Re: shapes-ISSUE-33 (Simonstey): Shifting section "Shape Selection" to introduction? [SHACL Spec]

On 4/2/2015 16:17, Simon Steyskal wrote:
> The creation of this issue was actually motivated by many requests I 
> received from colleagues of mine who had a look at the specification.
> Besides the general consensus of SHACL being a valuable addition to 
> the SW stack, they were primarily worried to what extent they have to 
> adapt their already existing RDF data in order to support SHACL. 

Yep, that's an important concern for our customers too.

> So they were skimming through the document looking for some actual 
> code examples and before they got to section 11.1 they bothered me ;)

I have rather recently (last week?) added a worked out example (1) to 
the Introduction, see

http://w3c.github.io/data-shapes/shacl/#introduction-overview

I intentionally designed it such that it defines both an rdfs:Class and 
a sh:Shape, to make sure that both class and shape fan clubs have 
something to cheer about. I was hoping that by using a Class I would 
convince newcomers that it is only a modest step from an RDFS/OWL model 
to a SHACL model.

If you then look at Example 2, you can see that rdf:type triples are 
used to link instances with their shapes. The text reads

"The following example code defines two instances of the 
class|ex:Issue|. When constraint validation is started, a SHACL engine 
may follow the|rdf:type|link of these instances to determine which 
shapes need to be used. (If|ex:Issue|were not a class but a|sh:Shape|, 
then the property|sh:nodeShape|would have been used instead)."

Given that these examples are quite new, I am wondering whether the 
feedback from your colleagues predates this revision, and we have it 
already addressed?

(And sorry I keep changing the document quite frequently, so it's 
difficult to keep up with such changes).

Cheers,
Holger


>
>> So my question is, what is missing to make this clearer? Could you
>> propose alternative prose for this aspect?
>
> I would just append the examples of 11.1.1 and 11.1.2 (or at least one 
> of them and refer to 11.1 for a more detailed discussion) to the 
> paragraph you cited.
>
> But that's just my 2 cents.
>
> cheers,
> simon
>
>
> ---
> DDipl.-Ing. Simon Steyskal
> Institute for Information Business, WU Vienna
>
> www: http://www.steyskal.info/  twitter: @simonsteys
>
> Am 2015-04-02 02:26, schrieb Holger Knublauch:
>> Hi Simon,
>>
>>  I have so far not emphasized the aspect of shape selection because it
>> was a controversial topic (you certainly remember the Class-vs-Shape
>> discussion which is still not completely resolved). In our brief
>> collaboration on the primer, Eric was very keen on introducing Shapes
>> without relying on either selection mechanism - because the validation
>> could also be triggered by custom mappings outside of the RDF data
>> model.
>>
>>  Having said this, the introduction [1] currently has the following
>> paragraph:
>>
>>  "One of the operations [2] that SHACL engines should support
>> validates that a given RDF node matches a given shape. This operation
>> can be invoked based on any control logic, i.e. applications can pick
>> their own mapping between RDF nodes and their shapes. SHACL also
>> provides two mapping mechanisms based on the RDF triples in the graph
>> being validated. Current proposals for these mechanisms include
>> selection based on sh:nodeShape and rdf:type triples."
>>
>>  So my question is, what is missing to make this clearer? Could you
>> propose alternative prose for this aspect?
>>
>>  Thanks
>>  Holger
>>
>>  [1] http://w3c.github.io/data-shapes/shacl/#introduction [3]
>>
>>  On 4/1/2015 19:20, RDF Data Shapes Working Group Issue Tracker wrote:
>>
>>> shapes-ISSUE-33 (Simonstey): Shifting section "Shape Selection" to
>>> introduction? [SHACL Spec]
>>>
>>> http://www.w3.org/2014/data-shapes/track/issues/33 [1]
>>>
>>> Raised by: Simon Steyskal
>>> On product: SHACL Spec
>>>
>>> Information on how to associate shapes with resources/data they
>>> constrain is imho very crucial and should be part of the
>>> introduction rather than being placed somewhere at the end of the
>>> specification.
>>
>>
>>
>> Links:
>> ------
>> [1] http://www.w3.org/2014/data-shapes/track/issues/33
>> [2] http://w3c.github.io/data-shapes/shacl/#operations
>> [3] http://w3c.github.io/data-shapes/shacl/#introduction