Switching between Tabs via TabStrip produces visible lag |
||||||
Issue descriptionApp Version (from "Chrome Settings > About Chrome"): 70.3538.44 beta iOS Version: 12.0 Device: iPad Pro 10.5-Inch Steps to reproduce: 1.) Open 2 NTP tabs 2.) Press on a tap and move the finger to switch between tabs 3.) Load URLs in both Tabs 4.) Repeat step 2 Observed behavior: Switching is fast on step 2, but have noticeable lag on step 4 Expected behavior: Switching should be fast on step 2 and 4 Additional comments: Reported by external Beta user.
,
Oct 10
This is always the case. Tested on M60.0.3112.72 and the behavior is same.
,
Nov 4
I wonder if this is simply because we trigger -tabWasTapped on touchUpInside instead of touchDown. rohitrao@ any historical reason why we trigger this on touchUp? It looks like most other browsers (desktop, safari, etc) go with touchDown. srikanthg@ Lets try this on a dev build with touchDown and see if you still see the difference (rather than landing this change for now)
,
Nov 5
,
Nov 6
Does touchDown make a difference? Does the video show a touch+drag? Why is touchUpInside being delivered at all in that case? I probably used touchUpInside because that's what we do for other buttons. It's ok to change if we think touchDown would be better. Right now can you longpress+drag to reorder a tab without selecting it? We'd lose that if we switched.
,
Nov 6
> We'd lose that if we switched. That's correct -- longpress + drag on desktop also selects first.
,
Nov 6
If you have a dev build with the CL, then please share it on Google Drive, so that I can take a look test your change.
,
Nov 6
+edchin for snapshot question Turns out snapshots are really slow here. Do we need tab_strip_controller::-tabViewTapped to take a snapshot? Note, we won't take that snapshot when changing tabs from the keyboard, or from swipe gestures, etc.
,
Nov 13
eugenebut@, shbarezer@, and I just looked at this on real iPad devices with M70+ and were not able to repro the lag. The switching was fast and expected. I'll capture video.
,
Nov 13
,
Nov 13
Did a VC with edchin@ to explain the issue, and filed 905065 for investigation. |
||||||
►
Sign in to add a comment |
||||||
Comment 1 by linds...@chromium.org
, Oct 10Labels: M-72
Owner: justincohen@chromium.org
Status: Assigned (was: Untriaged)