Created attachment 322926 [details]
Test case to illustrate the problem.
Simple test to animate font-size CSS property, which does not yet work in the hg repository for SMIL at http://hg.mozilla.org/users/dholbert_mozilla.com/smil-patches/
See Bug 216462, Comment #58 for more details.
As far as I know the bug 216462 only implements animation of attributes which are lengths.
Thanks, that's good to know. My understanding from discussion with Daniel was that I could create bugs to cover missing functionality.
(In reply to comment #2)
> My understanding from discussion with Daniel was
> that I could create bugs to cover missing functionality.
Yes, please do. (thanks for making this one)
Rather than overloading the "URL" field to point to the hg repository, I'm trying out a whiteboard syntax suggested by timeless, for dealing with hg branches: "[hgurl:c0]", which means "this is a bug on a specific hg repo, whose url is in comment 0".
(alternately, we could just put the whole hg URL in the whiteboard, but hg urls are probably too long for that)
Created attachment 375599 [details]
This test-case animates:
y and rx for rect
r for circle
rx for ellipse
points for polyline
points for polygon
Since that last testcase works perfectly in Opera shouldn't this bug be labeled parity-opera? (It almost works in Safari 4 beta as well.)
(In reply to comment #6)
> Since that last testcase works perfectly in Opera shouldn't this bug be labeled
> parity-opera? (It almost works in Safari 4 beta as well.)
I am not sure, do you suggest there is something in my test case that doesn't follow SVG standard?
Yes, it works in Opera - just like the first test case. However it wasn't written to match Opera's SVG implementation. It's as simply as the fact that Opera just supports SVG really well.
Lars: Yes, it should -- thanks.
Rafał: I think you misunderstood Lars' meaning from comment 6 -- he's simply saying that this bug is for a feature that Opera (and WebKit) already supports, and we should label it as such. (also -- thanks for the great testcase!)
*** Bug 491268 has been marked as a duplicate of this bug. ***
Ouch, I didn't understand Lars. Thanks for explaining.
I reported bug #491268 which is duplicate of this one as I didn't know x and y attributes of text are lists. I'll attach test-case for testing textx elements with single values in x/dx/y/dy.
Created attachment 375790 [details]
One more test-case for text elements.
Nothing special, just animating x/dx/y/dy attributes.
FWIW, all of this bug's testcases now work in mozilla-central nightly builds, except for the last two lines of the attachment 375599 [details]
(for 'points'), which is being tracked in bug 522308.
What about supporting begin="click" or begin="mouseover" in animations? As far as I know, they don't work yet. At least the following example didn't work, while it works in Webkit-based browsers and in Opera.
<circle id="circle" cx="240" cy="135" r="135" fill="black" pointer-events="all">
<set attributeName="r" attributeType="XML" to="240" begin="mouseover" dur="6s"/>
That's bug 485157, whose patches are currently undergoing review.
awesome - sorry my search didn't find it
Closing as there are no more types left.
Can we get rid of nsISVGValue/nsSVGValue and all related code now?
There's nsSVGStringProxyValue left which creates proxy objects for XUL I think for when we have attributes that won't parse e.g. x="?" which is then replaced. I imagine that could be rewritten though and nsISVGValue could go as part of that.
Is that actually still used? It looks to me like it wraps attributes which are themselves represented as nsISVGValues, and we've removed all those.