Editorial comment/question on the federation query draft...

Eric,

I was looking at the federation extension draft...

1. I am not sure your binding example (section 3) is correct. First of all, you define a binding for the ?human variable, which does not appear anywhere in the query. I would expect that you wanted to use ?species instead. Is that correct?

But even if so, and I look at the data up there at the beginning of section 2, the object for all iuphar:species are resources and not literals and, to add insult to injury:-), they are all blank nodes. Ie, a binding of ?species to the literal "human" seems to be a problem. I actually wonder whether a binding on species is of importance at all. If I look at the example in section 2 again, what you *really* 'transfer' as binding from the first service to the second are really the ?iuphar and the ?id variables; the values of ?species returned from the first service are blank nodes and therefore (I presume) they are locally scoped to each service...

I may be wrong somewhere, though.

2. This is really editorial. The way I read this document is after having looked at the new version of the query document and, well, it was suddenly much tougher to read simply because the examples are much more esoteric. Looking at the data, the queries, comparing with the results in the examples, etc, require to mentally remember terms like ligand, receptor, iuphar (I am not even sure what that word means...), etc, and this is an extra challenge. Compare it with such mundane terms like name, employee, title, price, etc, that are used in the query spec:-) I really believe we should try to rewrite the examples using similar terms...

Cheers

Ivan

----
Ivan Herman, W3C Semantic Web Activity Lead
Home: http://www.w3.org/People/Ivan/
mobile: +31-641044153
PGP Key: http://www.ivan-herman.net/pgpkey.html
FOAF: http://www.ivan-herman.net/foaf.rdf