293221 - CanvasRenderingContext2D has wrong default property values

Status: VERIFIED FIXED

(Core :: Graphics: Canvas2D, defect)

Description

[Jens Bannmann]

Some of CanvasRenderingContext2D's properties have wrong values according to the
spec at http://whatwg.org/specs/web-apps/current-work/#canvas . One property,
lineCap, even throws an exception upon reading.

Comment 1

[Jens Bannmann]

Attachment: test case showing actual and expected property values

Comment 2

[Jens Bannmann]

Note: The test case includes all properties, not only the ones with wrong values.

Comment 3

[Vladimir Vukicevic [:vlad] [:vladv] (needinfo me, slow to respond)]

Fixed with patch in bug 294400 - note that one of the values (the composite
operator) will still be wrong, because the cairo names are a little strange and
not inline with canvas/quartz usage.  That gets fixed once we move to a more
recent version of cairo.

Another one of these values has the correct default, it's just incorrect in the
spec :)

Comment 4

[Vladimir Vukicevic [:vlad] [:vladv] (needinfo me, slow to respond)]

fixed by checkin for bug 29440

Comment 5

[Anne (:annevk)]

Which one is incorrect in the specification? Would be nice if that specification
was fixed so we actually have clue against what we can develop test cases and such.

Comment 6

[Hixie (not reading bugmail)]

Vlad and I had a little talk. Nothing to see here. Move along now.

Comment 7

[Vladimir Vukicevic [:vlad] [:vladv] (needinfo me, slow to respond)]

(In reply to comment #5)
> Which one is incorrect in the specification? Would be nice if that specification
> was fixed so we actually have clue against what we can develop test cases and
such.

Sorry, is a bug in the testcase ;)  The default composite op specified by the
spec is "source-over", not "over".

There's a discrepancy between our impl and safari (they have "source-over", we
have "over" as the default, but they behave the same) -- cairo names these
things very differently at the moment, and that won't be fixed until we move to
a CVS version of cairo.  Once that happens we'll nail down exactly which
operators correspond to which. (I haven't spent time doing that up to now, since
we'll be moving to a new cairo as soon as 1.8b2 goes out the door.)

Comment 8

[Anne (:annevk)]

Thanks.