Re: The NoSPARQL use case

Hi Iovka,

> On 2 Mar 2015, at 13:11, Iovka Boneva <iovka.boneva@univ-lille1.fr> wrote:
> 
>> The conclusion from the workshop wasn’t that SPARQL is too “low level”. The conclusion was that SPARQL queries cannot easily be inspected and understood, either by human beings or by machines, to uncover the constraints that are to be respected.
>> 
>> A “macro” mechanism that wraps SPARQL queries into named parametrised templates should fully address this particular concern
> 
> Do not agree here. Named macros are not understandable by machines ... except if the human associates with the macro a formal description that algorithms can deal width.

Named macros backed by SPARQL are also understandable by machines

- if it’s a pure constraint checking machine that includes a SPARQL engine, or
- if knowledge of the macro has been hardcoded in the machine by a software engineer who has read the SPARQL or the informal description provided by the macro author.

I’m assuming that there will be well-known libraries of macros (including a “core library” produced by this group) that vendors may want to support in particularly clever ways that go beyond standard SPARQL evaluation, hence they may hardcode native support for these macros. This will also allow the language to continue to grow after the end of this WG.

> Macros might even be misunderstood by humans, even if the macro comes with a very carefully written description in English. Only languages with well defined formal semantics allow to express things w/o ambiguity.

If humans are involved, the ability to make unambiguous statements doesn’t guarantee correctness of the statement, nor does it guarantee that the statement will be understood as intended. Humans muddle through quite well with natural language.

Best,
Richard

Received on Monday, 2 March 2015 15:32:33 UTC