1328074 - [e10s][APZ] Images on EPIX.com, cips.org twitch when I autoscroll the page

Status: NEW

(Core :: Panning and Zooming, defect, P3)

Description

[arni2033]

>>>   My Info:   Win7_64, Nightly 49, 32bit, ID 20160526082509
STR_1:
1. Open url http://epix.com/
2. Hover mouse over any relatively small (narrower than 50% of the page) image
3. Move mouse away from image and immediately start slowly autoscrolling the page

AR:  The image twitches up and down for several moments
ER:  Images shouldn't twitch, just like before regression 1257641

Note:  (testcase)
That page does this on 'mouseover' event:
  TweenMax.to(elt, 15, {scale:1.2, ease:{_func: null, _type: 1, _power: 3, _params: [0, 0, 1, 1] }});
then (2) on 'mouseout':
  TweenMax.to(elt, 15, {scale:1, ease:{_func: null, _type: 1, _power: 3, _params: [0, 0, 1, 1] }});
where "elt" is hovered element with selector "div.block-item-bg".
Several moments after (2) is finished, you can see the lags/twitching described above.
If you want a standalone testcase, then download the page, then add the following 2 attributes to 
some "div.block-item-bg" element, and also apply stylesheet   .block-item-gradient{display:none}
onmouseover="TweenMax.to(this,15,{scale:1.2,ease:{_func:null,_type:1,_power:3,_params:[0,0,1,1]}});" 
onmouseout="TweenMax.to(this,15,{scale:1,ease:{_func:null,_type:1,_power:3,_params:[0,0,1,1]}});"

This is regression from bug 1257641. Regression range:
> https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=8827b7c4748b20120f7ad8816a546ebbeb82dd4b&tochange=1bdfcc34a189c1fd9a307814488b47a6c595d769


STR_2:
1. Open https://www.cips.org/supply-management/
2. Scroll the page so section "Events" is visible
3.A) Wait until "Events" section starts moving, then immediately autoscroll the page up or down
3.B) Click on right button in the section "Latest jobs", then immediately autoscroll the page up/down

AR:  Moving elements fly a bit to the top/bottom during autoscrolling
ER:  They should be animated in a normal way@ Kartikaya Gupta (email:[e-mail]):
It seems that this is a regresion caused by your change. Please have a look.

Comment 1

[Kartikaya Gupta (email:[e-mail])]

I can repro STR_2 pretty easily. I think I saw STR_1 as well but it's much harder to repro. Will focus on STR_2 in this bug.

Comment 2

[Milan Sreckovic [:milan] (needinfo for best results)]

Kats, is it OK to stop tracking this?

Comment 3

[Kartikaya Gupta (email:[e-mail])]

Yeah, I guess so. It's a legitimate bug but I don't think we have cycles to look into this right now.