Re: updates to PAQ doc for discussion

Hi Jim,

I think that from your point of view, and perhaps most people in the WG 
share this opinion, an entity can be an IVPof another entity even when 
the the second characterize a different version (state) of the first.

I don't share this view. In the example that you gave, although B and C 
are versions of the same document, I view them as entities that 
characterize 'different' things, and so C is not an IVPof B. In other 
words, the way I see it is that B is an entity that is a 
characterization of a document  A1, and C is an entity that is a 
characterization of a document A2. A1 and A2 are different things, 
despite the fact that one of them is a version of the other. Note also 
that A1 and A2 are not represented in the model, the only 
representations that we have of them is via the entities that 
characterize them, i.e., B and C.

Khalid



On 23/08/2011 14:03, Myers, Jim wrote:
>> I could not follow the example you gave, as I don't know what the semantics
>> of the edges is. Do they refer to IVPof.
> A is a document, B and C are specific versions of it (specific text strings).  B IVPof A, C IVPof A, C derivedFrom B. I see versionOf being applicable to B versionOf A, C versionOf A (i.e. versionOf as a subtype of IVPof), and I can see 'revisionOf' applying to C revisionOf B (i.e. revisionOf is a subtype of derivation).
>
> If I understand, you're talking about B and C when you say:
>
>>>> In my opinion versionOf implies that there were some changes that we
>>>> are aware of, and, therefore, the characterizations we end up with
>>>> are describing different entities. I would therefore prefer to
>>>> describe relationship between versions using derivation rather than IVPof.
> Is that correct? If so:
>
> How would you relate B and C to A? If D is produced from C by the same type of process (e.g. another text edit) as was used to produce C from B, would it be a revision? Would there be anything one could infer between D and A?
>
>   Jim
>>> I'm confused:
>>> Given
>>>
>>>                   A
>>>                /    \
>>>              B<---C
>>>
>>> I would say B and C are IVPof/versionOf A and C is derivedFrom B - I
>>> don't see how you would choose "derivation rather than IVPof" since
>>> they do different things. (To keep going, D derivedFrom C wouldn't
>>> automatically make D a version/IVPof A, so one can't just infer
>>> versionOf from derivation, etc.)
>> In that case, I would defined VersionOf as a subrelation of DerivedFrom,
>> instead of IVPof.
>>
>> Khalid
>>
>>
>>>     Jim
>>>
>>>
>
>

Received on Tuesday, 23 August 2011 17:01:22 UTC