- From: Holger Knublauch <holger@topquadrant.com>
- Date: Tue, 02 Jun 2015 09:28:54 +1000
- To: Data Shapes Working Group <public-data-shapes-wg@w3.org>
On 6/1/2015 20:20, Peter F. Patel-Schneider wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA256
>
>
>
> On 05/31/2015 11:00 PM, Holger Knublauch wrote:
>> Hi Peter,
>>
>> thanks for starting to bring structure into the suggested refactorings
>> (also in your later email). I do believe we need to tackle them one by
>> one, and break them into separate ISSUEs. So below are some references to
>> tickets (I raised a few of them recently):
>>
> On the contrary, I believe that tackling these issues one by one will end up
> with something that does not have a clean conceptual basis and thus that
> will be hard to understand and that it will be hard to determine whether the
> result is reasonable. I thus strongly feel that there first needs to be a
> determination of what the foundations of SHACL will be.
While a clean conceptual basis helps and will be especially important
when we have to explain our technology stack to the outside world, we
need to approach this issue from both sides. The times where we could
throw in some untested claims with hand-waving and some generic
statements ("some magic happens here", "everything is turned into a
single SPARQL query") is over. Some things only become obvious when you
look at specific examples and implementation details. SPARQL is quite a
complex language, and I wouldn't want to mess with analyzing syntax
trees, OPTIONALs and all these things.
Having said this I believe many of the ISSUEs are actually quite
conceptual and will help create a clean basis. In particular I believe
we have to answer the following tickets soon:
- ISSUE-1 (Inferencing)
- ISSUE-60 (Other Extension Languages)
- ISSUE-62 (General selection or filtering)
For example you continue to bring up dependencies on RDFS, but this only
works as long as the WG decides that this is realistic.
Holger
Received on Monday, 1 June 2015 23:30:58 UTC