Re: Reference to the HTML specification

On Fri, 2011-08-05 at 13:41 -0400, Arun Ranganathan wrote:
> On 8/5/11 11:52 AM, Philippe Le Hegaret wrote:
> > On Fri, 2011-08-05 at 17:18 +0200, Marcos Caceres wrote:
> >>> Again, what are the reasons to link to the WHATWG HTML version?
> >> If there is something you need that is not in the W3C spec, then it seems like a valid reason (e.g., PeerConnection API or some helpful concept).
> > Agreed, but no one has come up with such need so far.
> I refer to the HTML WG's work as normative, but in the File API's 
> Editor's Draft [0], I'd also like to link to the WHATWG document as an 
> informative reference for the Stream API [1]  and LocalMediaStream [2].  
> This is a pragmatic, and not a political, cross-referencing.

I have no issue with that. I would simply note that work is now ongoing
in the WebRTC group so you might want to monitor what they're doing.

> Stream API reuses blob: URIs; LocalMediaStream defines globally unique 
> identifiers in a way that I find useful for the opaqueString 
> production.  I'm tidying up normative and informative links, and in 
> general, I think the time is ripe for a good discussion of affiliated 
> specifications.  Another area for coordination that I'd encourage is 
> between W3C and Khronos, if it isn't happening already.  For instance, 
> File API makes use of ArrayBuffer [3] *normatively* which is defined at 
> Khronos [3] and which is implemented in some user agents.  Is there a 
> formal liaison?  This will benefit WebGL as well.

We don't have a formal liaison with Khronos at the moment. In fact, we
don't do formal liaison in general, we prefer to have technical liaisons
instead (ie directly from Working Group to Working Group). Why would we
need to have a formal liaison in order to reference a specification? Are
you aware of [1] ?



Received on Friday, 5 August 2011 18:19:58 UTC