[om] XML Namespace for OpenMath

David Carlisle davidc at nag.co.uk
Wed Jan 26 11:04:37 CET 2000

I suggested

Andreas suggested
 - how about openmath.org/OpenMath/XML/Namespace/

I don't think that either `xml' or `namespace' belong in the namespace
uri as xml is misleading, the `namespace is more or less the abstract
collection of openmath constructs, so if you did want another non xml
encoding, it would probably want to recycle the same
uri not divided up by markup language.

/Namespace is a possibility. It does keep the top level /OpenMath
URI free, but as I mentioned there is no obligation to put anything
at that position so it doesn't in principle block any future development
of the web site although in practice you might want to put something at
the namespace uri, if only `this is a namespace uri, you are not
supposed to link to it..'

In some ways OM is so verbose that it doesn't matter but
if you add the characters


on to every om object it all adds up.

>  - how about including a version number in the path

No! It is not at all unlikely that at some point one wants to have an
extended system, but you should only change the namespace URI if the
system is so different that you want to change the name of _every_
construct and ensure that no transform of query from or two the old
system `accidentally' works with the new system.

If the system is close enough to existing openmath that you still
want existing queries to succeed or want existing transforms
of openmath OMA to mathml apply to keep working then the namespace has
to be the same.

Bruce asked

> If anything were eventually needed at that URI, what would it be?

You don't want to know the lengths people have argued about that:-)
Currently the official line is: nothing is needed, but best practice says
to put an html file there saying that it isn't needed.

> How are later versions coped with?  Why /OpenMath ? (vs. /DTD 
> vs. /Namespace, or whatever) Or is this naming already conventional?

Well for comparison the mathml one is


and XHTML is


Don't be confused by the `1998' and `1999' they are not dates they are
just the w3c's attempts to confuse, and for example MathML2 due this
year will use the same http://www.w3.org/1998/Math/MathML namespace.

so ignoring the w3c's current top level `date-like' fixation the basic
idea of the namespace is that it `name' the language (not its particular
encoding or validation schema) which is why /OpenMath not /DTD, a w3c
XML Schema (or schematron schema) for validating OpenMath would use the
same namespace as a DTD based validation tool.


om at openmath.org  -  general discussion on OpenMath
Post public announcements to om-announce at openmath.org
Automatic list maintenance software at majordomo at openmath.org
Mail om-owner at openmath.org for assistance with any problems

More information about the Om mailing list