RE: Hypermedia - Why

 
> 
> You've mentioned mime types before but I think it's an 
> important point that most XML vocabularies are not defined in 
> terms of a mime type.

That's for sure!  But there are exceptions, like Atom for instance.

> The mime type is useful in some contexts (and MathML defined 
> that one at
> MathML3 but it had happily gone for a decade without). 

XML itself has happily gone on for more than a decade without defining
too many mime types.  But that is a fundamental issue on the web,
because the mime type is what defines and conveys the semantics of the message,
unless I'm mistaken.  On the web that is.  Other mediums would have
other signal mechanisms, I believe.

http://cm.bell-labs.com/cm/ms/what/shannonday/shannon1948.pdf

See paragraph 2 in the introduction.  The paper in general is way beyond me,
but perhaps it will mean something to you.  Also, maybe its out of date.


This is the reference I more or less understand:

http://www.w3.org/2001/tag/doc/mime-respect-20060412#media-type

> One 
> usually distinguishes Docbook from XHTML from TEI not by the 
> mimetype it is served with, but rather as you would 
> distinguish English from German or Fortran from C by looking 
> at the terms used in the document or by some prior agreement 
> of the language of communication. 

I guess this is like the browser sniffing the content to determine what it's 
dealing with.  Apparently that's not a good idea in the long run, see this thread:
sorry its a bit tl;dr.

http://lists.w3.org/Archives/Public/www-tag/2006Aug/0014.html


> That's why it is easy to 
> have an XHTML+MathML+SVG document even though a mime type can 
> only refer to the document as a whole.

Yes, it is all packed into text/html now.  Which is ok, great even, but a decade
ago if you had put all that stuff in text/html, no browser would work on it.

Peter

Received on Wednesday, 25 July 2012 16:00:04 UTC