Aligning with the syntax of rdf datasets

[doerthe]

Related to #1

Which of the three options would you prefer:

1. Allow the named graphs syntax in N3, ie say that

<graph> {:a :b :c}
is also directly valid in N3. Then we of course also need to fix the semantics.

2. Add a predicate to allow triple representation of named graphs in N3. Then we would have something like

<graph> :isNameFor {:a :b :c}

In that case it might be possible to add semantics to such triples by providing rule sets for the different existing semantics of named graphs. This of course depends on our solution for issue #8.

3. Ignore named graphs for now and maybe add them later.

What do you think?

[mielvds]

I'm still in favour of

4. Both 1 and 2, where <graph> {:a :b :c} is shorthand syntax for <graph> :isNameFor {:a :b :c} or most common predicate, possibly :isNameFor, if you have one for all possible semantics.

I'd not ignore them because citing graphs is something people would turn to N3 for.

[william-vw]

Is this something we still want to pursue, i.e., shorthand syntax for named graphs in N3, or defer this to a later point? Note that we have some ideas on how to support various named graphs semantics using N3 cited formulas, but distinguishing between them requires a predicate.