[webkit-reviews] review denied: [Bug 10405] REPRO: SVG Maze crashes in RenderPath code : [Attachment 10142] First attempt

bugzilla-request-daemon at opendarwin.org bugzilla-request-daemon at opendarwin.org
Mon Aug 21 08:57:45 PDT 2006


Eric Seidel <macdome at opendarwin.org> has denied Eric Seidel
<macdome at opendarwin.org>'s request for review:
Bug 10405: REPRO: SVG Maze crashes in RenderPath code
http://bugzilla.opendarwin.org/show_bug.cgi?id=10405

Attachment 10142: First attempt
http://bugzilla.opendarwin.org/attachment.cgi?id=10142&action=edit

------- Additional Comments from Eric Seidel <macdome at opendarwin.org>
There are a couple things which I would change here.

1.  The test case should ideally draw a green square behind the red one, so
that when the red one is cleared by the javascript, the green square shows
through as an indicator of success.

Second, I 'm not sure I agree with your strokeBoundingBox changes.

IMO it would be cleaner to have a one line if check at the very beginning:
if (isEmpty())
    return CGEmptyRect;

They way you've written it now, it's nondeterministic as to what sort of rect
you are returning.   Also, an if with an early return often reads a bit
cleaner.

Otherwise looks great!

Given that the rect you are returning is non-deterministic (and I'm not sure if
you added your copyright to KCanvasPathQuartz or not) I'm going to have to r-
this.



More information about the webkit-reviews mailing list