Re: provenance of provenance

Hi Satya,

On 05/17/2012 04:57 PM, Satya Sahoo wrote:
> Hi Luc,
>
>     In a sense, the approach is similar to Tim's approach
>     http://www.w3.org/2011/prov/wiki/Using_graphs_to_model_Accounts,
>     where he was using a hash as the graph identifier. If the
>     descriptions change, so does the hash. I think using hash is too
>     strong
>     requirement, hence, we have regular identifiers.  (A secure
>     version could use them).
>
>     If an application does not wish to distinguish between two bundles
>     (note: a bundle does not have state!), ... then
>     it shouldn't produce bundles.  There is no requirement to generate
>     bundles.
>
> I did not understand - is there a construct other than bundle for 
> handling multiple provenance assertions together? If bundle is the 
> only construct available then we need to avoid "over-specifying" it.

What is overspecified?

An entity is a physical, digital, conceptual, or other kind of thing 
with some FIXED aspects.

A bundle identifies a set of provenance descriptions. That's it's fixed 
aspect. Other descriptions, mean other entity.

>
>>     2. I am not sure why provenance location, a function for
>>     accessing provenance and very often application/domain dependent,
>>     should be part of DM? "A provenance locator is information that
>>     helps locate provenance descriptions. It may identify a service,
>>     or may consist of a URI where provenance descriptions can be
>>     found." What if location of provenance of an entity or activity
>>     is written down as set of instructions on file (or paper) -
>>     provenance of a car is located in a file in the filing cabinet of
>>     manufacturer? In this case, the application cannot use the
>>     currently defined hasProvenanceIn construct.
>
>     There is no claim that we covering all possible ways of accessing
>     provenance. That's why we have attributes to allow other ways
>     to be expressed (you could encode there your instructions to find
>     a paper for instance).
>
> So, "target", "service", "provenance" (very uncomfortable with this 
> label btw) are just specific examples of the generic "attributes" - 
> this is not clear from the current signature.

I think that target needs to remain as part of the definition.

I suggested to Jun that we could see prov:service and prov:provenance 
(which looks strange agreed, but it's the one chosen in the PAQ)
as optional reserved attributes.

Luc

>
> Thanks.
>
> Best,
> Satya
>
>
>
>     Luc
>
>
>>
>>     Thanks.
>>
>>     Best,
>>     Satya
>>
>>     On Thu, May 10, 2012 at 5:14 PM, Luc Moreau
>>     <L.Moreau@ecs.soton.ac.uk <mailto:L.Moreau@ecs.soton.ac.uk>> wrote:
>>
>>
>>         Dear all,
>>
>>         We are seeking feedback on text regarding bundles (allowing
>>         provenance
>>         of provenance to be expressed).
>>
>>         http://dvcs.w3.org/hg/prov/raw-file/default/model/working-copy/wd6-bundle.html
>>
>>         It is addressing ISSUES-257, ISSUE-260, ISSUE-88, ISSUE-297.
>>         We will respond to these issues individually, shortly.
>>
>>         Cheers,
>>         Luc
>>
>>
>
>     -- 
>     Professor Luc Moreau
>     Electronics and Computer Science   tel:+44 23 8059 4487  <tel:%2B44%2023%208059%204487>
>     University of Southampton          fax:+44 23 8059 2865  <tel:%2B44%2023%208059%202865>
>     Southampton SO17 1BJ               email:l.moreau@ecs.soton.ac.uk  <mailto:l.moreau@ecs.soton.ac.uk>
>     United Kingdomhttp://www.ecs.soton.ac.uk/~lavm  <http://www.ecs.soton.ac.uk/%7Elavm>
>          
>
>

-- 
Professor Luc Moreau
Electronics and Computer Science   tel:   +44 23 8059 4487
University of Southampton          fax:   +44 23 8059 2865
Southampton SO17 1BJ               email: l.moreau@ecs.soton.ac.uk
United Kingdom                     http://www.ecs.soton.ac.uk/~lavm