Re: Formal query about WG role and MathML-FAQ

>> can be achieved via XPath axes. You know very well that this content-like
>> syntax can be "easily" parsed to presentation one.
>
>I'm sorry but I really don't see what you are trying to achieve here. If
>I need a parser capable of parsing <textnode>a+b</textnode> then I may
>as well have a full text based system that uses such a parser, asciimath
>or TeX or some simplified subset of tex (or maple or mathematica, or any
>one of dozens of similar alternatives). I honestly see no benefit in
>having some of the expression marked up as XML and some not. If you like
>that style, then fine. This list is even a suitable forum to discuss it as
>it is for discussion of any math on the web, not just mathml, but
>whenever you post to a public list you just get comments or not
>depending on whether any other readers choose to comment, and personally
>I haven't really got anything to say about such a format.
>
>
>David

That is, it appears I am being trained that there is not room for
improvement. Apparently, my ideas are not good enough (curiously other
people think some points i did reserve further research) but I am
receiving none serious alternative for problems I am delineating.

Let me _virtually_ to abandon the CanonMath project. In short, I am being
"forced" to use:

- A full text based system that uses asciimath for students but cannot be
correctly extended due to graphical syntax and parser options and that is
not semantically oriented.

- TeX or some simplified subset/dialect of tex is not semantically
oriented for mathematicians and physicists; it offers *problems* for
transformation to MathML (There is experience on that). Do not work for
chemistry and others disciplines (therein needs for CML). It is difficult
to write/read [Robert Mayans. The Future of Mathematical Text: A Proposal
for a New Internet Hypertext for Mathematics. Journal of Digital
Information, Volume 5 Issue 1 Article No. 234, 2004-05-27], etc.

- Use maple or mathematica when preparing online computational documents.
Since neither ASCIIMath/TeX or MathML are designed for that.

- Work directly the MathML (double) syntax for servering on the web by
hand. Which is highly verbose and time consuming. To server some as
<mfrac> for presentation but instead some like <divide/> for content,
since computers are programmed to ignore that semantics for <mfract> is
just the same that for <divide/>.

- Since MathML is not designed for input. Use some tools. Tools do not
work well for semantics and that cannot be used for science and have
offered to me many problems. I have used a lot of tools, conversors,
programs, etc!

- Use tricks for displaying certain MathML contents, for example,
reactions containing images that currently I could not offer via SVG could
be not introduced into MathML due to absence of an image tag. Therefore, I
forced visual presentation, which is just a trick may be avoided in the
short run.

- Since that above is not sufficient use also at least a dozen of others
languages (use at least four languages for science: CML, CellML, STTL,
ThermoML, etc.), additional browsers, additional specialized tools.

> I honestly see no benefit in
>having some of the expression marked up as XML and some not.

If this is true, then some like

<doc>
<para>Next is TeX </para>
<TeX>1 \over {x+1}^2</TeX>
<para>This is other text
<br/>
This is other text </para>
</doc>

which one finds in several programs would also be ignored since one is
mixing XML markup with non-XML one and pure content with mixed one.

Maybe my approach is a pure nonsense. I do not know really, but at least
people who has worked in MathML specification could have answered some
pure questions. For example, it is interesting nobody has said nothing
about my criticism to the need for a double markup and possibility to
unify presentation and semantics in a more natural way.

Moreover, let me finalize clarifying once more that I was not troubled
with lacking of constructive comments, just by absence of reply to a
formal query. Someone like "I am not interested in your approach" or "I
have not time" -or similar- is very different from no reply at all.


Juan R.

Center for CANONICAL |SCIENCE)

Received on Tuesday, 14 March 2006 19:08:59 UTC