<sandro> PatHayes, can you formally define g-box for us?

REST defines a resource as a function from times to representations. So, a g-box is a resource whose representations are (a recognized interchange syntactic form of) g-snaps, ie RDF graphs. 

Will that do? 

That then has the nice feature that it is 'natural' to say that IRIs relate to g-boxes via HTTP GET and PUT, according to the standard Web architectural principles; but it is also natural to say that IRIs relate to graphs by naming, ie by reference. So we have two parallel worlds: 

IRI----HTTP/"identifies" ---- g-box 
IRI----denotes/names-----g-snap

and we can state as a semantic rule that when the first one works, the second one has to conform to it. So if your http IRI GETs something that parses to a graph, then the IRI refers to/denotes that graph. Which all has a nice http-range-14-ish flavor to it :-)

Pat

------------------------------------------------------------
IHMC                                     (850)434 8903 or (650)494 3973   
40 South Alcaniz St.           (850)202 4416   office
Pensacola                            (850)202 4440   fax
FL 32502                              (850)291 0667   mobile
phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes

Received on Thursday, 13 October 2011 04:32:00 UTC