Re: urgent question

Raph Levien wrote:
> 
> Chris Lilley wrote:
> > Both of these pretty much require an XML DOM[1], and most of the viewers
> > are using an XML parser which does provide a W3C-compliant DOM 1
> > implementation, so the foundations are there - it is just a case of
> > adding it.
> >
> > The CSIRO viewer comes with source, and the new version has moved to the
> > IBM XML4J parser which includes a W3C-compliant DOM 1 implementation so,
> > if you feel inspired, you could look at adding this functionality and
> > contributing it back to this project.
> 
> I'll add to this that Gill is very much DOM-centric, using Gdome, the
> Gnome DOM Engine.

Thanks Raph. I had forgotten that Gill is also open source and thus,
anyone interested should also consider helping out with Gill, too.

I was actually looking at the Gill pages again when your mail arrived,
coincidentally enough, since I now have Linux (Mandrake 6.0) and Gnome
installed on my laptop and thus, hopefully, will get to try out Gill
soon. Its the only SVG implementation I haven't tried yet, as far as I
know.

> We've been trying to go even further in embracing these standards, by
> making basing rendering on a Model/View system using DOM 2 events to
> intermediate the events and listeners. However, I've found some
> technical flaws in the current DOM 2 draft that make it not really
> suitable for implementing the Model/View pattern.

Which I hope you have forwarded to the relevant folks (DOM IG) so they
are aware of these problems ...

>  This is one of the
> things that's holding back Gill's rapid progress (the other major one
> being my thesis crunch). My discussion with the DOM working group is on
> the w3c dom mailing list archive.

 ... Aha. Good.

> 
> Nonetheless, draft DOM 2 events are what's being used in Gill, and it
> does use that mechanism for interactive manipulation of the DOM.

Do you use the error and mutation events from DOM 2? When SVG WG first
looked at these we didn't see much application for them in SVG, although
once we started looking azt how DOM-based and concurrent declarative
animations on the same element would interact, it became pretty clear
how applicable they were ;-)

> > in the second case, it is clearly the easiest way for the
> > viewer author to implement the declarative animation support.
> 
> Agreed. Trying to do SYYM animations without the DOM is probably a waste
> of time.

Right.

I wonder what a good authoring support model would be, for applications
that wanted to let folks author the SYMM animations in a WYSIWYG manner
and then write out the SVG?

--
Chris

Received on Wednesday, 22 September 1999 00:37:30 UTC