Data spaces draft

Sandro, all,

Topic: Sandro's draft document "RDF Spaces and Datasets"
   http://dvcs.w3.org/hg/rdf/raw-file/default/rdf-spaces/index.html#

First, great work, thanks a bunch for putting all this effort in!

In the light of the recent "making progress on graphs" discussion I've 
reread your draft with the intention to identify parts where we do have 
consensus. This is my first attempt:

SEC. 2: USE CASES

Great that you have re-formulated the use cases in the context of one 
example domain. Have to check whether it addresses all the important 
ones, but on first reading this appears to be the case. [I think it 
would be fair to ask others to work on the actual examples in the 
appendix. ]

SEC. 3: CONCEPTS

For me the subsection 3.1 (spaces), 3.2 (quads and quad sets), 3.3 (data 
sets), 3.4 (named graph), 3.6 (graph store) and 3.7 (union and merge) 
appear to be non-controversial. I suggest text like this should become 
part of RDF Concepts.

Personal note: after reading I felt quite comfortable with the term 
"space". Looks like a decent choice.

Sec. 3.5 (quadset/dataset relationship) is controversial in the sense 
that it interacts with ISSUE-22, as the ongoing debate shows. From my 
perspective this is an important but also quite specific/detailed issue. 
We should resolve it, but I don't think it should block us making 
progress in other areas.

Sec. 3.8 (untrusting merge): I'm not sure I fully grasp what you are 
trying to achieve here and for which use cases it is necessary. I assume 
it requires being explicit about graph equality.

SEC. 4: SEMANTICS

The controversial part here is directly related to the 3.5 
(quadset/dataset relationship) debate. The rest appears to me to be 
non-controversial (by I'm a lay person in the semantics area). I guess 
the real debate will be about what we want to add here (or not).

SEC. 5: DATASET LANGUAGES

I really like the fact that you included concrete syntaxes as part of 
the document. Some of the details of the grammar rules will no doubt 
lead to debates, but this is a useful starting point. As Primer editor 
I'm very happy with this :-).

For what it's worth,
Guus.

Received on Wednesday, 16 May 2012 12:24:58 UTC