Re: XMLLiteral in OWL

The proposal by Ian [1], rephrased today by Jeremy [2], to
obsolete the Full version of test 205, indeed seems to
describe the minimal way to solve the problem noted 
in connection with XMLLiteral in OWL.

Three further changes were also proposed [3] in connection 
with this problem, which are purely editorial, and which
make explicit and clarify points which would otherwise 
seem to be very difficult to note for an unguided reader of 
the OWL documents.
In the interest of the clarity of the spec, on behalf
of Philips I hereby repeat the proposal to make
also these three further changes:
- see the first two changes precisely described by Jeremy 
in [4];
- add a sentence just before Section 5.1 of S&AS that in
case XMLLiteral is not in the datatype map, the direct
semantics takes precedence over the RDF-based semantics.

See [3] for more background.
Jeremy [5] wrote as follows about making only the minimal 
change of [1] and [2]:
>I would be happy with Ians suggestions, although I suspect most readers 
>would miss the implicit requirement. There is an explicit requirement to 
>list the supported datatypes which I think is the key defence against 
>surprising interoperability problems.


Herman

[1] http://lists.w3.org/Archives/Public/www-webont-wg/2004Jan/0062.html
[2] http://lists.w3.org/Archives/Public/www-webont-wg/2004Jan/0085.html
[3] http://lists.w3.org/Archives/Public/www-webont-wg/2004Jan/0045.html
[4] http://lists.w3.org/Archives/Public/www-webont-wg/2004Dec/0100.html
[5] http://lists.w3.org/Archives/Public/www-webont-wg/2004Jan/0066.html

Received on Wednesday, 28 January 2004 11:31:28 UTC