[topicmapmail] Wiki/WebLog application built on top of topicmaps...

[Guy Murphy]

Replies inline.

[snip]
> You have lost me here.  Even if you created a bi-directional relationship,
why
> would "Meg Ryan" have any more-or-less involvement?  I think you are
adding
> your own semantic interpretation to omni-directional vs. bi-directional
links,
> and that is what is creating the issue.

One is always left with the end user applying their own semantic
interpretation to an asserted link. But I'm not applying the interpretation
you seem to think.

Before continuing though it's really important to acknowledge that what
matters is the semantic interpretation that users place upon topicmaps. Any
notion of a "pure model" of interpretation is of no use to a user presented
with links.

My point with the meg example is that an assertion can be very useful when
viewed from one side, and of no interest -- further to that an actual
hinderance -- when viewed from the other side.

> What you actually need to do is track
> who created an assertion, so that they are responsible for managing it.

No you have it the wrong way around.

Primarily the created association is so that the reader of a topic can see
who created it. In many cases a person who has created a large number of
topics will not be interested in the fact that they created Topic A at all,
because they may have created 100s of topics.

The association is important on one side, and not important on the other....
one can view applying a direction to such associations as merely a special
case of weighting with a weight of 1 on one side and 0 on the other.

> You
> could do this via topic map scopes, since they can be used to define from
whose
> point of view an association is made.

I'll consider this further, as it's not something I've thought about.

> At the moment, you are running with the
> idea that the source of a omni-directional association can also be
construed as
> the creator.

No I'm not... the creation association was used merely as a solid case
example that applies to the application I'm working on. Forget creation if
it's muddying the field, and simplt consider a one-to-many relationship of
your choice where there "manies" are so numerous as to loose their value in
the noise they create.

Remember that I've considering topicmaps from the point of view of Topic
Navigation Maps, and not merely a datastore to be queried... I'm creating
navigable views, and any view with 100s of links is no use to man nor beast.

> I cannot help but wonder if that assumption might not let you
> down at some stage.
[snip]

Cheers,
    Guy.