakness
[16:12] <mzanetti> om26er: the test_clickTileNotClose() clicks on it and *immediately* checks if if the tile is still here
[16:12] <om26er> mzanetti, ah? what's that ?
[16:12] <mzanetti> om26er: while most likely it would take a couple of ms to actually close the app
[16:12] <mzanetti> and this test wouldn't notice
[16:14] <om26er> mzanetti, we are verifying that twice, first we make sure the running app is not in the model and only then check if the tile has finally vanished
[16:15] <mzanetti> yeah. but if there is one async operation in closing the app both checks would run before
[16:16] <mzanetti> om26er: as we're using mocks, most likely everything in the backend is syncronous. But remember, an autopilot test like this woule *never* fail even though the app would be closed when clicking on the tile
[16:20] <om26er> mzanetti, I have a hard time understanding that, do you mean we need an autopilot test for that or telling me qmltests are better ? :D
[16:21] <mzanetti> om26er: neither of that. actually scratch the autopilot comment
[16:22] <mzanetti> om26er: what I meant is that both, qmltests and autopilot tests are not really useful if written in such a way
[16:22] <mzanetti> om26er: as it's really easy to break the code without getting noticed by the test
[16:23] <om26er> mzanetti, in autopilot I would really make sure that the tile disappeared and the number of tiles previously and now is different
[16:24] <mzanetti> om26er: yeah that's the thing. this test checks if the number is still the same as before, and it does that immediately
[16:24] <mzanetti> om26er: so if the number of tiles changes a few milliseconds later, for example because of an async operation somewhere in the code, the test checks have already ra
[16:24] <mzanetti> n
[16:24] <mzanetti> and after the test passed the tile goes away
[16:25] <om26er> mzanetti, is there something like "Eventually" here ?
[16:25] <mzanetti> om26er: unfortunately not really
[16:25] <mzanetti> om26er: one of the rare cases where we allow to use wait() in qmltests
[16:26] <om26er> mzanetti, does that take a time parameter or it has its own timeout ?
[16:26] <mzanetti> om26er: because with wait(0) you trigger the event loop without really sleeping. so the test would still catch single async operations. still not ideal, I agree
[16:26] <mzanetti> om26er: wait(milliseconds)
[16:27] <mzanetti> whereas 0 has the special meaning of triggering the event loop and continue _at the end_ of the scheduled things in the next event loop run
[16:28] <mzanetti> see QTimer docs for that
[16:33] <om26er> mzanetti, ack. I'll update this one, and add a new test for the other branch
[16:33] <mzanetti> om26er: thanks