Re: Parsing of HTML fragments

On Apr 24, 2007, at 2:16 PM, Sam Ruby wrote:

>
> Edward O'Connor wrote:
>>> One thing I strongly believe need to be defined clearly is how HTML
>>> fragments, such as those in Atom documents, are to be parsed.
>> If this WG adopts the WHATWG draft, I imagine this would be  
>> handled by the
>> innerHTML parsing mode.
>
> This presumes that all existing markup in existing feeds is  
> retroactively to be interpreted as HTML5?
>
> I would have tended to agree with this, but this seems contrary to  
> the position of some workgroup members that opt-in versioning is  
> mandatory.

Unlike full documents, HTML fragments today don't indicate  
versioning, so they seem to be a hole in any opt-in versioning  
scheme. IRight now you can't even tell if an embedded HTML fragment  
expects standards mode or quirks mode.

I alluded to this a bit with my mention of compound documents. At the  
very least they rule out using a doctype for versioning, but if what  
is included is not the root element (or indeed not an element at all  
but a fragment) then there is no reasonable way to indicate a version  
at all.

Regards,
Maciej

Received on Tuesday, 24 April 2007 21:50:27 UTC