mozilla :: #fx-team

20 Apr 2017
04:55dietrichanyoe have ideas on who to ni? to get a quick fix to 53 release page? https://bugzilla.mozilla.org/show_bug.cgi?id=1358003
04:55firebotBug 1358003 NEW, nobody@mozilla.org 53 release page has a typo
10:23Gijsmikedeboer: async await, eh
10:23mikedeboerGijs: not working for ya?
10:24Gijsmikedeboer: funny, that, it looks like that change makes tests fail, maybe because head.js helpers for CUI still use Promise.jsm and Promise.defer()
10:24Gijsmikedeboer: still looking if it's not just me being an idiot
10:24mikedeboerthat *should* work
10:24Gijsmikedeboer: we should really just wholesale convert that test dir
10:24mikedeboer+1
10:24GijsI might look at this once I have something that works for this bug
10:27Gijshm, yeah, I think it's something else
10:27GijsWell, I don't know yet
10:27Gijsbut fixing those helpers didn't help
10:28mikedeboer:*(
10:28mikedeboerheh, where did that asterisk come from? I'm crying snowflakes :P
10:28Gijsoh
10:29Gijsodd, the test works in isolation but not when run as part of the suite
10:29GijsI swear it worked yesterday before I started fixing the review comments though
10:29* Gijs shelves changes and tries again
10:31Gijsnope, now it's just perma-sad even after shelving my last few changes
10:31Gijssigh.
10:38mikedeboerGijs: perhaps I've got something to distract you for a bit ;)
10:38Gijsoh dear
10:38Gijsmikedeboer: style changes to subviews?
10:38mikedeboerGijs: no, the proposal I talked about yesterday
10:39Gijsoh, ehm, refresh my memory? Sorry, juggling a bunch of things :)
10:40mikedeboerGijs: it's in a review request, don't worry - easy to find, just code
10:41Gijsoh shoot, you're forking?
10:41mikedeboeryeah, as in proposing to
10:42mikedeboermight seem reasonable - please read on for a bit - happy to jump on Vidyo for a bit to talk about it
10:43GijsI can't do this without tea
10:43Gijsmikedeboer: vidyo in 5 minutes? :)
10:43mikedeboerGijs: hehe, of course
10:48Gijsmikedeboer: in my room :)
12:00Gijsgaaahh
12:00GijsI seem to have broken mozreview
12:01Gijshttps://reviewboard.mozilla.org/r/131626/
12:01Gijslists 2 issues and I can't autoland
12:01Gijsbut the issue list has 0 unfixed issues
12:03Standard8Gijs: congrats!
12:03Standard8;-)
12:03Standard8#mozreview are fairly responsive normally
12:04Gijsyep, there already, no response so far
12:04Standard8I see the same as you fwiw
12:04mikedeboerGijs: yeah, had that before, smacleod fixed it for me before
12:04mikedeboer2x 'before', I apologize humbly.
13:56mikedeboerGijs: I&#39;m lost. Why does `.panel-viewcontainer[panelopen] > .panel-viewstack > panelview[current]` match the first <panelview>, but not the second view in the same container?
13:56Gijsmikedeboer: the current attribute gets set on the current container, I think?
13:56Gijserr
13:56Gijscurrent panelview
13:56Gijsso it&#39;ll only match 1 panel at a time
13:56Gijsif you remove the attribute from one view and stick it on another one, that one will match, I expect
13:56Gijsdoesn&#39;t it?
13:56mikedeboerGijs: it also doesn&#39;t work when I remove the [current] part
13:57Gijserr, don&#39;t know then...
13:57mikedeboerand I set the &#39;current&#39; attribute
13:57Gijsyou can try to use element.matches() to see which bit of the selector it&#39;s failing on, maybe?
13:57Gijsprobably want to sanity check that it works correctly by also checking against the second panelview
13:58Gijs(or the first, or whichever one *is* matching)
15:22dolskemikedeboer / jaws: does theming still need the cedar branch?
15:22mikedeboerdolske: nope
15:22dolskemikedeboer: cool. imma take it then.
15:23mikedeboerfeel free to :)_
15:27jawsit&#39;s all yours :-)
15:42Gijsmconley / florian: is there some way to flag stuff like bug 1358130 for profile interpretation attention for folks on photon-perf / qf?
15:42firebothttps://bugzil.la/1358130 NEW, nobody@mozilla.org Loading Expensify brings the browser to a halt
15:42mconleyGijs: yep - please add the [qf] tag to the whiteboard
15:42mconleywe&#39;re doing a triage tonight
15:43Gijsmconley: done, thanks!
15:45mconleythank you
16:20sfostermconley: will you look at that :) https://treeherder.mozilla.org/#/jobs?repo=try&revision=a0e65171a7d267ac70d6df21e9e58b5c53bf920d
16:20sfosterclean and green.
16:20sfostereslint errors are just from my logging. I&#39;ll strip that and attempt to land. 3rd time lucky
16:22Gijssfoster++ for the ridiculous amount of effort that single bug ended up requiring
16:22sfosterGijs: quite a bit of it self-inflicted sadly
16:22GijsI promise, not all Firefox development is that annoying
16:22Gijswellllll
16:22sfosterbut its a learning curve right
16:23Gijssfoster: honestly, I don&#39;t think it&#39;s you, it&#39;s us/the platform/whatever
16:23sfosterthere&#39;s no shortcuts on this stuff. You just have to make all the mistakes sooner or later
16:24sfosterGijs: was going to ask who would be a good person to give me a whistle-stop tour of the download toolbarbutton and how that works. I&#39;m on the hook for implementing the new photon animation
16:24sfosterI have dim memories of dealing with downloads in Metro
16:24Gijssfoster: a little bird told me that animation might be changing... but paolo would be the best person, and I can sub if he&#39;s not available
16:25MossopGijs, sfoster: Mind giving me a quick overview of what made that bug so much work?
16:25sfosterpaolo sounds like a european timezone.
16:25Gijssfoster: UK, same for me
16:26sfosterMossop: it was mostly writing a test that would catch a regression (and verify the patch). Using the WindowUtils elementsStyled.
16:27sfosterBut I borked a merge and spent a couple days testing a bad patch - in which the test was trying hard to tell me it was broken but I assumed erratic test results.
16:28sfosterMossop: also some back and forth on how much solution was enough, naming, and the usual stuff.
16:28MossopAh ok
16:29sfosterI shuldn&#39;t speak too soon, its not landed yet. But that last try run is encouraging.
16:34mconleysfoster: heyo
16:35mconleysfoster: might be worth a bunch of retriggers to be sure
16:35mconleysfoster: out of curiosity, what change did you make?
16:36sfostermconley: 2 changes. The first was actually applying my patch to browser.js which got lost in a duff merge :/
16:36sfosterThe 2nd was applying your suggestions to wait for window focus before snapshotting intial elementsStyled count.
16:37mconleycool
16:37mconleysfoster: just as an added precaution, I highly advise a few retriggers to be safe. Good stuff.
16:37sfosterwilld o.
16:37sfosterdo.
16:55sfosterdao: thanks for pointing me at bug 767133. I&#39;m building that revision, but it looks like that will clip the contents as they slide in?
16:55firebothttps://bugzil.la/767133 FIXED, dao+bmo@mozilla.com Add slide-in animation for arrow panels
16:55sfosterI&#39;ve ni?d neal but no response from him yet. I&#39;m happy to move this out to another .css file, but unless I duplicate it across all the themes/platforms its not obvious where to.
16:56sfosterThe current animation is implemented in xul.css
16:56sfosterArguably it shouldnt be, according to the comments on xul.css
17:27jawswhich talos group should be chosen to run &#39;tabpaint&#39; ?
17:28jawsmconley: you&#39;re listed at https://wiki.mozilla.org/Buildbot/Talos/Tests#tabpaint maybe you would know?
17:36Aryxjaws: &quot;other&quot;: https://dxr.mozilla.org/mozilla-central/source/testing/talos/talos.json#25
17:37jawsAryx: thanks!
17:50mconleyjaws: sorry, was in a meeting
17:50jawsno worries :)
18:06jawssfoster: http://searchfox.org/mozilla-central/source/browser/base/content/browser.js#1133
18:06jawssfoster: actually, http://searchfox.org/mozilla-central/source/browser/base/content/browser.js#1249
18:16sfosterjaws: I filed bug 1358215
18:16firebothttps://bugzil.la/1358215 NEW, nobody@mozilla.org Add pref to facilitate early landing of photon-animation work ahead of v57
18:22dolskejaws / squib: belated thought... if that talos issue is actually due to animations now being measured because of the prefname change... we should probably still look into that separately (maybe the perf team can help). Because that would imply animations have been slowing down real-world tab load, which isn&#39;t great.
18:24sfosterwhere can I download a copy of v16? Do we have an archive of earlier releases somewhere?
18:24Mossopsfoster: Firefox 16?
18:24sfosteryes
18:24Mossopsfoster: http://ftp.mozilla.org/pub/firefox/releases/
18:25sfosterthanks
18:54jawsMattN: did you intentionally clear the needinfo for giorgios from bug 1352501? it seemed the question about future UITour usage of reader mode hasn&#39;t been answered
18:54firebothttps://bugzil.la/1352501 ASSIGNED, mconley@mozilla.com Remove Reader Mode feature promotion panel
18:55MattNYes, it wasn&#39;t relevant to the ReaderParent change that was being discussed
18:55MattNUITour doesn&#39;t use showReaderModeInfoPanel
19:04mconleyMattN: ah - so it shows its own panel, gotcha
19:04MattNyes
19:04mconleythanks for clearing that up
19:04mconleysorry for introducing confusion. :)
19:05MattNno worries
19:12dolsketimdream: hmm, that&#39;s unfortunate. The new EmojiOne release switched to a non-free license. :/ https://emojione.com/blog/introducing-emojione-v30-at
19:26jawsMattN / mconley: okay thanks! well then my note about UITour in my review doesn&#39;t apply. still r+ though
19:27mconleyr+, but I guess I can be a little more vigorous with my patch
19:27mconleyI&#39;ll have a new one up soon
19:27MattNHmm I agreed with you about removing the variable you mentioned
19:27MattNbut it&#39;s in ReaderParent
20:06spohlare there instances where we use workers in .jsm files? I&#39;m wondering if there is any way to move the loop at https://dxr.mozilla.org/mozilla-central/source/toolkit/modules/GMPInstallManager.jsm#427 to a worker thread to avoid blocking the main thread
20:07spohlGijs suggested expanding nsIZipReader&#39;s API with an async version of extract, but before I do that I want to make sure that I can&#39;t simply use a worker thread here
20:15mconleyspohl: here&#39;s an example: http://searchfox.org/mozilla-central/source/toolkit/components/promiseworker/PromiseWorker.jsm#173
20:18spohlmconley: thank you
20:18mconleyspohl: np
23:02bdanforth4.1.1. id refers to the specific permission (ex: geolocation, web camera...). Are these the only two permissions?
23:02bdanforth4.1.2. Currently this test performs this check only for the geolocation permission, but it should be performed for all permissions associated with the WebRTC framework (see Question 1 to clarify which permissions to check)?
23:02bdanforth4.1.3. When you say &#39;WebRTC permissions use a different codepath&#39;, what do you mean?
23:03bdanforth...Did I just send that to everyone :D
23:14felipeis there a bug about the progress on the download arrow flickering?
21 Apr 2017
No messages
   
Last message: 91 days and 7 hours ago