[Martin Taylor 2017.09.04.09.56]
[From Bruce Nevin (2017.09.03.13:36 ET)]
Well said, Dag. Succinct and to the point, and grounded in
experimental method, not just theory. 
However, over time a number of people have complained about
limitations of our listserv environment. You who have voiced
those complaints, it would be helpful if you were to add your
voice to this discussion now. Control requires clear and
explicit perception of what is and of what is desired.
/Bruce
Well, going from my ECACS experience, three useful facilities not
available from CSGnet are (1) continuity of concept over a thread
that can be be months or years between messages. (2) a modular
structure of topic areas, and (3) the ability to cross-link
(hyperlink) messages across threads and topic areas. A global search
within topic area or more wide-spread is also quite useful.
Facility (1) would avoid the kind of periodic recycling of arguments
that happens on CSGnet because of its very short institutional
memory. Nobody is likely to search the archives to see whether topic
X has been debated before, but if someone interested in a topic
knows it is seems as though it belongs in, say,
“Theory->Interacting Control Systems->Social”, it’s not so
hard to find that, for example, the interactions among collective
control modules was discussed seven years ago, and see whether any
useful conclusions were reached then, or whether questions remain
open.
Facility (2) allows people to see whether their topic of today's
interest seems to have a place within the existing structure or
whether it suggests that maybe a new second-level subtopic could be
useful. Probably an administrative decision would be necessary at
higher topic levels. On ECACS, administrators occasionally
subdivided threads when the conversations seemed to be diverging
into a new area. There’s an instance of this in the “Stigmergic
Systems” example I mention at the end of this message. The point
here is that the Forum should have a structure reminiscent of the
control hierarchy, with some kind of administrative metaphor for
reorganization.
Facility (3) turns the thread structure into a network or mesh, and
allows topics that initially seemed conceptually separate to be
merged, as, for example might happen if Rupert built a robot based
on some theory discussed in the basic theory area and described it
in the “Application development” or “Hardware Implementation” topic
area.
I think text search can be useful, but it's often hard to find ways
to describe what you are looking for with a net wide enough to find
it and narrow enough not to find dozens of irrelevancies. Topic
structure and hyperlinking takes care of some of it. Search within
topic refines the search to remove a lot of irrelevancies, and
hyperlinking may help to find something related to what you found
that was not exactly what you were looking for.
Earlier, I suggested that a Wiki might be useful, and someone
pointed out that there’s a PCT area on Wikipedia. That’s something a
bit different. Wikipedia wants things that are academically secure
and buttressed by peer reviewed (or at least published) material.
Wikipedia does not allow original material. The CSGnet replacement
is a place for reporting theoretical developments, simulations,
physical and biological implementations, etc., all things that are
not suited to Wikipedia but that are suited to a specialized Wiki
intended for original material. Cross-topic hyperlinking might serve
some of the same function, especially if an area of the forum were
provided for this “nexus” function. But maybe that’s a bit too much
to ask.
Here's an example of one fairly typical ECACS sub-topic area under
(simplified topic titles) “Current_Interests → Theory →
Mathematical Dynamics”. It has 171 messages under the (actual)
headings “Stigmergic Systems” “A starter for brain dynamics”
“Attractors in the brain” “Emotion, neurobiology, and dynamic
systems”.
All the messages in those sub-topics were posted in 2004-5, but the
threads remain (or remained until 2014) open for further
contributions and for back-linking from new messages on other
topics, perhaps from new members with different idea, perhaps
because new knowledge from other sources changes something about
what was then understood by the contributors. “Stigmergic Systems”,
for example, though not using those words, has been a matter of
CSGnet interest from time to time recently. Maybe something
discussed then would have been useful to the current discussion. On
re-reading that old discussion now, I find that I would like to
continue it as I think I understand better how stigmergy relates to
PCT than I did about 12-13 years ago.
Martin
···
On Sun, Sep 3, 2017 at 11:45 AM,
Dag Forssell csgarchive@pctresources.com
wrote:
[Dag
Forssell (2017.09.01 15:30 PDT)]
Rick, If you have not followed this thread with care,
or not at all, or don’t quite remember, please just
review this thread at the forum archive.
If you cannot do that with ease for this thread or any
other you want to catch up with, there is your answer.
Best, Dag
[From Rick Marken (2017.09.02.1820)]
RM: Could someone please remind me why we're
doing this?
Best
Rick
