Remove CSS_SHADERS compile-time flag.

Remove CSS_SHADERS compile-time flag.

The compile-time was on by default in Chromium builds. The feature remains guarded by a runtime flag.

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=147953

[Max Vujovic (mvujovic), 2013-04-09 18:02:17]

Thanks for the perl script, Eric! It removed most of the occurrences.

Do you think you could review this?

[eseidel, 2013-04-09 18:30:08]

looks great.  I believe we're supposed to have you sign something?  I'm very new
to this chromium-process.
http://www.chromium.org/developers/contributing-code/external-contributor-che...

https://codereview.chromium.org/13839016/diff/1/Source/WebCore/css/CSSValue.cpp
File Source/WebCore/css/CSSValue.cpp (left):

https://codereview.chromium.org/13839016/diff/1/Source/WebCore/css/CSSValue.c...
Source/WebCore/css/CSSValue.cpp:230: #if ENABLE(CSS_FILTERS) &&
ENABLE(CSS_SHADERS)
CSS_FILTERS?

https://codereview.chromium.org/13839016/diff/1/Source/WebCore/css/CSSValue.h
File Source/WebCore/css/CSSValue.h (left):

https://codereview.chromium.org/13839016/diff/1/Source/WebCore/css/CSSValue.h...
Source/WebCore/css/CSSValue.h:164: #if ENABLE(CSS_FILTERS) &&
ENABLE(CSS_SHADERS)
Are CSS_FILTERS always compiled in as well?

[jamesr, 2013-04-09 18:31:36]

On 2013/04/09 18:30:08, Eric Seidel (Google) wrote:
> looks great.  I believe we're supposed to have you sign something?  I'm very
new
> to this chromium-process.

Adobe has signed the CLA.

[Max Vujovic (mvujovic), 2013-04-09 18:53:14]

Thanks for the review, Eric!

As James said, my name is under Adobe's CLA with Google/Chromium, so we should
be all set.

Do you think you could commit this patch, since I don't have svn credentials
yet? Let me know if the patch gets stale, I'll happily rebase and reupload.

I'm assuming svn credentials will be distributed to external contributors when
Blink is a little older.

https://codereview.chromium.org/13839016/diff/1/Source/WebCore/css/CSSValue.h
File Source/WebCore/css/CSSValue.h (left):

https://codereview.chromium.org/13839016/diff/1/Source/WebCore/css/CSSValue.h...
Source/WebCore/css/CSSValue.h:164: #if ENABLE(CSS_FILTERS) &&
ENABLE(CSS_SHADERS)
On 2013/04/09 18:30:08, Eric Seidel (Google) wrote:
> Are CSS_FILTERS always compiled in as well?

CSS_FILTERS are always compiled in as well. I'll be removing the CSS_FILTERS
flag next. 

CSS_SHADERS are a subset of CSS_FILTERS. This line is actually wrong in WebKit.
It should only check CSS_SHADERS because in WebKit, we assume if CSS_SHADERS are
on in a build, then CSS_FILTERS are on.

[eseidel, 2013-04-09 18:54:49]

thanks.  lgtm.  I'll figure out how to commit this for you. :)

[eseidel, 2013-04-09 19:10:21]

Committed patchset #1 manually as r147953 (presubmit successful).

[eseidel, 2013-04-09 19:12:20]

The fact that our tools currently make no attempt at attribution of the commit
seems super-lame.

https://chromium.googlesource.com/chromium/blink/+/5945f325bfca6ddce8bf0a50f7...

Another thing to fix. :)

[eseidel, 2013-04-09 19:23:53]

I filed https://code.google.com/p/chromium/issues/detail?id=229526

[jamesr, 2013-04-09 19:27:53]

Has an API review been scheduled for CSS_SHADERS?

[eseidel, 2013-04-09 19:32:55]

It appears it's already tracked on the dashboard:
http://www.chromestatus.com/features

We're the first to implement it (prefixed), but it has support.

Presumably to follow our new feature policy we should be moving away from
prefixing and making sure it's disabled by a runtime flag in Stable builds.

[Max Vujovic (mvujovic), 2013-04-09 20:22:40]

On 2013/04/09 19:10:21, Eric Seidel (Google) wrote:
> Committed patchset #1 manually as r147953 (presubmit successful).

Thanks Eric!

[Max Vujovic (mvujovic), 2013-04-09 20:23:42]

On 2013/04/09 19:27:53, jamesr wrote:
> Has an API review been scheduled for CSS_SHADERS?

Not yet- I'll email meh to schedule one.

[Max Vujovic (mvujovic), 2013-04-09 20:41:30]

On 2013/04/09 19:32:55, Eric Seidel (Google) wrote:
> It appears it's already tracked on the dashboard:
> http://www.chromestatus.com/features
> 
> We're the first to implement it (prefixed), but it has support.
> 
> Presumably to follow our new feature policy we should be moving away from
> prefixing and making sure it's disabled by a runtime flag in Stable builds.

Yeah, it's disabled by a runtime flag. It's prefixed in two ways, through the
CSS filter property, and the new filter at-rule, which is currently only being
parsed.

The currently functional syntax looks like:

-webkit-filter: custom(url(page-curl.vs) mix(url(page-curl.fs) normal
source-atop), 10 10 attached, direction 90, amount 0.5);

For various good reasons, the spec has recently changed to use a new syntax
(shown unprefixed):

filter: custom(page-curl, direction 90, amount 0.5);

@filter page-curl {
    src: url(page-curl.vs), url(page-curl.fs);
    parameters: direction 180, amount 0;
    mesh: grid(10 10 attached);
    mix: multiply source-atop;
}

This and other topics sounds perfect for the API review :)