W3C home > Mailing lists > Public > public-qt-comments@w3.org > March 2005

Re: [XQueryX] reuse MathML Content Markup to represent XQuery in XQueryX

From: Frank Yung-Fong Tang <franktang@gmail.com>
Date: Wed, 9 Mar 2005 09:24:57 -0500
Message-ID: <2e4dfd6905030906241ece35dd@mail.gmail.com>
To: Paul Libbrecht <paul@activemath.org>
Cc: public-qt-comments@w3.org, www-math@w3.org

First of all, please do not bring in OpenMath into the discussion. I
never think it is a good idea to mix OpenMath with XQueryX. And brnig
OpenMath in will confuse the discussion.

Second of all, I am asking have any one research about this idea (a
fact), I am not asking about 'is this a good idea'. (an opinion)

Third, the question is when we express XQuery syntax and sematic into
a XML form, as what it should be done in XQueryX, can it be serialized
to a set of mark up that reuse some of the MathML content markup. In
other word, instead of reinvent many tags in XQueryX, can some of the
tags reuse the content markup tags spec out in MathML Content Markup.
This does NOT mean someone XQueryX need to fully  implement MathML
Content Markup because only a subset of MathML Content Markup will be
used to express XQueryX. Also, to express XQueryX you also need to
introduce some other tags because MathML content markup may not
include some of the structure thing that XQuery need.

Just like XQueryX will use XML syntax to store XQuery expression,
XQueryX may be able to use 'some' MathML content markup to store 'some
major part' of the XQuery expression. (notice I said 'Some', not
'all')



On Wed, 9 Mar 2005 10:50:38 +0100, Paul Libbrecht <paul@activemath.org> wrote:
> XQuery is, I suppose, fully implementable.
> I don't think a math-evaluator that would do all is fully implementable.
> 
> And it was never the intent of MathML-content or OpenMath to produce
> something with an implementation for evaluation!
> There's much more that can be done with mathematical objects encoded in
> MathML-content!

Agree. That is exactly why I feel interest about this topic. If MathML
content model can only encode subset of mathematical information than
what XQuery can do , then it won't be interesting. If MathML Content
model can only encode 'exactly the same amount of mathmatical
information' that XQuery can do, then I won't feel interest NEITHER.
Reuse some general mechanism in a special context does not mean your
implementation have to fully implement the general mecanism. For
example, XML is a general mechanism, and MathML use XML syntax to
store math information, but a math processor which know MathML are not
require to fully implement ANY XML language.

> 
> XQuery's evaluations are specified, I think, so there's no such thing
> as undecidability there, I believe.
> 
> paul
> 
> Le 9 mars 05, à 01:01, Frank Yung-Fong Tang a écrit :
> 
> > If such danger exist, then why W3C ever release MathML?
> > If MathML content will easily become unparseable for XQuery
> > interpreter , then in what sense it could be parseable for other
> > application?
> > If MathML content will easily become inefficiently executable for
> > XQuery interpreter , then in what sense it could be efficiently
> > executable for other applications?
> >
> > I am not saying that the issue you concern about is not there. But if
> > they are valid concern for XQuery, then it probably will be valid
> > concern for any other single thing ...
> 
> 


-- 
Frank Yung-Fong Tang
譚永鋒
Šýšţém Årçĥîţéçţ
Received on Wednesday, 9 March 2005 14:25:29 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 27 March 2012 18:14:37 GMT