Re: on constraints

There are still problems.

The document is still confused as to what a constraint component is.

In Section 4 there are things like:
"Constraint Component: sh:ClassConstraintComponent "
which give clear examples of constraint components.
However in the terminology list there is
"A constraint component represents a part of a constraint that is determined
by the values of one or more properties."
These two do not seem to be talking about the same thing.

I suggest that someone in the working group go through the document as a whole
to ensure that shape, constraint, property constraint, focus node constraint,
constraint component, parameter, and related terminology are all defined
correctly and used correctly.

Peter F. Patel-Schneider
Nuance Communications


On 10/02/2016 11:20 AM, Dimitris Kontokostas wrote:
> Thanks again for your feedback Peter,
> 
> I tried to incorporate your comments on the latest commits
> https://github.com/w3c/data-shapes/commits/gh-pages
> 
> let me know if these editorial changes resolve your comments
> (this is not an official WG response but will be tracked by the WG)
> 
> On Fri, Sep 30, 2016 at 4:48 PM, Peter F. Patel-Schneider
> <pfpschneider@gmail.com <mailto:pfpschneider@gmail.com>> wrote:
> 
>     If you remember, it used to be that constraint parameters could not be
>     directly attached to shapes.  Instead there had to be a "local constraint"
>     atttached to the shape and then constraint parameters were attached to that.
> 
>     Now instead constraint parameters (*not* constraints) are attached directly to
>     shapes and the shape itself takes the role that used to be played by the
>     "local constraint".
> 
>     Even the use of "attach" is dodgy here.  Attached how?  It is better to talk
>     about triples with a particular subject, object, and predicate.  Shorthand
>     constructions may be easier to read as English, but it can easily turn out
>     that they don't have the necessary precision.
> 
> 
>     There has also been a terminology shift, again for precision.  The object of
>     an sh:shape triple is a shape, which is now also a constraint.  So for
>     sh:shape to group focus node constraints (not constraint parameters) something
>     like
> 
>       sh:shape [ sh:shape [....] ; sh:shape [...] ]
> 
>     is required.
> 
> 
>     At least this is my understanding of the current mechanisms and terminology.
> 
> 
>     Peter F. Patel-Schneider
>     Nuance Communications
> 
> 
> 
> 
>     On 09/30/2016 01:43 AM, Dimitris Kontokostas wrote:
>     > Thank you for the feedback Peter,
>     >
>     > I tried to address some of your comments here:
>     > https://github.com/w3c/data-shapes/commit/8d8df0e57361558674bd65492cebd9c7358140c7
>     <https://github.com/w3c/data-shapes/commit/8d8df0e57361558674bd65492cebd9c7358140c7>
>     >
>     > On Fri, Sep 30, 2016 at 3:46 AM, Peter F. Patel-Schneider
>     > <pfpschneider@gmail.com <mailto:pfpschneider@gmail.com>
>     <mailto:pfpschneider@gmail.com <mailto:pfpschneider@gmail.com>>> wrote:
>     >
>     >
>     >     "Focus node constraints are attached directly to Shapes"
>     >
>     >     Not any more.
>     >
>     >     "sh:shape can be used to group focus node constraints."
>     >
>     >     Not any more, except by using sh:shape itself, I guess.
>     >
>     >
>     > Can you please elaborate a bit further on these two comments?
>     >
>     > Best,
>     > Dimitris
>     >
>     >
>     >
>     >
> 
> 
> 
> 
> -- 
> Dimitris Kontokostas
> Department of Computer Science, University of Leipzig & DBpedia Association
> Projects: http://dbpedia.org, http://rdfunit.aksw.org, http://aligned-project.eu
> Homepage: http://aksw.org/DimitrisKontokostas
> Research Group: AKSW/KILT http://aksw.org/Groups/KILT
>

Received on Monday, 3 October 2016 00:00:38 UTC