Re: Attention Users! (2 in a series)


> DISCUSSION: rdfms-literals-as-resources
> Should literals  be considered a type of resource, possibly 
> "data:" URIs rather than a special case in the model?
This may be interesting from a theoretical point of view. If I understand
that corerectly, it results in identity of indentifier and "content".
In other words, the description identifier IS the resource. (Doesn't this
make literal-resources still different from regular resources?)

However, I see several problems in practical use of this:

Treating literals as resources results in another inconsistency with XML,
which does not treat character data in the same ways as elements. Thus we
have different meaning in triple and XML form, although they are supposed
to be identical.

Literals could become very long, as in dc:Description contents.
Comparison of such data: URIs against other URIs will be very costly.
Also, if you want to use them as subjects, you have to replicate them
for every triple, which is a waste of memory.

It may be usefull to annotate literals, for example to add language
information, but in my opinion this is a process on the description
(meta-information) layer, not on the actual "content" layer. It
should be semantically equivalent to other resources, so the best way
would be to use an anonymous resource, and point at the "content" with
an rdf:value.

Regards, Karsten

Received on Thursday, 12 July 2001 13:35:28 UTC