Re: Last Call for additional SHACL Tests

Hi Peter,

I have meanwhile merged your recently submitted tests (pending one that 
has a conflict and the one about pre-binding which is under review and 
will be handled by separate emails). While merging your tests I have 
made fixes to your tests because they contained errors (see commit 
history if interested). I have fixed these errors on a good will basis.

We are now clearly beyond the Last Call date. I have accepted the tests 
above because you had originally submitted them before the deadline. 
Unless I hear otherwise from the WG, I will no longer accept any other 
new test cases from anyone. This includes #81 which you have submitted 
today. I have no doubts we could spend another year or two discovering 
more "interesting" test cases. But the WG will by default shut down on 
May 31 and we are quickly out of time. The WG members and the 
implementers need a stable base line to work against. The test suite 
isn't going to be 100% perfect. SHACL isn't going to be 100% perfect. So 
be it. Such is life. Time to move on.

Thanks,
Holger


On 28/04/2017 9:54, Peter F. Patel-Schneider wrote:
> I have submitted the following tests.
>
> Tests of various aspects of pre-binding that needed fixing
> - resulted in change to SHACL
> - failed by at least one implementation
>
> ./sparql/pre-binding/values-2
> ./sparql/pre-binding/minus
> ./sparql/pre-binding/bind
> ./sparql/pre-binding/values
>
> Tests of SHACL Core processors on SHACL-SPARQL constructs
> - resulted in clarification to SHACL
>
> ./sparql/core/sparql-conform
> ./sparql/core/sparql-nonconform
> ./sparql/core/component-nonconform
> ./sparql/core/component-conform
>
> Tests for property paths that share structure or are unused
>
> ./core/path/path-complex-002
> ./core/path/path-unused-001
>
> Tests on nodeshapes for datetime data values, repeated components, allowed
> qualifiedValue parameters
> - test interesting cases for SHACL Core
>
> ./core/node/class-003
> ./core/node/minInclusive-002
> ./core/node/minInclusive-003
> ./core/node/qualified-001
>
> Tests of WG shacl-shacl shapes graph, shared sub-shapes, subclasses of SHACL
> syntax classes
> - test interesting cases for SHACL Core
> - test of large shapes graph created by working group
>
> ./core/complex/shacl-shacl
> ./core/complex/shared-001
> ./core/complex/subclass-of-nodeshape
>
> Test of other value for true
> - test of interesting case for SHACL Core
>
> ./core/property/uniqueLang-002
>
> Test of possibility of sharing results in validation report
> - test of different possibliity for validation report
>
> ./core/validation-reports/shared
>
>
> I don't view them as low quality at all.
>
> peter
>
>
> On 04/25/2017 03:13 PM, Holger Knublauch wrote:
>> Hi Peter,
>>
>> you had submitted a number of pull requests and I have tried to merge them all
>> in. Unfortunately many of the test cases you have submitted had serious
>> problems and overall low quality. Errors included invalid Turtle files (did
>> not even parse), mismatching URIs in tests and data, malformed manifest files,
>> logic errors. I had spent several hours fixing most of them, and notified you
>> about the various errors.
>>
>> Your most recent pull request had again similar errors, see
>>
>> https://github.com/w3c/data-shapes/pull/66/files#diff-68d02e684d9100fc33e1e1ed8c3bc111
>>
>>
>> So I had merged them in, tried to run them, but gave up and reverted the
>> merge. There is only so much detective work that I am willing to make.
>>
>> Feel free to correct the errors and resubmit. Having said this, we are now
>> after the original last call date because we are trying to reach a stable
>> state in which implementers can submit full coverage. Also, note that the WG
>> has complete discretion to approve or reject tests.
>>
>> Holger
>>
>>
>>
>> On 26/04/2017 1:12, Peter F. Patel-Schneider wrote:
>>> I have a bunch of tests that I submitted but were not accepted.
>>>
>>> peter
>>>
>>> On 04/19/2017 07:16 AM, Sandro Hawke wrote:
>>>> The WG is eager to receive people's SHACL test cases, both single-feature unit
>>>> tests, and more complex edge cases and real-world examples.  If you have any
>>>> of these, please send them along.    See
>>>> http://w3c.github.io/data-shapes/data-shapes-test-suite/ for details.
>>>>
>>>> Given the short timeline, the group is going to need to stop accepting
>>>> submissions for this round soon.  If you can, please send in your test cases
>>>> by *Tuesday, 25 April*.   If this deadline is too soon, please let us know
>>>> about your plans, and we'll try to work something out.   Test cases submitted
>>>> later will still be helpful to implementors, but will not be used as part of
>>>> W3C's Recommendation Track process.
>>>>
>>>> Thanks!
>>>>
>>>>         -- Sandro  (on behalf of RDF Data Shapes WG)
>>>>
>>>>
>>

Received on Friday, 28 April 2017 22:22:46 UTC