Re: axiom URIs

On 23 Jan 2007, at 16:12, Pat Hayes wrote:

>> Oh, thinking about this some more reminded me of something.
>>
>> One (alternative) way to serialize axiom annotations would be to  
>> embed them in XMLLiterals. Then it'd be quite easy to have URIs  
>> (or not) for axioms (though this identity would be independent of  
>> the content...yuck!)
>
> Why XMLLiterals?

Going with the most specific appropriate type and trying to stick  
with built in types? Also, I keep hoping RDF toolkits will get better  
at dealing with XMLLiterals :)

> Why not use xsd:string literals?

Eh...then I'd use plain literals.

> Probably they introduce fewer parsing problems; and all that the  
> RDF has to do is not lose them, most of the time, since some  
> external code will have to dig inside them to get the comment out  
> anyway, no?

You misunderstood. I speak of putting the *axiom* in the XMLLiteral  
in RDF/XML form (for example). Thus reifying (in a sense) the axiom  
but also isolating its representation, etc. etc.

Cheers,
Bijan.

Received on Tuesday, 23 January 2007 16:24:32 UTC