[topicmapmail] Are Facets Really Simple After All?

[Kal Ahmed]

On Mon, 2003-12-01 at 00:24, Murray Altheim wrote:
> Kal Ahmed wrote:
> [...]
> > The discussion itself is useful if it is to help us distinguish
> > different types of "information relevant to a subject" and I encourage
> > that - indeed I have found this whole thread a very interesting one. My
> > only point is that you seemed to me to be implying that ISO 13250
> > constrains the use of occurrences in a way which from reading the
> > definitions I feel it clearly does not.
> 
> I'm sure we've both read over the standard enough times between us
> to make it a bestseller. In my reply today to Thomas (at 6:32pm),
> I tried to make it clear that simply I am *very* uncomfortable
> with what I consider a misuse of the term "occurrence" -- not that
> the standard states clearly what constraints may be put on what is
> or what isn't a Topic occurrence. It of course makes clear not to
> put any constraints at all, but that doesn't mean there aren't
> inherent, semantic constraints within a specific application. I
> think it would have been severely limiting to create any explicit
> constraint within the standard, as the applications of Topic Maps
> may be *very* wide, and this kind of thing is quite application
> dependent.
> 
> But by the same token, I think simply a pragmatic analysis of
> what you're proposing should suggest some best practices. By that
> I mean, a simple reading in English of what you're suggesting.
> I can understand that an occurrence of the Topic "Paris, France"
> might be a latitude/longitude for Paris, an important date in
> its history, a photograph of the Champs Elysees, etc. The standard
> rightfully doesn't constrain any of this kind of thing; it would
> make Topic Maps much less interesting and valuable. But I can't
> imagine that a "correct" occurrence of the Paris Topic would be
> "2003-11-23T11:34:55", "245,190", or "#bce033" (i.e., the creation
> date of the Topic in my application, the location of the Topic
> node in my TouchGraph visualization, or the colour of the displayed
> node, resp.). That just isn't right. Now, if I never planned to
> share that Topic Map with anyone, hell, I can do what I like*),
> but it doesn't *to me* express a correct interpretation of the
> standard. Even in the inimitable vagueness of its definition of
> "occurrence", a property of a Topic and an occurrence of a
> Topic simply aren't the same thing. Certainly not semantically.
> 

In that case we agree. But you talk here of meta data for the topic (as
opposed to meta data regarding the subject) for the first time. I was
simply pointing out that subject meta data is valid "information
relevant to the subject". I would expect topic meta data to be expressed
as occurrences of another topic whose subject is the topic construct
that you are making the assertions about. 

Cheers,

Kal
> -- 
> Kal Ahmed <kal@techquila.com>
> techquila