Re: [TTL] Differences between SPARQL and Turtle.

On 24/04/11 05:51, Ivan Herman wrote:
> To be more general: is there any argument in any of the cases against
> _not_ to align Turtle on SPARQL? It strikes me that the SPARQL
> grammar has undergone quite some scrutiny at its definition, and the
> feedback on that grammar (eg, employee:12345) are pretty relevant for
> Turtle, too. The differences seem to be mostly minor, I do not see
> cases that would seriously break existing and deployed RDF serialized
> in Turtle...
>
> My proposal would be to issue a Turtle FPWD with a complete alignment
> on SPARQL and see if the community would come up with "breaking"
> points...
>
> (Would be good to declare victory on that one:-)

+1

Highlighting the matter with a note (big red box) in the Turtle FPWD 
stating an intention to align Turtle with SPARQL would be good.  Asking 
the community for compelling reasons why not would push the issue along; 
and maybe list the areas of known incompatibility.

If anything comes up, I think we should take the hit, and fix SPARQL 
1.1. (apologies to the chairs of SPARQL-WG!).

What I think is a bad outcome is to address each difference, one-by-one 
in isolation.  A way forward would be for RDF-WG to resolve to align to 
SPARQL unless there are compelling reasons why not.

 Andy

PS Of the differences,

   (1 2 3 4) .

should probably be removed from SPARQL.

Received on Sunday, 24 April 2011 16:50:25 UTC