[html5] HTML5 without namespaces

[J David Eisenberg]

On Wed, Jun 1, 2011 at 7:51 AM, Henri Sivonen <hsivonen at iki.fi> wrote:
> On Wed, 2011-06-01 at 09:44 +0200, Jesper Tverskov wrote:
>> As I have understood it, something more or less similar is under way
>> to support RDFa inside HTML5 served with mimetype "text/html".
>>
>> http://dev.w3.org/html5/rdfa/rdfa-module.html
>>
>> I don't know if RDFa has to become part of the HTML5 spec to be
>> integrated into HTML5 served with mimetype "text/html" or if there is
>> some other way to allow RDFa inside HTML5?
>
> RDFa is not integrated into HTML parsing. The group specifying RDFa uses
> attributes of the form xmlns:foo, but those attributes are not namespace
> declarations in text/html. The group specifying RDFa has been notified
> about this repeatedly. While they have introduced alternative syntax,
> they have refused to get rid of the xmlns:foo syntax even though
> xmlns:foo is represented differently in the DOM in the text/html case
> compared to the application/xhtml+xml case.
>
>> Now my main question:
>>
>> What are our options, the day we would like to support other
>> applications in HTML5, like, let us say "musicScoreML", "VoiceML",
>> etc,  (I'm just making them up, for the argument, we don't care if
>> they exist or if they are relevant to include or not)?
>
> SVG and MathML were pre-existing languages and already implemented in
> (some) browser engines. That's why they stayed in their namespaces. If
> you are designing new vocabulary additions to the Web platform today,
> putting the additions in the (X)HTML namespace avoids the problem of
> having to introduce new namespaces.
>

I don't see that as an entirely satisfactory solution. From reading
http://wiki.whatwg.org/wiki/Namespace_confusion, one might get the
impression that namespaces were introduced solely to create bugs and
mystify everyone. For example, there's a reference to a page where an
O'Reilly author says "I can report that 90% of the technical questions
from readers involve confusion related to namespaces." However, if you
read that entire page
(http://www.w3.org/2004/04/webapps-cdf-ws/papers/verity.html), you'll
see that the author namespaces help to solve the problem of handling
compound documents.

With some mechanism for clearly identifying the parts of compound
documents that have varying document types (and I am *NOT* saying that
it must be namespaces, nor that namespaces are the ideal method), you
solve the problem of collisions between element names. Such a
mechanism would also enable a UA to hand off a section of markup
belonging to some other document type to a plugin that is registered
to parse/interpret/display it.

Simply "putting the additions in the (X)HTML namespace" doesn't seem
to solve the name collision problem, and it appears to make (X)HTML
more monolithic than modular; it is as if there is no longer any such
thing as a compound document; it's all (X)HTML.

> --
> Henri Sivonen
> hsivonen at iki.fi
> http://hsivonen.iki.fi/
>
> _______________________________________________
> Help mailing list
> Help at lists.whatwg.org
> http://lists.whatwg.org/listinfo.cgi/help-whatwg.org
>