Re: PROV-ISSUE-473 (generating-activity): Unique generation events and activities [prov-dm-constraints]

Hi,

There has been no further discussion of the changes proposed to resolve the issue, other than support from Daniel and Curt.  (Tim mentioned in an offline exchange that it is OK, although the "simultaneous" constraint on the generating events surprised him).  I am going to close the issue now.

Graham, I had the impression that we intentionally avoided talking about events in other documents (moving the discussion of them to constraints) to *avoid* adding complications that are not needed to explain e.g. how to use prov-n syntax or prov-o terms.  Conversely, this means that prov-dm and so on should avoid describing constraints or events, so as to avoid conflicting definitions.  If there are places in other documents where there is a conflict, or overlapping specifications that might be ambiguous, these should be raised as editorial issues on them, I think.  If you perceive a problem with  the way things are described in prov-constraints, please make a new issue saying what you think should change in prov-constraints.

Incidentally, prov-constraints describes events mainly in the non-normative rationale section, and the conformance statement says that if there is an apparent inconsistency between the informal text and formal constraints, the latter takes precedence for the purpose of validity/equivalence checking.  I think it would be sensible to broaden this to say that "if there is an apparent inconsistency between formal text of constraints and informal description in this specification or others, the formal text takes precedence".  Would this avoid the issue you perceive below?

--James

On Aug 13, 2012, at 9:25 AM, Graham Klyne wrote:

> On 10/08/2012 11:45, Luc Moreau wrote:
>> As a reminder, nowhere in prov-dm, we define a notion of event.
>> Events are only defined in prov-constraints.
> 
> I noticed this, and had been wondering if I should comment.
> 
> I don't think it's a technical problem, but did wonder if it might be source of confusion (events being described somewhat differently across the two documents).  I think I had previously suggested that events might be defined explicitly in DM, making them easier to talk about, and reference in (say) the ontology, but I'm not sure offhand if that opens a different can of worms.
> 
> #g
> 
> 
> 


-- 
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.

Received on Thursday, 16 August 2012 12:37:48 UTC