Re: SHACL is a Candidate Recommendation

For anyone working on implementations:

The Test Suite is work in progress but I believe the tests that are 
already there do have quite a good coverage. So please take a look and 
feel free to submit (intermediate) implementation reports using the EARL 
format as outlined in the Test Suite document:

     http://w3c.github.io/data-shapes/data-shapes-test-suite/

Note that this includes engines that only return true/false conformance 
instead of the full report - these would still be of interest.

If you feel that certain test cases are missing, please submit or 
request them.

In order for SHACL to proceed to the final stages of the W3C process, we 
need to demonstrate that each feature of SHACL has been implemented by 
at least two independent implementations. No single implementation must 
support all features (although TopBraid's SHACL API does AFAIK).

Please contribute where you can for these important next steps.

Thanks,
Holger


On 12/04/2017 0:51, Sandro Hawke wrote:
>>
>>
>>   Shapes Constraint Language (SHACL)
>>
>>
>>     W3C Candidate Recommendation 11 April 2017
>>
>> This version:
>>     https://www.w3.org/TR/2017/CR-shacl-20170411/
>> Latest published version:
>>     https://www.w3.org/TR/shacl/
>>
>
> Now the real work starts.    :-)
>
> (Seriously, now we need to finish up the test suite and get some 
> interoperable implementations, as soon as possible.)
>
> On Peter's formal objections, my summary of the Director's decision 
> (apologies for oversimplifying):
>
>> 1. Node shapes and property shapes have different features. Details at
>> https://www.w3.org/2014/data-shapes/wiki/CRFO1.
>> On this, the WG took some steps to address the Objection, but stopped
>> short of satisfying the objector.  The Director considers the remaining
>> substantive issue to be support for Generalized RDF, which he believes
>> is important.  As such, he is requiring the Working Group to avoid
>> unnecessarily obstructing the use of literals as subjects.  To allow for
>> this change to be properly investigated, certain syntactic restrictions
>> have been marked At Risk and may be removed during CR.
>>
>> 2. Require a mode that checks if a shapes graph is ill-formed. Details
>> athttps://www.w3.org/2014/data-shapes/wiki/CRFO2.
>> On this, the Director agrees that syntax checking of shapes graphs is
>> important, but not that it should be a mandatory feature.  As a
>> middle-ground solution, the Director required the group to provide a
>> shapes graph which can be used to check shapes graphs, where practical
>> within the language. This is now included as a normative appendix in the
>> specification and will be part of the test suite.
>>
>> 3. Pre-binding issues are not sufficiently resolved.  Details at
>> https://www.w3.org/2014/data-shapes/wiki/CRFO3.
>> On this, given that relevant part of the specification is marked At
>> Risk, the Director has decided to overrule the objection.  The Director
>> hopes the relevant questions can be sufficiently settled by
>> implementation experience during CR.  Failing that, the features may be
>> moved to a non-Rec-track document.
>
> Not perfect, but still good work, folks!
>
>       -- Sandro