Bug 48288 – Fix LayoutTests/canvas/philip/tests/2d.composite.operation.clear.html

• Status: RESOLVED
• Resolution: WONTFIX
• Component: Canvas

[Mike Lawther]

This layout tests fails. See master bug: https://bugs.webkit.org/show_bug.cgi?id=46506

[Kulanthaivel Palanichamy]

Created attachment 94273 [details]
Proposed patch.

This patch removes three compositor operators (clear, darker and highlight) that are not valid by the canvas spec.  Operator "clear" seems most popularly used and there may be backward compatibility issues on removal.  However, to support this operator and confirm to the spec, the implementation has to rename it something like "webkit-clear'.

[Matthew Delaney]

Yea, this has come up many times. Here's the issue: removing clear (or even renaming it to webkit-clear) agrees with the spec but then will break sites that rely on "clear" as it were. There are many clients of WebKit such as DashBoard widgets on mac that rely on this and don't plan to change. Thus, although it's not ideal, the consensus is that it's best to leave these operators there for backward compatibility.

[Kulanthaivel Palanichamy]

So, is the consensus then is to keep all 3 operators or just 'clear'?  If it is former, the patch could be reworked to reduce the scope.  Just for my knowledge: is there a way to mark a bug to "don't attempt to fix" or withdraw so other will try not to fix them.

[Matthew Delaney]

(In reply to comment #3)
> So, is the consensus then is to keep all 3 operators or just 'clear'?  If it is former, the patch could be reworked to reduce the scope. 

I would argue that the only reason we should consider dropping support for any one of them is if both:
a) They are getting in the way of properly implementing other more important features.
AND
b) We're *very* confident that very few pages rely on their functionality.

There are still many other vastly more important features that we need to implement and/or improve upon. So, unless these 'old' ones gets in the way, there's no benefit to our users to drop them.

>  Just for my knowledge: is there a way to mark a bug to "don't attempt to fix" or withdraw so other will try not to fix them.

Status > Resolved > Wontfix.

[Eric Seidel (no email)]

Comment on attachment 94273 [details]
Proposed patch.

Cleared review? from attachment 94273 [details] so that this bug does not appear in http://webkit.org/pending-review.  If you would like this patch reviewed, please attach it to a new bug (or re-open this bug before marking it for review again).