Re: MathML and SVG in text/html

Hi, Folks-

Along these lines, several of us have just had a conversation on IRC, 
where we discussed a possible extensibility point:
  http://krijnhoetmer.nl/irc-logs/whatwg/20080401#l-441

We hashed out some ideas on the WHATWG wiki:
  http://wiki.whatwg.org/wiki/Diagrams_in_HTML

The main open question is detailing how to handle tokenizer errors and 
tree construction errors.

Note that this is not set in stone by any means, but is a positive 
potential step forward.

Regards-
-Doug Schepers
W3C Team Contact, SVG, CDF, and WebAPI


Doug Schepers wrote (on 3/31/08 9:33 PM):
> 
> Hi, CDF WG-
> 
> The editor of the HTML5 specification is not convinced that MathML and 
> SVG are the right choice for use in HTML.  He apparently considers LaTeX 
> and other formats as equally well suited as MathML [1], and VML and 
> Windows Metafile format as equally well suited as SVG [2].  I don't know 
> if he's genuine in this belief, or if he's merely setting expectations 
> low so as to gain concessions to the markup and features allowed in 
> text/html.
> 
> I am inclined to believe that they are the most suitable formats 
> (particularly SVG, though I find the arguments of the MathML advocates 
> compelling, too); they have been designed from the ground up to be 
> compatible with other Web technologies, specifially (X)HTML (and by 
> extension HTML).  However, he may be right that they do not fit within a 
> vision for HTML which is a monolithic generalized language covering all 
> domains of expression, as opposed to a framework of multiple 
> interlocking languages where each performs a dedicated function with 
> applicable semantics.
> 
> This is the essence of CDF, of course, so as I mentioned at our last 
> F2F, it may be that the best place for this to be specified is the the 
> CDF WG, working closely with the HTML, XHTML, MathML, and SVG WGs.  The 
> CDF WG understands the importance of preserving the original formats of 
> each language, and limits itself to defining the interactions between 
> technologies.  The HTML5 specification need only concern itself with the 
> legacy requirements of the HTML language, and could provide a single 
> point of extensibility, such as an <ext> element or a set of locations 
> or circumstances under which other languages could be inserted.
> 
> Naturally, the goal would still be to have the same DOM serialization in 
> XHTML as in text/html.  It would be a disservice to authors to introduce 
> confusing incompatibilities.  It's my belief that the markup itself 
> should be a close as possible to the original formats as well, for the 
> same reason.
> 
> Given the momentum behind development in HTML in the browsers today, I 
> think this may be the biggest bang for our buck, and I'd like to discuss 
> this in our next telcon.  What do you think?
> 
> (Since we've agreed to act in public when the new charter goes through, 
> I'm sending this to the public list, but also BCCing the CDF, MathML and 
> SVG WG lists to make them aware.  If you prefer to respond on the 
> member-only list, I certainly respect your privacy.)
> 
> 
> [1] http://lists.w3.org/Archives/Public/public-html/2008Mar/0267.html
> [2] http://lists.w3.org/Archives/Public/public-html/2008Mar/0266.html
> 
> Regards-
> -Doug Schepers
> W3C Team Contact, SVG, CDF, and WebAPI
> 
> 

--

Received on Tuesday, 1 April 2008 03:41:36 UTC