Re: Reusable Shapes

Reusable could mean lots of things.  Addressable could also mean lots of 
things.  The Working Group probably needs some separate document where 
approved requirements are elaborated.

peter

On 11/19/2014 06:22 AM, Eric Prud'hommeaux wrote:
> * Peter F. Patel-Schneider <pfpschneider@gmail.com> [2014-11-19 04:38-0800]
>> Is there a pointer to documentation where the working group accepted
>> either of these requirements?   The closest that I can see is the
>> acceptance of R103, which includes naming of constraints and
>> recursion.
>>
>> http://www.w3.org/2014/10/30-shapes-minutes.html
>>
>> However, these minutes have not yet been approved, at least
>> according to the main WG wiki page.
>>
>>
>>
>> I don't think that either of these are best described as reusability
>> of shapes or of rules within shapes.  I don't think that there is
>> even any resolution that indicates that there will be rules within
>> shapes.
>
> I think the relevent resolution is
> [[
>    RESOLVED: accept R147 "addressable schemas" and R148 "addressable
>    constraints" as requirements
> ]]
>
> It doesn't per se mandate reuse, but it madates addressability, which
> I would expect would be the chief technical barrier. A related example
>
> [[
> <oslc-change-request> a rs:ResourceShape ;
>          …
>  rs:property
>   <oslc-change-request#dcterms-title> ,
>   <oslc-change-request#oslc_cm-status> .
>
> <oslc-change-request#dcterms-title> a rs:Property ;
>  rs:propertyDefinition dcterms:title ;
>  … .
>
> <oslc-change-request#oslc_cm-status> a rs:Property ;
>  rs:propertyDefinition oslc_cm:status ;
>  … .
> ]] — <http://www.w3.org/Submission/2014/SUBM-shapes-20140211/#ex_change_req>
>
> illustrates that if folks assign URLs to rs:Propeties, it would be
> trivially follow that another schema could re-use those defns.
>
> [[
> MY:change-request a rs:ResourceShape ;
>          …
>  rs:property
>   <oslc-change-request#dcterms-title> ,
>                  [ rs:propertyDefinition MY:disposition ;
>                    rs:occurs rs:Exactly-one
>                  ] .
> ]]
>
>
>> peter
>>
>> PS:  Let me say, yet again, how much I miss CommonScribe.
>>
>>
>>
>> On 11/19/2014 03:47 AM, Eric Prud'hommeaux wrote:
>>>
>>> On Nov 19, 2014 12:28 PM, "Dimitris Kontokostas"
>>> <kontokostas@informatik.uni-leipzig.de
>>> <mailto:kontokostas@informatik.uni-leipzig.de>> wrote:
>>>>
>>>> I came late and this thread became so big that is hard to pick it up properly.
>>>>
>>>> I would like to raise another related issue regarding Shapes reusability.
>>>> Assuming I have X defined shapes and Y applications profiles that each
>>> profile can reuse any of the X defined Shapes. Is this case something that
>>> this WG would like to cover?
>>>> If yes, what would be the proper approach to store & define Shapes?
>>>
>>> We have, if I recall, accepted requirements to have both reusable shapes and
>>> reusable rules within those shapes. So far, all of the proposed technologies
>>> enable that, though of course we'd want to then consider practical ways to
>>> overload, extend, and maybe even retract parts of reused rules.
>>>
>>>> Best,
>>>> Dimitris
>>>>
>

Received on Wednesday, 19 November 2014 14:51:02 UTC