RE: PROV-ISSUE-1 (define-resource): Definition for concept 'Resource' [Provenance Terminology]

A few thoughts in reply...

 Jim

> there is also a discussion on whether an Information Object has the
same resource status as a resource as a physical object

I'll randomly note that 'physical objects' such as statues often exist
as multiple versions/copies (Brancusi's "The Kiss" for example), so
digital resources are not unique in that aspect.

> I personally agree that any notion of provenance refers to a specific
state of a resource.  Naturally here we mean "observable 
> state". I have not seen the notion of observer introduced in this
discussion (I have yet to catch up with the others!), but it seems 
> natural that provenance is relative to an observer.

I claim that what constitute state depends on the observer and their
processes of interest (my earlier post) and we should make that
explicit. Accounts capture the sense of an observer's perspective, but I
don't think we've explicitly noted that what we consider resource versus
state of resource is observer/account/process dependent.

> - can we also assume that provenance is /monotonic/ wrt the state
evolution of the resource it refers to.

I think any versioned resource could have a graph of 'states' rather
than a single line and the provenance graph could then have merges and
splits. Aggregate/collection resources would similarly have the
potential for complex graphs. All would be directed and acyclic though.
I could also imagine accounts of a file, one of which captures textual
changes (edits) and the other format translations (not changing text)
where there would be insufficient info (no timestamps) from the two
accounts to decide on the full intermediate text+format states of that
file... :-) 


> I do have a problem with "containers" as a separate notion from
resource, however.
>Isn't a database a container? and a resource? 

A database resource can play a container role w.r.t. some
processes/accounts....

Received on Wednesday, 1 June 2011 20:11:04 UTC