Early feedback on SHACL Spec appreciated

Hi all,

in the F2F today, I mentioned that the current draft of the Spec

     https://w3c.github.io/data-shapes/data-shapes-core

may now be ready for a first round of feedback. Needless to say it is 
early, has unfinished places (marked in red) and definitions may not be 
precise enough yet for those who love to drill into such details. All 
this needs fixing over the coming year.

Yet, Arnaud indicated that we should aim at producing some First Public 
Working Draft soon, and *we have to start somewhere*, so I am exposing 
the current draft now in its raw form.

Compared to the Primers, I hope I have addressed some issues:
- We now use sh:Shape consistently
- Controversial issues about rdf:type/sh:nodeShape are marked as 
controversial
- SPARQL is now more clearly separated into its own native language 
mechanism
- There is no "OO" or "Modeling" story line - it's just a plain language 
spec.

Regard this document as a "SHACL Full". We should probably also create 
another deliverable for a lighter sub-set of SHACL, especially for those 
who prefer formal abstract definitions of the core language and worry 
about expressivity or complexity in general. This should be perfectly 
doable via sh:Profiles.

I will continue to add bits here and there, this is still work in 
progress. Yet I would appreciate feedback (even on details), so that we 
can determine how far away we are from a FPWD. I will mark sections that 
have received substantial comments in the document (and we can open red 
Issue blocks for the time being).

Many thanks
Holger

Received on Thursday, 26 February 2015 23:36:13 UTC