Re: PROV-ISSUE-52 (provenance-source-equal-treatment): why handling provider and third parties differently? [Accessing and Querying Provenance]

The reason for handling provider and third parties differently is that the 
resource provider who also provides provenance can be presumed to cooperate with 
the requirements for providing provenance when serving a resource.

A third party, by my definition, is one who does not have cooperation of the 
resource provider for enabling discovery of provenance information.

One might say:  why not use third party mechanisms for everyone?  I suggest:
(1) the third party solutions are more complex to deploy
(2) the third party solutions impose considerable additional network traffic and 
processing overhead compared with provider-based solutions

And in any case, the third party solutions *still* leave the issue of actually 
discovering the third party service URI, so don't buy that much.

...

To your example, they may amount to the same outcome at a trivial level, but the 
  deployment and trust issues involved in achieving that outcome are quite 
different.

#g
--

Provenance Working Group Issue Tracker wrote:
> PROV-ISSUE-52 (provenance-source-equal-treatment): why handling provider and third parties differently? [Accessing and Querying Provenance]
> 
> http://www.w3.org/2011/prov/track/issues/52
> 
> Raised by: Luc Moreau
> On product: Accessing and Querying Provenance
> 
> Let us assume I have the identifier for the document that I currently hold.
> Why do we have two different methods to obtain its provenance-uri?
> 
> HTTP HEAD if we want to obtain provenance-uri(s) from the document provider.
> 
> SPARQL query if we want to obtain provenance-uri(s) from a third party.
> 
> Ultimately, it's the same operation (give me provenance-uri(s) for the current document), why two different methods? My view is that we should have a SINGLE operation, whoever we are dealing with, document provider or third party.
> 
> 
> 
> 
>

Received on Thursday, 28 July 2011 21:40:25 UTC