[Om3] Being pragmatic about the semantics of, eg, variables and functions

[David Carlisle]

> if MathML is such that that means <condition> is allowed anywhere in
> strict markup since CD verification is optional 

Certainly that is the case: we need to say essentially by schema
(relaxng, or xsd or dtd) what is and is not strict mathml.


> (*) "Rule 2" --- "Note that the order of bound variables..." (4.2.3.2 of
> MML2) in place as the means of tying up the x in the integrand with the x
> (or whatever) in the domain;


more or less, although it's probably more helpful to reference the
current draft's 4.3.3.1

http://monet.nag.co.uk/~dpc/draft-spec/chapter4.html#contm.domainofapplication.qualifier

as that would be the starting point for any wording change.


> (a) This wouldn't preclude OM developing intcond etc. later;

Nothing MathML does constrains OpenMath, unless the OpenMath society
chooses to keep consistent with MathML 3 as a matter of policy.
But currently this is all open, I believe.

> (b) Since Strict will be isomorphic to OM, these will therefore be
> part of strict;

No, after MathML3 is done we will go out of charter and may not have the
charter (or the inclination) to do a MathML4 in the near future. So for
example if (as I think it should) <bind> matches OMBIND and only allows
one child after the qualifiers, and OpenMath later changes OMBIND as in
your proposal to allow more than one, then in order to encode constructs
using the new version of OMBIND in MathML3 you'll have to rewrite them
to use the current encoding. Future changes to OM do not change MathML3.


> (c) Therefore pragmatic->strict COULD be (pace David, I won't say WOULD)
> be enhanced to use these in the future?

Only if we (or someone) recharters the MathML working group to do a
MathML4. 


In response to the above you may be tempted to say that we should leave
the MathML schema for bind open now to allow for any future extensions
but I do not think that is possible. If we say that bind with multiple
children is allowed now, we have to say what it means, say what its
default rendering is and do all the other specification work, in other
word that forces our hand into accepting the proposal for the chaneg at
this stage, we'h have to fully specify it now, in MathML3 and then may
still find that OpenMath rejects this proposal and doesn't extend
binding in this way.


David

________________________________________________________________________
The Numerical Algorithms Group Ltd is a company registered in England
and Wales with company number 1249803. The registered office is:
Wilkinson House, Jordan Hill Road, Oxford OX2 8DR, United Kingdom.

This e-mail has been scanned for all viruses by Star. The service is
powered by MessageLabs. 
________________________________________________________________________