cc: Pixel test image results for https://codereview.chromium.org/14244021/

cc: Pixel test image results for https://codereview.chromium.org/14244021/

R=enne
BUG=171734
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=194626

[enne (OOO), 2013-04-17 16:14:22]

lgtm!  These are great.  :)

[enne (OOO), 2013-04-17 16:18:14]

Actually, scratch that.  I have two changes to suggest:

(1) Because of the gl clamping issue that you mention on the other bug, I think
it might be useful for the mask to extend all the way to the edge of its
texture.  It might turn something else up.

(2) Maybe you can reuse the same image for these, but what about content scales
differences?

[danakj, 2013-04-17 16:53:16]

On Wed, Apr 17, 2013 at 12:18 PM, <enne@chromium.org> wrote:

> Actually, scratch that.  I have two changes to suggest:
>
> (1) Because of the gl clamping issue that you mention on the other bug, I
> think
> it might be useful for the mask to extend all the way to the edge of its
> texture.  It might turn something else up.
>

I tried explicitly to avoid that, cuz it's currently a bug, and if we fix
it these results won't change.

I could add another one that demonstrates the bug if you want? Would that
be useful for what you're thinking? Like the repro on
https://code.google.com/p/chromium/issues/detail?id=232195

>
> (2) Maybe you can reuse the same image for these, but what about content
> scales
> differences?
>

The ImageMaskOfLayer uses an ImageLayer with contents at a scale 4x larger
than the output being displayed, so it should cover this, unless you're
thinking of something else.


>
>
https://codereview.chromium.**org/14299010/<https://codereview.chromium.org/1...
>

[enne (OOO), 2013-04-17 16:55:24]

I'm a little behind the times and saw that you filed a bug for this.  I guess
when it gets fixed, the unit test can be updated, and the pixel tests shouldn't
change.

Ah, I missed the fact that the image mask was larger.  So, great!  Ignore my
previous comments and land this puppy.

[danakj, 2013-04-17 17:46:23]

Committed patchset #1 manually as r194626 (presubmit successful).