Re: A-, T- and S-nodes WAS [xtm-wg] A challenge on "the graph"

[Lars Marius Garshol <larsga@garshol.priv.no>]

* Lars Marius Garshol
|
| [my objections to the SRN graph model]
| That everything is reduced to three kinds of nodes, and that one of
| those kinds of nodes (S-nodes) seems *beep*. Turning resources into
| topics is just *beep* *beep* *beep*. There is more, but I have no
| time right now to really explain this properly.

* Sam Hunting
| 
| Well, Lars, since this is the key technical point of the discussion,
| I really think that you should try and find the time!

I could try, but I am very heavily loaded at the moment, in fact so
much so that I ought not to participate in this discussion at all.
Also, next week I will be going on two weeks of holiday, so there is
precious little time in which to put together a proper criticism of
the SRN PM.
 
| Well, I'm hearing two things then. One is that the PM shouldn't look
| like an implememtation; the other is that it is too far "removed"
| from an implementation. What am I missing here?

I don't know. Does it not make sense to you that a PM should use a
convenient representation of the abstract structure of topic maps in
order to not make the model too hard to comprehend and connect with
the actual form of implementations? 

And does it not make equally good sense to require it to not take the
form of a particular implementation, which would tend to constrain
implementors too closely? Or at least seem to do so.
 
* Lars Marius Garshol
|
| [the graph model]
| It also requires considerable mental effort to piece together
| what is going on there.
 
* Sam Hunting
|
| But this is a pedagogical issue, not a technical one.

It is both. The choice of model formalism greatly affects how easy it
is to understand the resulting model. You can represent nearly any
model using very nearly any notation, but certain notations fit the
task at hand better and so are easier to comprehend. 

So the technical choices have pedagogical implementations. This is
also the case with the XML Recommendation. They could have chosen to
use pure BNF to specify the grammar, which would have made the grammar
easier to feed into parser generators, but instead they chose to use
an extended BNF notation, since it is far easier to read.

It is no different with the topic map model.
 
| The markup is optimized for human consumption -- and the DTD is
| getting a lot of traction (market penetration) so presumably that
| portion of the effort has been successful.

So it seems. Now, does it not make sense to follow a similar approach
in the processing model? 
 
| Since the DTD represents a topic map in one dimension, and the PM
| has to represent the topic map in many, there should be no surprise
| that the two representations are not isomorphic.

Of course they cannot be! On the other hand, they are different
reflections of the same underlying concepts, and so there must
inevitably be certain similarities.

I'm not advocating a model that follows the syntax in every detail,
but I am advocating a model that represents all the various features
of topic maps clearly. If topic maps have such a thing as base names,
why not represent them as base names, instead of turning them into
something else, requiring people to map back to the intended concept?

If there is anything that should be optimized for human consumption it
is the processing model.

--Lars M.


------------------------ Yahoo! Groups Sponsor ---------------------~-~>
Secure your servers with 128-bit SSL encryption!
Grab your copy of VeriSign's FREE Guide,
"Securing Your Web site for Business." Get it now!
http://us.click.yahoo.com/KVNB7A/e.WCAA/bT0EAA/2n6YlB/TM
---------------------------------------------------------------------_->

To Post a message, send it to:   xtm-wg@eGroups.com

To Unsubscribe, send a blank message to: xtm-wg-unsubscribe@eGroups.com 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/