Re: An HTML language specification vs. a browser specification

I agree with what Roy says as well[1]. I think this also suggests some  
ways the spec could be broken out and modularized to great advantage.  
I see several distinct parts of HTML5 that could stand on their own.

1) a markup parsing and serialization specification — with thoroughly  
specified error handling — that could apply as much to SGML (if DTD  
support was added back into it) as it applies to HTML
2) modified HTML language and DOM specification
3) a web browser behavior specification (as Roy called it) including  
the thorough specification of DOM method and attribute processing  
algorithms


All three of these parts could stand on their own. The HTML language  
and DOM specification — as it currently stands — is also the weakest  
part of the draft. It fails to take into consideration the needs of  
users and authors and shows a drastic misunderstanding of HTML  
authoring. IMO, it would be best break the draft out into these three  
separate recommendations. Though there are dependencies here, these  
three have clear boundaries that allow their separation.

  a ) While HTML and its DOM need a parsing and serialization  
specification, it can potentially (and does) support multiple parsing/ 
serialization approaches.
  b) While the browser behavior (and the precise specification of  
processing algorithms) is dependent upon the HTML language and DOM  
specification, it is likewise dependent on a great number of other  
language and DOM specifications.
  c) Conversely, we should not assume that the HTML language and DOM  
will be processed only by a browser.

Take care,
Rob


On Nov 14, 2008, at 12:46 PM, Philip TAYLOR wrote:
>
> Mark Baker wrote:
>
>> I think we've had this discussion before 8-)
>> As often happens, Roy says it better than I could;
>> http://lists.w3.org/Archives/Public/public-html/2007Nov/0430.html
>
> And Roy also summarises my feelings perfectly.
> Philip TAYLOR
>

[1]: <http://lists.w3.org/Archives/Public/public-html/2007Nov/0430.html>

Received on Friday, 14 November 2008 19:33:20 UTC