W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > November 2001

Re: Proposal to drop S from consideration

From: Graham Klyne <Graham.Klyne@MIMEsweeper.com>
Date: Tue, 27 Nov 2001 09:43:08 +0000
Message-Id: <>
To: Dan Connolly <connolly@w3.org>
Cc: w3c-rdfcore-wg@w3.org
OK, I mis-read your previous comment (I read "number" for "numeral").

As for the rest... I am not yet convinced that the CC/PP style should be 
dismissed so easily.  I would be very surprised if CC/PP is alone in this 
approach.  (After all, I think it was you who asked us to use the XML 
schema datatypes -- if we'd used the original proposal of ccpp:integer, 
etc., we could have resolved the issue by requiring that type to be a 
numeral string and letting the application interpret that as denoting an 

My purpose in saying this is not a game of point-scoring, but to indicate 
that I believe that the usage <age>10</age> to indicate a value that will 
ultimately be interpreted as an integer and described by reference to the 
XML schema integer datatype is one that I think many people will expect, 
and to prohibit such usage would lead to confusion and 
non-interoperability.  Given the extent of support for the S proposal 
--which I'm not (yet) opposing-- I'd like to see more consideration given 
to this issue.


At 11:39 AM 11/26/01 -0600, Dan Connolly wrote:
>Graham Klyne wrote:
> >
> > At 08:36 AM 11/26/01 -0600, Dan Connolly wrote:
> > >Note that it'sn not the XML parser that type converts "40", but
> > >the application that knows about <age>. The analagous situation
> > >in RDF is: the object of <age> is a string, and the range
> > >of the age property is a numeral, not a number. [this
> > >is the case in S]
> >
> > Er, I'm missing something here.  I thought it was precisely this that the S
> > proposal does not allow.  Indeed, I thought the only proposals to allow
> > this are P/P++.
>The P/P++ proposals are about writing <age>10</age>
>and having the value be a number -- the 10th integer --
>not a numeral -- the two character string '1' followed by '0'.
>In S, all* literals denote strings. The only way to
>express a number is ala "the number whose decimal
>representation is '10'".
>In CC/PP-as-written, the instance data says <age>10</age>
>and the schema says
>         age range integer.
>The P/P++ proposal makes sense of CC/PP-as-written-style instances
>and schemas (at an unacceptable cost, to me: the
>denotation of literals is all mucked up).
>To use S, you'd either have to change the CC/PP-style schemas
>to say
>         age range integerNumeral
>to match <age>10</age> or
>or change the instance to say
>         <age dt:integer="10"/>.
>to match age range integer.
>* all except the parseType="Literal" ones, which
>denote structured XML thingies, to me; but that's
>another issue altogether.
>Dan Connolly, W3C http://www.w3.org/People/Connolly/

Graham Klyne                    MIMEsweeper Group
Strategic Research              <http://www.mimesweeper.com>
      /\ \
     /  \ \
    / /\ \ \
   / / /\ \ \
  / / /__\_\ \
/ / /________\
Received on Tuesday, 27 November 2001 05:01:12 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 20:24:06 UTC