Re: Support Existing Content

On Thu, 3 May 2007, Jeff Schiller wrote:
> What do the editors think of:
> a) put the HTML5 language in one spec (including the DOM tree):
> audience = content authors, user agent developers
> b) another spec that has guidelines for user agents to process 
> legacy/non-conforming content into an HTML5-compatible DOM tree: 
> audience = user agent developers

There's far, far more to handling errorneous content than just the parsing 
of non-conforming content. Every attribute, every combination of 
attributes, every element, every API, has text for how to handle the 
errorneous case. For example, the spec for the loopEnd DOM attribute has 
to say that when it is set to a value smaller than the loopStart, and that 
definition is right in there with the definition of loopEnd when it is 
_not_ smaller than loopStart.

If the syntax vs parsing was all that you wanted split out, then it 
already exists -- this is for authors:

...and this is for user agents:

Ian Hickson               U+1047E                )\._.,--....,'``.    fL       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'

Received on Thursday, 3 May 2007 22:46:04 UTC