- From: Sam Pinkus <sgpinkus@gmail.com>
- Date: Wed, 16 Jul 2014 21:05:12 +1000
- To: Paul Murray <pmurray@anbg.gov.au>
- CC: public-rdf-comments@w3.org
On 07/16/2014 12:38 PM, Paul Murray wrote: > Or just a "a set of triples forms a graph (of nodes and arcs)". > Perhaps simply "is a graph"? No: a set of triples is a set of triples, > it "is" a graph no more than it "is" a file on disk. Seems reasonable to me that a set of triples can be a "graph", a serialization in some format, and a "set of triples" too. My point was it seemed "graph" was the term being used to describe a set of triples but it was never stated explicitly and should be. I still think it is, and if not should be. After all, we have "multiple graphs" (section 3.5) not "multiple sets of triples". > Perhaps "is equivalent to"? There's a mathy word for when you have two > models that have strictly the same underlying from (doing some > operations on A is the same as going from a to B, doing the equivalent > operations there, then going back to A again) - 'homeomorphic'? But > you probably don't want to bother with that term. Your describing an invertible function? But I think the word your after may be isomorphism. Anyway mute point. > Informally, the triples form a graph with connections. Formally, they > may form a graph with disjoint subgraphs. > Good point. Thanks, Sam.
Received on Wednesday, 16 July 2014 11:17:46 UTC