W3C home > Mailing lists > Public > public-html@w3.org > August 2009

Re: HTML 4.02

From: Robert J Burns <rob@robburns.com>
Date: Sun, 23 Aug 2009 15:28:26 -0500
Message-Id: <56C28547-F81F-43D6-8D4F-57F90F57E40A@robburns.com>
To: "public-html@w3.org WG" <public-html@w3.org>
Cc: tai@g5n.co.uk
Hi Toby,

The HTML 4.02 specification is here:

	http://doctype.be/402.html

It's 13 lines long, including the heading.

Features:

	* RDFa
	* ARIA
	* Ruby
	* a few other nice bits from other (X)HTML specs
	* DTD validation (though doesn't cope with xmlns:*)

Any valid HTML 4.01 Strict document can become an HTML 4.02 document
by just changing the doctype.

What do people think?
I think what you're doing here is precisely the solution we need to  
this WG. I have long been involved with the WG, but have recently  
dropped out when I concluded the W3C has ceded the development of HTML  
to a cartel of browser vendors with no interest in the needs of  
authors and users (not even their own users).

I think the effort you've begun with HTML 4.02 is just what authors  
and users need. Ian Hickson and the WhatWG have shown they have no  
interest in the needs of authors and users but only the needs of the  
anointed browser vendors (which they have taken great pains to control  
in turn by getting their members involved with undermining the various  
browser development projects).

So I think authors and users need to take back HTML for authors and  
users. Ian is going to define HTML5 UA conformance without regard to  
the needs of authors and users, but in a way where the features  
authors and users need are still there for legacy reesons. However,  
this legacy support means we can then define an HTML specification   
like you have done with HTML 4.02  that give authors a conformance  
target to author against (ideally with conformance checking based on  
machine DTD, RelaxNG, XSD as well as customized conformance checking).

HTML4All had made some efforts in this direction, starting with a  
prose describing to authors how to author such HTML (eventually hoping  
to get to a DTD and other schema as you have done). Much of what we  
were looking at is included in in your HTML 4.02 with some other  
features that degrade gracefully in current browsers and some  
additional features that would require UA enhancement to work at all.  
Unfortunately HTML4all has had some attacks on its wiki and so the  
bulk of the work is currently unavailable for viewing. However some of  
it is still viewable at: <http://html4all.org/wiki/index.php/Parsing_tHTML 
 >. The overall plan with the HTML4all effort was to do much as you  
have done: combine some of the best of HTML5, XHTML2, XForms, Ruby,  
ARIA, and new enhancements into one specification that addressed the  
most pressing needs of the web community. This included separating  
HTML into separate specification modules: 1) serialization; 2)  
parsing; 3) namespacing; 4) the HTML infoset; 5) HTML DOM interface;  
and 6) presentation. Among other things this allows authors to focus  
on a simplified serialization model and the HTML infoset vocabulary  
while insulating authors from the intricacies of parsing (which only  
UA parser developers need consider).

Thank you for engaging in this effort. I think it will prove very  
useful to the entire web community. Eventually we will have  
conformance checkers that genuinely meet the needs of authors rather  
than the narrow misguided partisan interests of the WhatWG.

Incidentally, I think what you have included in HTML 4.02 is probably  
just what the doctor ordered. I'll mull this over further to see if  
there's anything else that might be appropriate for such an  
incremental release. As for the xmlns:* issue I think there is a  
greater need to make DTD based conformance checkers namespace aware  
than to enhance the DTD expressibility to meet that issue. In that way  
a namespace aware DTD based conformance checker can easily discern  
between the element and attribute names in the namespace the  
conformance checker is currently directed against (e.g., HTML in this  
case) and ignore the elements and attributes beyond its DTD-provided- 
knowledge (e.g., xmlns namespaced elements and attributes). So in that  
respect nothing else is needed from your DTD but some enhancements are  
needed to DTD-based (and other scheme based) conformance checkers to  
make them namespace aware.

Take care,
Rob
Received on Sunday, 23 August 2009 20:29:44 UTC

This archive was generated by hypermail 2.3.1 : Friday, 10 October 2014 16:24:51 UTC