Re: Dealing with distributed nature of Linked Data and SPARQL

On 06/08/2016 11:46 AM, james anderson wrote:
> if the goal is to leave room for the judgement call,
> assuming that dimension is free, to place each in its own graph gives
> one the latitude to make the judgement call, develop some systematic
> which depends on provenance, reflect the question to a manual choice, or
> just project FROM both and allow a naive merge.

Absolutely agree.

>
> which would see to make it the general best practice.
> is there some case which argues against it?

Not AFAIK.  I think named graphs are one of the best tools that we have 
available, and should be considered a best practice.   I just meant to 
caution that named graphs may not *fully* solve the problem.  But they 
are the best starting point, IMO.

David Booth

Received on Wednesday, 8 June 2016 18:45:15 UTC