Marc de Graauw

Scoped Occurrences (6)

• It's useful. (term-theme)
• This may constrain subjects, so I would be hesitant to do this. (subject-vs-resource)
• I would tend to the any-subject view for backward compatibility if there is a reasonable way around the problems with the set of all topics (or similar contructs). (term-scope-def)
• Say little that is normative, and you can say a lot that is non-normative. I would not want my subjects to be constrained in any normative way. (term-subject-def)
• Apart from all the subject/topic stuff there is another argument to drop the 'topic', 'association' and 'occurrence' PSI's. They are most uninformative. Since any topic is an instance of class 'topic' by default (unless specified otherwise, says XTM), there is no information involved that is of any use to anybody. The user of the Topic Map will not want to see the class 'topic' listed among the classes of a topic; therefore a Topic Map engine will have to suppress it, and it makes no sense to me to introduce a property which only needs to be suppressed later. For the Topic Map engine there is also no useful information in the PSI, since Topic Map engines 'know' what topics are (i.e. how to process them) without using this PSI. So to me it seems a piece of redundant information. Same goes for the 'association' and 'occurrence' PSI's. (psi-generics)
• Since the definition is broken anyway, and there is no clear need for it, I suggest it be dropped. (term-subject-identity)