====================================

Good Build:

36.0.1917.0    Base Position: 260463


Bad Build:

36.0.1930.0    Base Position: 262250

=====================================

Able to repro this issue on Windows 7, MAC (10.11.6) & Ubuntu Trusty (14.04) for the Google Chrome Stable Version - 52.0.2743.116

This is a regression issue broken in M36, below mentioned is the bisect info:

CHANGELOG URL: https://chromium.googlesource.com/chromium/src/+log/65931348282198089ee550c681c5b3b7fdcea704..6b9c3bd1ad6ff9c702be378ba0e8d1bc2cb6e523

BLINK CHANGELOG URL: https://chromium.googlesource.com/chromium/blink/+log/74531855e974dbf4a849ab27fc306f60f260803a..901a0d538709c60c743f6efaf34d951654c6e12f

Suspecting Commit: fadb854d3b738ec55647b9a85557658ce5e8b079

Review URL: https://codereview.chromium.org/222473005

@esprehn: Could you please look into the issue, and if it has nothing to do with your changes and if possible please do assign it to the concerned owner.

Thank you.

The glitch occurs due to the a transition: transform.

There is this rule:
		.card:hover .card-image {
			transform: scale(1.1, 1.1);
		}

and this rule:

		.card.selected {
			z-index: 1;
			transform: scale(1.11, 1.11);
			cursor: default;
		}

So when the hover state changes we seem to believe that there is a transform transition and we animate the transform from 1.11 to 1.11, in the process causing some snap behavior introducing the glitch.

First question is for the animation team: Should we even be doing the transition in this case? It seems not since the transform is not actually changing.

As per comment #1, this is a regression since M36. If that is true, then this won't block M53 Beta/Stable , correct?

schenney@ You have 1.1 and 1.11, so it would transition by 0.01 ?

@govind, the confusion was that in comment #1 the M-53 label was added. Normally a bisect would add the M label corresponding to the regression.

@schenny, see comment #7.

This is probably not even a regression since the range flagged includes "Enable unprefixed CSS Transforms by default", and the testcase uses only unprefixed transforms.

I think the 1.1 vs 1.11 difference is a red herring. They're not applying to the same thing, so I don't think anything is transitioning from 1.1 to 1.11. Even if they're the same value, the text glitches.

When I change them both to 1.0, and zoom in, I can see the hover causing the "GROUP" text in box 1 to jump, and the border-radius corner rounding in boxes 2 and 3 to momentarily disappear. This still happens if I remove the transform on .card.selected. Then removing the transform on .card.selected .card-content changes things again, but the border-radius change is still happening.

It's not clear to me whether this is an animations issue or not. Sending back over to Paint folks for the next pass. Also changing from Bug-Regression to Bug per #9.

This issue has been Available for over a year. If it's no longer important or seems unlikely to be fixed, please consider closing it out. If it is important, please re-triage the issue.

Sorry for the inconvenience if the bug really should have been left as Available. If you change it back, also remove the "Hotlist-Recharge-Cold" label.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

This appears to be fixed. At 100% on Linux there is no change in the left text box when the center and right are hovered. This might be due to the fix for sub-pixel layer positions.

At 125% there is a minor text layout change in the left box - the "i" moves a little. That is almost certainly some snapping issue but not worth fixing.

If your experience differs from this, please re-open the bug.