Re: CURIEs (was Re: Several minor problems in the grammar for the functional-style syntax)

On 23 Mar 2009, at 09:59, Ian Horrocks wrote:

> It seems obvious that we should report our difficulties to the  
> CURIE guys.

I can draft an email.

> It also seems obvious that we can't risk waiting for them to fix  
> the problem and that we thus won't be able to rely on the CURIE  
> spec. But I'm not convinced that developing a bespoke abbreviation  
> mechanism is the right way to go. What about Peter's "just go back  
> to QNAMES" proposal?

That *is* a bespoke mechanism. Any abbreviation mechanism needs:

1) A syntax
	for declarations @prefix in turtle, "Namespace: rdf <" in Manchester  
syntax, etc.,  and references.
	This includes legal characters for prefixes and "local names".  
Peter's proposal is just one (common) bespoke mechanism.
2) A processing model
	How to expand, what do do with multiple declarations
	I think our current specs underdescribe this...at least it wasn't  
obvious to me what happens in MS when you have:

Namespace: foo <http://www.ex.org/1>
Namespace: foo <http://www.ex.org/2>

So, since we're in this game anyway, let's 1) do a proper job and 2)  
do it consistently across serializations.

I think we were subconsciously relying too much on the (future) CURIE  
spec. It's pretty clear we can't so rely on it.

Cheers,
Bijan.