PROV-ISSUE-469 (zero-timespan-entity): Can entity's have zero timespan? [prov-dm-constraints]

PROV-ISSUE-469 (zero-timespan-entity): Can entity's have zero timespan? [prov-dm-constraints]

http://www.w3.org/2011/prov/track/issues/469

Raised by: Stian Soiland-Reyes
On product: prov-dm-constraints

wasInvalidatedBy strictly follows wasGeneratedBy - do we have WG
consensus?  Activities can be zero-length, is that OK? Need justification for the difference.

Has this been discussed with the WG?


>From Stian's review: http://lists.w3.org/Archives/Public/public-prov-wg/2012Aug/0021.html


> Constraint 38 (generation-precedes-invalidation)
> IF wasGeneratedBy(gen;e,_a1,_t1,_attrs1) and wasInvalidatedBy(inv;e,_a2,_t2,_attrs2) THEN gen strictly precedes inv.

Why is this relation in particular *strictly precedes*? This needs to
be justified (beyond "we need some strictly in there so we can do our
loop testing"). So an entity can't have zero lifetime, but an activity
may?

An activity can use an entity at the same time as it was generated,
but it can't invalidate it then? This implies some kind of minimal
planck time on entities, which is probably OK for most applications of
PROV.


After some discussion with fellow geeks, I have however come go agree
that an entity can't have an empty lifespan, to avoid problems and to
ensure time moves forward. I think we need to formulate this by using
the description of an entity as characterizing and fixing aspects for
some duration. But currently the DM descriptions seem to imply the
opposite, entities can have zero lifespan, but activities cannot!

http://www.w3.org/TR/prov-dm/#term-entity
> An entity ◊ is a physical, digital, conceptual, or other kind of thing with some fixed aspects; entities may be real or imaginary.
> An activity ◊ is something that occurs over a period of time and acts upon or with entities; it may include consuming, processing, transforming, modifying, relocating, using, or generating entities.

I do however see bigger use for a zero lifespan activity, because it
can be used to describe transitions of entities. So can we add to
Constraint 38 some kind of remark about why an entity must have a
non-zero lifespan? Something like:

"Constraint 38 implies that an entity must have a non-zero lifespan by
using 'strictly precedes', that is the entity cannot be invalidated at
the same instant as it is generated. The reasoning for this is that a
meaningful entity is a thing with some aspects fixed. For these
aspects to be fixed, the entity must exist for some (possibly
infinitesimal) time. Note that this requirement does not apply to
activities."

Received on Monday, 6 August 2012 15:31:14 UTC