Re: prov-issue-11: what is version?

Hi Stian,

Comment interleaved.

On 06/29/2011 12:19 AM, Stian Soiland-Reyes wrote:
> On Tue, Jun 28, 2011 at 17:03, Luc Moreau<>  wrote:
>> If we agree with the definition I suggested (possible Jim's too, I am not
>> sure), it shows that version (or is revision of) is not a primitive
>> notion in PIL, but can be derived from more primitive concepts.
>> I think we still need to take a view on this concept, since it is part of
>> the charter,
>> and we can't simply ignore it.
> I like your definition, and I still think there could be room for the
> concept, even if is just a composition of other concepts. I would also
> call it 'revision' instead, because 'version' suggests a parent-child
> relationship rather than a sibling one.
> For reference:
>> A thing B is a version of (or should we say revision of) a thing A if:
>> * B is derived from A
>> there exists a thing C, such that:
>> * B is IVP of C
>> * A is IVP of C
>> * statements about C are optional
> In particular because revisions can be done in many ways, and people
> could disagree about whether B is a revision of A or not (by which
> authority, etc), it would be a very useful thing for an asserter to
> say that "I think B is a revision of A" - and to do that without
> defining the abstract C which could be cumbersome and lead to
> conflicts about my C (not) being the same as your C.
> A Revision can also have additional properties, such as what/who
> generated it (which might be different from who made the revised
> thing). For our journalist example, Bob can assert that chart c2 is a
> revision of chart c1 - while the newspaper don't want to do make such
> a strong statement even if they (later) want to acknowledge the
> existence of c2.
> Uh oh - digression follows: I just realised that here c2 is *not* a
> revision of c1, even for bob.
> c2 is derived from f2
> f2 is derived from d2
> d2 is derived from d1
> d1,d2 are IVPs of 'd0'
> so only d2 is a revision of d1. f2 is not a revision of f1 by your
> definition, and neither is c2 a revision of c1 - but I feel that they
> probably should be.

Correct that they are not revisions, according to this definition.
Simply because f1 and f2 occur in different, independent branches of the 

We could have an extra inference rule,

  d2 is a revision of d1
  c1 isDerivedFrom d1 (by specific transformation t)
  c2 is derivedFrom d2 (by same specific transformation t)

then c2 is a revision of c1

> If Bob explicitly says that c2 is a revision of c1 - then he's
> implicitly saying that there's a common IVP of both c1 and c2. Is that
> 'd0', some kind of transitivity through f1/f2, r1/r2 etc, or just some
> abstract idea about the "chart of d"?


Professor Luc Moreau
Electronics and Computer Science   tel:   +44 23 8059 4487
University of Southampton          fax:   +44 23 8059 2865
Southampton SO17 1BJ               email:
United Kingdom           

Received on Wednesday, 29 June 2011 07:15:25 UTC