Re: more sloppiness in the SHACL document

(As a general comment, I think it would be most helpful if you could 
include actual suggestions in your emails, in addition to pointing out 
the problems that you see. This would significantly streamline the process.)

On 6/12/2016 12:05, Peter F. Patel-Schneider wrote:
> This fixes one problem but introduces another.
>
> There can be more than one focus node in play at any given time.  (Yes, this
> is itself quite sloppy, but the new wording has brought temporal issues into
> play.)   There might be more than one shape or constraint currently in play
> for any of these focus nodes.

I guess with temporal issues you are now referring to the use of the 
term "when", i.e. that a SHACL processor may at a given time run 
multiple shape validations in parallel?

In 3.4.2.1 we explain how sh:focusNode is populated. In 3.4.2.2 we refer 
to that value to explain sh:resultPath. I have applied the same 
mechanism now and dropped the "when the result was produced" part. If 
this does not meet your standards, please suggest alternative wording 
that would.

https://github.com/w3c/data-shapes/commit/579c7c029642bbfd02a51075070eaf3127d8a129

>
> The part of the document on validation reports has quite a few similar issues.
>   Aside from those that I have already pointed out there is the issue that
> various validation results are supposed to be temporary but the value of
> sh:result MUST include them all.

We had recently made some edits that eliminated many uses of MUST. I 
don't see a relevant MUST in the section on validation results anymore. 
But I guess you are referring to the section on sh:result. I have added 
a crosslink to the paragraph that explains that nested results may be 
excluded (unless as sh:details):

https://github.com/w3c/data-shapes/commit/ef1d8afea1e17c3c2836811317d47e7918cfac6e

Thanks,
Holger


>
> Peter F. Patel-Schneider
> Nuance Communications
>
>
> On 12/05/2016 05:35 PM, Holger Knublauch wrote:
>> I have rewritten that section to implement the convention that constraints and
>> shapes "are" nodes (as you correctly point out) and to no longer exclude
>> bnodes (which was a mistake):
>>
>> https://github.com/w3c/data-shapes/commit/5b6ef796f453b3a9578442e901099fc5c0565eaf
>>
>>
>> Please let me know if this does not address your issue.
>>
>> Thanks,
>> Holger
>>
>>
>> On 5/12/2016 6:15, Peter F. Patel-Schneider wrote:
>>> "Validation results may link to the IRI of the constraint that has caused the
>>> result, specified via the property sh:sourceConstraint, and at the IRI of the
>>> shape that has declared the constraint, via sh:sourceShape."
>>>
>>> However, shapes and constraints don't have IRIs and many shapes and
>>> constraints aren't IRIs.
>>>
>>> Aside from that, the sentence has major semantic issues, e.g.,
>>> sh:sourceConstraint doesn't specify a linking.
>>>
>>> Peter F. Patel-Schneider
>>> Nuance Communications
>>>
>>

Received on Tuesday, 6 December 2016 03:33:23 UTC