Re: Misusing shacl for sparql examples?

> On 16 Feb 2024, at 10:21 pm, Jerven Tjalling Bolleman <Jerven.Bolleman@sib.swiss> wrote:
> 
> Hi All,
> 
> I have a question, we at SIB kinda misuse shacl to represent our query examples.
> This started with the UniProt sparql  endpoint and spread to others and we are considering this as an internal standard for all our query examples.
> 
> e.g.https://sparql.uniprot.org/sparql/?query=DESCRIBE+%3Chttps%3A%2F%2Fsparql.uniprot.org%2F.well-known%2Fsparql-examples%2F16%3E&format=html&query=DESCRIBE+%3Chttps%3A%2F%2Fsparql.uniprot.org%2F.well-known%2Fsparql-examples%2F16%3E
> 
> We use the sh:SPARQLExecutable, sh:SPARQLSelectExecutable, sh:prefix, sh:namespace, sh:select, sh:ask, sh:SPARQLAskExecutable etc.
> Unfortunately we also  invented our own sh:describe as in the example linked.
> 
> I was wondering if this use of the ShACL vocabulary is acceptable or should be discouraged. As well as general ideas on making this more usable in ShACL native tools.

I think for symmetry reasons alone, the SHACL namespace should define all relevant variations.

Please feel free to record an Issue at https://github.com/w3c/shacl/issues so that the CG and future 1.2 WG can address that.

Meanwhile I believe what you are doing makes practical sense, for this particular case. Needless to say this should NOT encourage everyone else to use undefined terms from the sh: namespace.

Holger


> 
> Regards,
> Jerven
> 
> 
> 
> Jerven Tjalling Bolleman
> Principal Software Developer
> SIB | Swiss Institute of Bioinformatics
> 1, rue Michel Servet - CH 1211 Geneva 4 - Switzerland
> t +41 22 379 58 85
> Jerven.Bolleman@sib.swiss <mailto:Jerven.Bolleman@sib.swiss> - www.sib.swiss <http://www.sib.swiss/>