Re: DAML+OIL (March 2001), CVS

From: Richard Fikes (fikes@ksl.stanford.edu)
Date: 03/05/01


> I just read over the axiomatization and I am rather disappointed.  I was
> hoping that there would be at least the beginnngs of a treatment of some
> issues having to do with XML Schema datatype.  However no mention of XML
> can be found in the document at all.  In fact, there is no treatment of
> datatypes at all!

Point taken.  We only had a few hours last week to devote to updating
the axioms and wanted to make sure that an updated version was ready to
include with the new language release.  So, we opted to make the changes
in the axioms that corresponded to the changes in the DAML+OIL spec.  I
think we did that and we will fix right away any bugs in the axioms that
are deviations from the new spec.  Axiomatizing XML data types needs to
be done and is on our "to do" list.  However, I have not yet looked at
the XML data types spec and so have no idea how difficult or well
defined that task is.  Comments on that issue are welcomed.  Peter, did
you specify a denotational semantics for the XML data types?

> There are also a number of problems with the separation into two
> sub-domains.  Most of these are minor (things like union and oneof), but
> the treatment of complement is very different from the treatment in the
> denotational semantics, and will break lots of things.

I will take a look at the differences regarding "complement".

Richard


This archive was generated by hypermail 2.1.4 : 04/02/02 EST