[topicmapmail] Intelligence is not optional [Re: Announcement of XML Schema ...]

[Martin Bryan]

Bernard
>
> " When people start to create topic maps they will almost certainly create
> a single type of occurrence,
> and will not assign them roles " etc ...

How on earth do you make the above statement equate with:
>
> I wonder how you can assert such predictions, with others remarks in
> various posts carrying - unless I get you wrong all along this debate -
the
> understatement that most authors and users won't be able to understand the
> very essence of the tool, and make meaningless use of it.  ?

What I said above was that, as is always the case, people will start by
using the minimal set of functions they need to get started, and only start
to realise their limitations when then come do try to do something more
fancy, like combining their topic map with someone elses.

> My position is that users and authors are not to be considered as fools to
> begin with and for ever, and that using a smart tool is somehow helping
> users to build up intelligence and knowledge *for themselves* ...

True, but first they need to see some examples of "good practice" to help
them learn the best way to create topic maps

>I aknowledge that may be incompatible with the stress
> of a production environment. But in that case, my position is that what is
> wrong with that is that very stress.

Ah, there speaks an academic :-) Unfortunately I'm an over-stressed systems
administrator for much of my time these days!

> In the long run, time passed to let
> people understand the tool they use is far from lost time.

True, but practice teaches that the average user does not bother with the
theory - he will try to make the program work without reading the manual if
he possibly can.

> And not for pure knowledge's sake, but because I'm deeply convinced that
we
> (and there I don't speak of our small community, but of mankind at large)
> have no choice for the century and millenium to come that to become
> quickly, effectively, collectively and widely more intelligent that we've
> been up to now, if we don't want to hit the wall in a very productive and
> effective way.

I could not agree with you more. As Tim Berners-Lee stated "The semantic web
is a computer system, a distributed machine which should function so as to
perform socially useful tasks." The key is what we define as socially
useful. The anarchist will claim that instructions for making a bomb is
socially useful, but there are those who would disagree!

The problem I am seeking to solve is how to intergrate ontologies with XML
DTDs and Schemas in a way that does not require you to reorganize your data,
or rename the logical units you are currently working with. Its possible,
but difficult to do. Neither RDF, OIL or XTM help solve this problem, but
there are techniques around based on AFs that can if we are very careful.

Martin