mozilla :: #apz

11 Sep 2017
11:05mstangebotond: I retract all my opposition to the apz frame delay; it seems I flipped it back on in my main profile and didn't realize it, the only thing I realized was that the header parallax on twitter seemed better than usual
11:42mstange*the only thing I noticed
14:26botond|homemstange: great! :)
15:21botondmstange: does that mean you'd be ok with it riding the trains in 57?
15:21mstangebotond: yes
15:23botondkats: hey, welcome back :)
15:24katsbotond: thanks :)
15:24botondkats: how was your PTO?
15:24katsbotond: it was alright. my parents were visiting
15:24botondkats: cool
15:25katswe drove up to quebec city and montreal
15:25katsweather was a bit lousy but overall still good
15:26botondkats: yeah, it's been a bit cool and rainy at time here too. fall starting early, i guess...
15:26katsbotond: yeah trees were already going red
15:27botondkats: anyways, when you have a few minutes, there are a couple of things i'd like to run by you
15:27katsbotond: sure, go ahead
15:28botondkats: so first, i'm not sure if you recall this optimization: http://searchfox.org/mozilla-central/source/gfx/layers/apz/util/APZEventState.cpp#208
15:29botondkats: i refreshed myself on it and summarized it here: https://bugzilla.mozilla.org/show_bug.cgi?id=1385789#c18
15:29firebotBug 1385789 FIXED, bwerth@mozilla.com stylo: remove HasStateDependentStyle warnings
15:30katsbotond: your summary and conclusion seems accurate to me
15:31botondkats: it's been suggested to me by :emilio that the optimization might not be all that great of an idea, because the necessary style check is expensive
15:32katsbotond: i'd be ok with removing the optimization
15:32botondkats: (http://logs.glob.uno/?c=mozilla%23developers&s=6+Sep+2017&e=6+Sep+2017&h=StateDep#c1652683)
15:32botondkats: and so that we might be better of removing it
15:33botondkats: if that sounds reasonable to you, i can go ahead and do that
15:33katsbotond: sgtm
15:33botondkats: great!
15:33botondkats: second thing, i wanted to touch base with you about apz.frame_delay.enabled
15:34botondkats: we've had enabled on Nightly for over a month now, with no complaints, except from Markus who retracted his objection today (see above)
15:35botondkats: it did cause a regression on Fennec (which I haven't looked into yet, partly because I just got my new Android phone from jlin today), but that's specific to the dynamic toolbar, and there's no reason to believe it would cause any correspoding issue on desktop
15:35botondkats: so, i was wondering what you thought about having it ride the 57 train
15:36katsbotond: would you make it ride 57 for desktop only? or fennec as well?
15:36botondkats: desktop only (sorry, should have clarified that)
15:36botondkats: for fennec it can ride the trains once i fix the dynamic toolbar issue
15:37katsbotond: that sounds fine to me. i'd slightly prefer if it rode the trains everywhere together but there are advantages to having them ride separately as well (if it turns out you can't fix the dynamic toolbar issue in 57)
15:38katsbotond: the "advantages" being that we still exercise the non-frame-delay codepaths in a shipping product so if we need to flip it off later we can do so
15:39botondkats: i can try to prioritize the dynamic toolbar issue and fix it before 57, but if it's going to ride the trains on desktop, i'd prefer to land that patch sooner rather than later (partly because there's apparently a "soft freeze", starting, depending on who you ask, September 18th, or today :))
15:39katsbotond: yup makes sense, rs=me to let it ride on desktop
15:39botondkats: ok, great - thanks!
16:37botondkats: btw. - will you be going to the All Hands now that it's in Austin?
16:37katsbotond: probably not
16:37botondmstange: ^ how about you?
16:38mstangeI don't know :(
16:38mstangeI'd really like to attend another all hands
20:57botondrbarker: ping?
21:04rbarkerbotond: pong
21:04botondrbarker: hey - remember the dynamic toolbar regression that apz.frame_delay.enabled caused (bug 1390145), which was temporarily worked around by disabling that pref on Fennec?
21:04firebothttps://bugzil.la/1390145 FIXED, rbarker@mozilla.com Content jumps around when hiding/showing urlbar
21:05rbarkerbotond: yep
21:05botondrbarker: you said at the time on IRC (http://logs.glob.uno/?c=mozilla%23apz&s=1+Aug+2017&e=28+Aug+2017#c37838): "so I think the issue is this:", and then "actually i'll put the issue in the bug so it doesn't get lost to sand of IRC"
21:06rbarkerbotond: and promptly forgot about :)
21:06rbarkerit
21:06rbarkerlet me go do that right now :)
21:06botondrbarker: thanks :)
21:45rbarkerbotond: https://bugzilla.mozilla.org/show_bug.cgi?id=1390145#c8
21:45firebotBug 1390145 FIXED, rbarker@mozilla.com Content jumps around when hiding/showing urlbar
21:46rbarkerbotond: let me know if you need more details.
21:48botondrbarker: thanks!
12 Sep 2017
No messages
   
Last message: 9 days and 20 hours ago