Re: Deciding in public (Was: SVGWG SVG-in-HTML proposal)

Henri Sivonen wrote:
> ...
> It was developed in a "distributed" (the one party away from others) 
> manner as an extension. It was developed without a blessed framework for 
> doing that kind of thing.
> 
> If there had been a blessed framework, would it have solved the lack of 
> a Path object and the lack of a fallback mechanism from the start?
> ...

Define "blessed framework".

Are you saying that a good extension not only requires peer review in an 
open manner, but that it needs to be run through a *specific* group?

(NIH?)

>> What distributed extensibility gives us is disambiguation. But that 
>> doesn't mean that things won't get peer review.
> 
> Would we be better off if <canvas> had its Apple origin unambiguosly on 
> it for the rest of the existence of the Web? Would the Web platform be 
> better if it were <apple:canvas 
> xmlns:apple="http://www.apple.com/2004/07/namespaces/webkit/dashboard#"> 
> instead of <canvas>?

Actually, yes. It would allow the W3C to now standardize <canvas> 
without having to deal with that legacy.

>> And yes, I talk from experience. For instance, WebDAV uses URI-based 
>> extensibility all over the place, but most extensions I'm aware of 
>> happen within the IETF process anyway.
> 
> 
> What does the URI cruft buy WebDAV if it's mostly within the IETF 
> anyway? Experimental CSS and JS API features seem to be doing fine 
> without URIs and with simple prefixes without an indirection layer.

It's not "cruft".

The point is that although the extensibility model is based on URIs, 
many parties *still* want peer review and go through the IETF. But it 
*allows* parties not do so, and this works fine as well (and indeed 
happens).

BTW: WebDAV is just one example. From my work experience, I could also 
report from XSLT, JCR or xml2rfc if you're interested. Thinking of it, 
EXSLT is an excellent example: when the W3C dropped work on XSLT 1.1, 
implementors worked together to collaborate on much-needed extensions. 
And they could, because XSLT 1.0 got extensibility right.

BR, Julian

Received on Friday, 1 August 2008 15:47:08 UTC