Saturday, November 06, 2004
New discussion about CoreTalk and data regularity in context
On the Nature of science and
cultural Polemics à
On the discussion of data
regularity à
On the Stephenson Cyber Attack
Taxonomy à
Communication from John DeArmon
Paul,
I completely agree with a
functional and practical approach to implementing the basic Cubicon structure
directly. The transformations with the outside can take place as
needed.
Why not have a disciplined
approach to defining the structure itself in a grammar that can be
expressed and processed in a parallel context which at first can be understood
by a community representing Cubicon and a larger close community?
Language is only a tool which at some point has to be realized for completing
an implementation by some people, some where. As you point out, how would
you know how to ~optimally instantiate the icons~ without better
understanding how they operate with one another, or the outside world, in
a representative known abstraction? Otherwise you implement, see how things
work, and then implement again, iteratively. Too much time and resources,
and you continue to be unable to explain how it works at *some* communal level
in the beginning.
An example: If I told
you I had the next generation technology for lighting but you had to understand
my unknown physics, you may find it difficult to proceed. If I
simply said I have a low voltage, high radiance
LED technology that can near permanently replace light bulbs you
would be intrigued because it relates directly to you. Then, if I produced a
description of schematics and known physics/(related grammar/? in our
case) for implementation on specific identified devices you would open you
wallet if you were an investor.
Of course, if I could
demo an LED that would increase the value and close the deal. Getting it
as close to the existing real world, OR the differences, in a formal manner
helps all parties in understanding.
I am interested in your
collective views on these points.
Perhaps a different
computational operating model will persist given Linux, etc. but I do not see
any reason to dialogue on a non computational platform unless we can develop a
common schematic, language or other method of describing the behavior. It
is just not that difficult.
John
I look forward to comments
from the others; Richard Ballard and Peter Stephenson in particular. I talked to Sandy on the phone today, and he
reminded me that his input is stated clearly in the Macromedia presentation
http://www.bcngroup.org/beadgames/safeNet/CoreTalkIntro.zip
I also look forward to
continuing discussions with John on the notion of a universal transformational
theory between CoreTalk and any other language.
The generative methodology
that Adi has helped me to see is perhaps a methodology that will assist in the
development of this “Application Programmers Interface” between any code base
in any programming language to the behavioral theory of type that has become
agreed on by the CoreTalk user community.
I suspect that John has some specific ideas as to how to achieve a
universal transformational theory. As
an exercise, this is likely to be important to the programming community and
would provide a means to help the programmers make a transition to other lines
of work, as programming is reduced in complications to the point where
non-computer scientists are always the ones who actually “configure Cubicon
icons as new software is deployed.
In CoreTalk the
responsibility for agreement on a behavioral theory of type is laid at the feet
of user communities. In the past, up to
now, the responsibility has been laid at the feet of the programming community. The “utility function” for evolving optimal
and responsible code bases was hijacked by the interest of the programmers in
job security. CoreTalk changes this
fact on the ground. The large social issues
are now at the forefront of political capital expenditures by the American
Administration, and as such there is a selling point that the federal
government, and not parts of the IT Industry should be the supporter of the
completion and deployment of Core Systems.
I look forward to a discussion with John and members of the core team on
how to present this opportunity to the Administration, now that the elections
are over.
I should also talk a bit
about stratified theory and the grounding of a new information science based on
an understanding of the nature of algorithms.
The need is for grounding information theory in the natural world, and
not in the abstractions that comes from classical information theory or from
mathematics. One can review my general
comments on this elsewhere in the bead game threads. I will point only to Chapter 2 of the Foundations
of Knowledge Science so as to, hopefully, make this point that
abstraction is a process of “leaving something behind” as Richard has pointed
out recently.
The human must be
involved.
You have said, John (above)
<quote>
Perhaps a different
computational operating model will persist given Linux, etc. but I do not see
any reason to dialogue on a non computational platform unless we can develop a
common schematic, language or other method of describing the behavior. It
is just not that difficult.
</quote>
I have to ask how one
justifies the claim that a common schematic is not difficult. This seems to be a theoretically grounded
statement that does not reflect something, giving the past experiences since
the 1960s and 1970s.
Sandy’s CoreTalk concept has
fully worked out the social science issues in a fashion that is acceptable to
me, and will be acceptable to many social scientists. It seems to me that the technology compatibility to the software
industry is not the most important consideration. The most important consideration is that the information science
be more true to natural science.