Re: SVG animation and relationship to SMIL 3.0 Animation

On May 27, 2012, at 7:20 PM, Brian Birtles wrote:

> Hi Dirk,
> Thanks for bringing this up. I tend to agree with the idea of getting 
> rid of redundant examples and text from SVG.
> However:
> * Getting rid of the examples means authors have to do a lot of 
> cross-referencing to write a simple animation. Maybe we can still keep 
> some example(s)?
The chapters that I mentioned don't have examples at all. When we just mention the used attributes and values, and link to the detailed description, we have more space to add more examples that are missing at the moment. Also, I'd like to have smaller examples instead of the big examples that need a whole chapter on their own. Big examples don't help to understand a concept. They just help experts.

> * SMIL Animation is a special self-contained profile which omits 
> features such as time containers etc. Referring to SMIL 3.0 makes things 
> a little more complicated because it's bigger and there's a lot of stuff 
> that doesn't apply to SVG even within the Timing and Synchronization and 
> Animations chapters. Maybe that's ok though?
SMIL always give the host language the possibility to define things. Especially time containers should get denied by the host language according to SMIL. And IIRC we are still missing a thought-out model in SVG 1.1 (didn't look at SVGT 1.2). But this is something that your Web Animation spec will change anyway, correct?

> In practice, implementers still refer to SMIL 3.0 and import updated 
> definitions into the subset outline in SMIL Animation.
That is fine, right now we have a lot of duplication between SVG Animation and SMIL Animation. I just want to remove the duplication, not the SVG specific behavior. That helps implementers that rely on SMIL 3.0.

> Also, if we are to support time containers in SVG 2 and rely on SMIL for 
> the definition, we *have* to refer to something other than SMIL 
> Animation anyway.
That is the part of the Web Animation spec. And Cameron already mentioned on another thread that this will be the time we have to change SVG Animate anyway.

> It's really a matter of timing. Ultimately, we're seeking to replace 
> SVG's dependence on SMIL altogether with Web Animations. I'm not sure 
> how the timing is going to work out, but ideally I'd like to see SVG 2's 
> animation defined in terms of Web Animations only.
I think we are on the same side. I just want to clean up SVG Animate module and not redefine current behavior. Like I said before, SMIL Animation is tolerant about other timing concepts.


> Best regards,
> Brian

Received on Monday, 28 May 2012 03:11:01 UTC