[xtm-wg] Re: Summary of PM discussion

[<guy.a.lukes@frb.gov>]

[guy]
>>Each level is grounded on the level above it.  Unfortunately, whether
>> it is the right grounding can only be determined by looking at the >
>> detail of what it supports.

[sam]
>Sorry, I'm missing the "above" here. Aren't the three metaphors --
>containment, paths, and multidimensional positions three ways of
>looking at the same thing? Where are the "level"s here?

I think, at least for XTM,  the three metaphors have a layered
relationship, with graphs being the most general and property
sets supported by prose being the most specific.


[sam]
>The problem is, that not all people "grok" the same metaphor. That is
>why I am trying to separate:
>
>   - model
>   - notation
>    - prose
>
>Surely describing "federated global knowledge interchange" requires
>more than one metaphor?

This is exactly right, keep each metaphor separate, and let it do
what it does best.

I was actually think about a simple high level graph, supporting a
moderately detailed set of  UML class diagrams, supporting a detailed
property set with prose.

We should try and keep the "grand unification stuff" out of the
property set layer and keep the detailed attributes for identification
and verification out of "the graph".

I'm not a "grove person" so I hope I'm referencing
 property sets correctly?

[guy]
>> The discussion has congealed around the three basic metaphors for
>> modeling information (paths, containers, and positions).

[sam]
>For "metaphor" should I be thinking "API"? Or is that category
>confusion?

I was thinking more about Cognitive Psychology.  The API would
probably be an elaboration of the UML and would be the last thing
you do, once your convinced everything else is correct.  It's very
difficult to recover from screwing up an API.

[guy]
>>The class diagrams of UML and their inheritance and aggregation
>> concepts are a formalization of the container metaphor.  While they
>>don't provide the global integration of a graph, they provide a more
>> intuitive abstraction of the more detailed components (objects) of
>> the system. This is very useful to someone designing the structure of
>> an XTM engine.

[sam]
>Does this translate to a claim that the UML notation is not suitable
>for specifying "the graph"? I think Eliot would disagree...

UML has Object and Collaboration diagrams and can be customized to
do just about anything.  However, "the graph" is not primarily meant
for hardcore programmers, so something with less baggage might
be better.  Something more philosopher friendly.

As someone outside the process, I don't feel like its my place to
comment on specifics, but it sounds like a lot of this stuff is
already done.

Guy


------------------------ Yahoo! Groups Sponsor ---------------------~-~>
Find software faster. Search more than 20,000
software solutions on KnowledgeStorm. Register
now and get started.
http://us.click.yahoo.com/556S5A/RNSCAA/2h4EAA/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/