[sc34wg3] Draft Reference Model

[Marc de Graauw]

* Marc de Graauw
|
| > * Graham Moore
| >
| > | I think we came to some consensus in montreal that scope is no more than
| > | untyped associations between an association and a set of topics.
| >
| > I wasn't in Montreal and wouldn't know about the decisions there. But I
| > wouldn't object to this conclusion at all, only I wouldn't say scope is an
| > _untyped_ association, just that it is not explicitly typed in the syntax.
The
| > implicit type of the association of course would be "extent of validity"
as
| > ISO 13250:2000 and XTM say.

* Sam Hunting

| Not sure I understand this.
|
| The assertion derived from such an association would be *explicitly*
| typed, would it not?

The way I see it we could say that the <scope> element in XTM is shorthand for
an association of type "extent of validity", with roles "scoped characteristic
assignment" (one role player) and "scoping topic" (one or more distinct role
players). Just in the way <instanceOf> is shorthand for class-instance
associations. That would mean the assertion in a TM graph would be explicitly
typed, which is desirable. (It would also mean the association in XTM for
which <scope> would be shorthand would be explicitly typed.) It also would
mean backward compatiblity is maintained, as existing <scope> elements in
Topic Maps remain valid, only the semantics are made more explicit. (Same for
HyTM.)

| 3.6.4.1.2 in the RM permits "untyped" assertions -- while it may be that
| all scopes are untyped associations between an association and a set of
| topics, does it the reverse statement hold -- that all untyped
| associations between an association and a set of topics are scoped?

No! Please no!

| (Note also that the RM can be used to give an operatioanl definition of
| "extent" in terms of graph constructs.)

Not sure I understand this...
Could you expand?

Marc