W3C home > Mailing lists > Public > www-rdf-interest@w3.org > December 2001

Re: Cutting the Patrician datatype knot

From: Jonathan Borden <jborden@mediaone.net>
Date: Tue, 4 Dec 2001 08:22:37 -0500
Message-ID: <01a201c17cc6$c21fb030$0a2e249b@nemc.org>
To: "Peter F. Patel-Schneider" <pfps@research.bell-labs.com>, "Pat Hayes" <phayes@ai.uwf.edu>
Cc: <www-rdf-interest@w3.org>, <joint-committee@daml.org>
Pat Hayes wrote:

> >
> >[...]
> >
> >>  OK, but now what is the problem with my MT extension, again? Of
> >>  course, with THIS sense of "union", one cannot treat [integer union
> >>  string] as in any sense the simple class-union of the classes
> >>  [integer] and [string]. So as far as the RDF reasoner is concerned,
> >>  [integer union string] is just another datatype, which might as well
> >>  be called [foodle]. If someone were to assert that
> >>
> >>  [integer] rdfs:subClassOf [integer union string] .
> >>
> >>  that would be correct, but that would cause no problem in the MT,
> >>  since [integer union string] agrees with [integer] on numerals; they
> >>  have the same lexical-to-value mapping on anything that would map to
> >>  an integer. Similarly, it would be correct to assert
> >>
> >>  [string] rdfs:subClassOf [string union integer] .
> >>
> >>  though in that case it seems rather pointless since in this case the
> >>  datatypes are identical, since even numerals will be mapped as
> >>  strings by that union. But again, this poses no problems for the MT.
> >>  On the other hand, if someone were to assert
> >>
> >>  [string] rdfs:subClassOf [integer union string] .
> >>
> >>  then that would be simply false, as all numeral strings are in the
> >>  former value space but not the latter.
> >
> >I don't see how this follows.  The value space of the XML Schema datatype
> >[integer union string] is precisely the union of the value spaces of
> >integer and string.
> Yes, Ive come to see that. I simply did not conceive that anything
> could be this badly defined. This aspect of XML datatypes really is a
> crock. It *defines* the domain and range of the datatyping functions
> so that they cannot possibly be the real domains and ranges. What a
> bloody silly thing to do, I'm amazed that Thompson put his name on it.

It should be noted that:

a) Henry was not the editor of the XML Schema datatypes spec
b). I can't say that I've thought it through in extruciating detail but my
gut instinct would be, given that XML is ultimately a _text_ format, to
derive all types, including simple types, not from the _value space_ but
from the _lexical space_. (this distinction is made in XML Schema
datatypes). The reason to have a distinct value space is to provide binary
datatypes (e.g. integers that fit into 32 bits) and to ease the definition
of comparison operators that operate on numbers (as opposed to patterns of
characters). It seems to me that one could (simply?) define number
comparison operators (it seems that mathematicians had gotten along
reasonably well before the advent of 32 bit integers and 64 bit floating
point math coprocessors), in the same way that one defines lexical
comparison operators for such things as string sorting, ditto for dates etc.
c) simple types such as "integer" "float" etc have been defined long ago and
will be defined in the future (i.e. this is not the sole domain of XML
Schema datatypes).
d) there are (currently) _serious_ problems assigning URIs to general XML
Schema datatypes. The URIs given for the simple XML Schema datatypes have
been explicitly created. The XML Schema formalism (WD) proposes a new URI
syntax to resolve these issues, but this syntax is not comparible with other
XML fragment identifier syntax proposals including XPointer. Sigh.

> So, my argument does not hold if we are obliged to conform to XML
> datatyping rules. I am inclined to just give up and leave this
> decision to others, at this point. There is no point trying to be
> rational when one is obliged by mandate to conform to irrationality.
> Let me know what y'all decide.

I agree but should mention that these are _XML Schema 1.0_ datatyping rules,
not _XML_ datatyping rules (that is to say that many in the greater XML
community feel that the last word on _XML_ datatypes has yet to be written).
XML 1.0 itself has a barebones discussion of element types (which are merely
element tag names) and attribute types (e.g. CDATA,ID,IDREF and tokens). My
suspicion is that if this group attempts to adopt XML Schema datatypes en
bloc, that such problems will continue to plague these discussions for a
long time.

Received on Tuesday, 4 December 2001 08:23:07 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:52 GMT