Re: [topicmaps-comment] TAO vs. ERA

[Bernard Vatant <bernard.vatant@mondeca.com>]

Robert

> For associations I can choose a scope to constrain its validity,
> for instanceOf I can not. In other words, the scope of an instanceOf
> is then the unconstrained-scope? Do I interpret something into the
> standard which is not here?

I think you are right. There is no <scope> in <instanceOf>, which means when you
declare an instantiation this way, it's an "absolute" one. If you want to scope
it, you have to create explicit class role and instance role. IMO this is an
inconsistency in the spec, since <instanceOf> is a shortcut for an association,
and hence should be "scopable".

BV
> > | It can be worse. Consider "documentation" that can be considered as
> > | a topic type (a set of documents), as a role (in a
> > | subject-documentation) or as this latest association type itself
> > | (the fact of documenting something) ... There again, you should
> > | define three different topics:
> > | 1. documentation-as-topic-type
> > | 2. documentation-as-role-specification
> > | 3. documentation-as-association-type


LMG
> > I think 1 and 2 are the same here, while 3 is probably not...

rho

> Hmmm. This was not really a practical problem yet, so I hope
> this is an academic discussion. ;-)

Unfortunately, it is not academic, it is a practical problem for me *right now*.
As said before, "documentation" is a term of the thesaurus I have to import, and
I have to decide from the beginning how many topics to create for class, role,
and association. And it is not the only example, I've also to deal with
"publication" and "production" and "technical solution" and many more ... A
careful methodology is to create the three to begin with, and have separate bags
of topics for classes, role types and association types, as Tom pointed out. You
can merge them anytime after if you like. But you can't make it the other way
round.

Bernard