Re: Reservations about <mchar>

Hi.

I wanted to add a few comments to the <mchar> vs entities debate.  Part of the
reason there is controversy here is that two important technologies
are pulling in opposite directions -- Schemas and DTDs.  Regardless of
what we do with MathML2.0, this will be an issue for people using
MathML in both contexts.

There are good reasons to add some schema-compatible mechanism.
First, if we don't, we will likely run into trouble at W3C -- it is
after all a consortium, and lots of important players are pushing
schemas.  Also, if we don't, people will start to come up with
something on their own, and there will be no consistency, which is a
disaster from the standards biz view point.

Hence, it seems fairly inescapable that we need to make *some* statement
about how to be schema-compatible.  Among the candidates, I like the
<mchar> solution as well as any, and better than most.

So to me the question is what to do about the status of the 1.0 entities.  The 
alternatives are:

   a) leave in entities and leave out <mchar>
   b) leave in entities and also have <mchar>
   c) we can have <mchar> and deprecate entities

As I said, I believe a) is neither wise nor politically feasible.  Between
b) and c) I have a slight preference for c).

Note that since MathML in XHTML is served as XML, it is possible to do
simple XSL processing to change <mchar>s to entities, and vice versa.
Since user agents are already beginning to use XSL to translate
complicated things like MathMl content markup so <mchar>s ought to seem
easy by comparison.

Finally, as Neil said, you are free to choose between MathML 1.0 and
2.0.  If you chose 1.0, you get entities, but will have trouble in
Schema environments. However there aren't any schemas at the moment,
and I expect most software to continue being biased to the 1.0
construction if/until schemas become important.

Just my two cents.


--Robert

----------------------------------------------------------------
Robert Miner                          http://www.webeq.com
Geometry Technologies, Inc.           email: rminer@geomtech.com
                                      phone: 651-223-2884
----------------------------------------------------------------

Received on Thursday, 4 May 2000 01:13:51 UTC