[
Lists Home |
Date Index |
Thread Index
]
Hi Thomas,
Thomas said:
Note that there is nothing here about randomness. The random aspects of
a queue (which Roger suggested in a previouspost) do not qualify it as
such a system.
Didier replies:
Thomas you are right. This is a stochastic phenomenon.
Thomas said:
It seems that Roger is moving towards inventing an XML-Simula...
Didier replies:
Again, right on target...
In another experiment, as you know, I developed an OO language based on XML.
The object state is expressed with XML elements and the objects behavior
with ECMAScript. With that language I made some experiments with adaptive
object models (ref: http://adaptiveobjectmodel.com ). What we are now
experimenting with Roger is using XSLT as a transformation rule language to
modify the state of a set of objects. We are using XSLT transformation in a
processing loop or cycles to simulate a system. If we pay close attention to
the structure we can notice that:
a) the entities states and the transformation rules are expressed in two
different files and may be expressed in two different language. I will
publish quite soon the specs of the XML OO language I designed and will
translate Roger's experiment into that language, both under the prototype
theory and class theory paradigms (the language supports both).
b) Up to now, we saw two kind of processing thread:
b-1) single run, for example an XML document transformed into HTML using
an XSLT template (one transformation applied to the original document).
b-2) pipeline, for example an XML document transformed into another
model, then transformed into HTML (two transformation applied to the
original document)
We are now adding a third processing model: the loop.
What is still to experiment is conditional processing. Probably the next
step is to create the language that will express these different processing
model and we will have a strange beast ready to model and process complex
adaptive systems :-)
Cheers
Didier PH Martin
http://didier-martin.com
|
