[whatwg] Re: repetition model

Malcolm Rowe malcolm-what at farside.org.uk
Thu Jun 24 10:12:22 PDT 2004

Jim Ley writes: 

>> That seems to me to be a fair argument that text/html has to either
>> conform to the HTML4 or XHTML 1.0-appendix-C specs.
> I do not believe that was either the intention of the authors or what
> it says (since in other places it illustrates identifying HTML 2.0
> etc.)

I'm not saying that it's perfect (and there are plenty of places where the 
RFC's and W3C recommendations don't tie up - just see the TAG on RFC 
2396bis), but that's the place that canonically defines what content in 
text/html means, and it doesn't say 'and you can put any old rubbish you 
want in'. 

It does say that content marked as text/html is unlikely to be conformant to 
the specs (it even says that you have to be "bug-compatible"), but it 
doesn't give any latitude for including XForms, SVG, Web Forms 2, or 
anything else, and legitimately marking it as text/html. 

>> Agreed. Actually, to whomever has control of the text/html MIME type -
>> I'm not sure if that's actually the W3C or IETF.  Perhaps IETF with W3C's
>> agreement.
> It's the W3C - see the RFC.

It's the W3C who submitted the text/html spec, but it's the IETF who manage 
MIME type registrations. You're probably right though. 

>> There are, of course, exceptions for stuff that's in development,
>> otherwise you'd never be able to develop anything :)
> No, that's what the various x. etc. trees are for. (or prs.hixie or
> vnd.opera :-)

Well, yes, but I don't think that anyone would suggest that if someone were 
to develop HTML5, they'd have to use anything but text/html while they're 
developing it. 

>> But yes, generally, I agree. We need to submit this to a standards
>> organisation after we've got a stable spec, and after we've got two
>> interoperable implementations.
> That seems rather late, seen as the implementations have to be of
> release quality, and they'll be supporting something which is almost
> bound to change before making it through the standards org.  Or do you
> really think they'll just rubber stamp it?  the W3 process doc doesn't
> support such.

I'm not an expert in W3C process, and I'm not sure what the eventual process 
will be, since this is something that's not really been attempted before. 
The PNG spec was submitted to W3C, who essentially rubber-stamped it, but 
that wasn't an extension to something W3C was already involved in. 

People who write UA's to specs that aren't standardised yet do expect to 
make changes while the specs are standardised, since that's what 
implementation experience is all about. On the other hand, if the W3C were 
to standardise WF2, I wouldn't expect them to make many (or any) changes, 
unless there was something really major we missed. 


More information about the whatwg mailing list