W3C home > Mailing lists > Public > public-data-shapes-wg@w3.org > January 2016

Re: ISSUE-115: Did we rush this decision?

From: Holger Knublauch <holger@topquadrant.com>
Date: Fri, 15 Jan 2016 09:05:01 +1000
To: "public-data-shapes-wg@w3.org" <public-data-shapes-wg@w3.org>
Message-ID: <5698299D.1070105@topquadrant.com>


On 15/01/2016 9:02 AM, Holger Knublauch wrote:
> As I am trying to implement the resolution on ISSUE-115 to use the 
> following syntax
>
> ex:MyShape
>     a sh:Shape ;
>     sh:closed true ;
>     sh:ignoredProperties ( rdf:type ) ;
>     ...
>
> I cannot help but think that we may have rushed this decision and have 
> forgotten a few aspects:
>
> 1) With the new syntax it is no longer possible to specify 
> sh:priority, i.e. closeness is always an Error and cannot be a Warning

Correction: sh:severity

Holger


>
> 2) With the new syntax we have lost the ability to specify filters so 
> that closeness only applies to certain instances. The work-around 
> would be to use different scopes, but that's not the same thing and 
> would lead to duplicate shape definitions (not user friendly).
>
> 3) With the new syntax we have lost the ability to abbreviate the 
> common case of ignoring rdf:type (as above).
>
> In my proposed syntax, SHACL would have included shortcut resources 
> (instances of sh:NodeConstraint) so that the syntax would simply be
>
> ex:MyShape
>     a sh:Shape ;
>     sh:constraint sh:closed ;
>     ...
>
> or
>
> ex:MyShape
>     a sh:Shape ;
>     sh:constraint sh:closedIgnoringRDFType ;
>     ...
>
> If the main reason for the adoption of Arthur's proposal was a 
> user-friendly syntax then I believe the syntax above is better. As an 
> added bonus, it is far more consistent. Currently I would need to 
> hard-code sh:closed as a special case.
>
> Sorry to be a pain, but I believe the topic should be reopened, at 
> least as a quick check whether we are really sure about our resolution.
>
> Thanks,
> Holger
>
Received on Thursday, 14 January 2016 23:05:37 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:30:29 UTC