From: thunk <gmkoller@...>
Date: 2010-03-28T11:15:05+09:00
Subject: Re: Boid writeup idea


Sorry,

I just reread my posting that I fired off a little to fast.  I meant
to write:

I'm not ashamed, rather I'm PROUD that I have NOT had to use Prolog or
something to get useful work done.

and that's not exactly right either, I'd love to use Ruleby or the
Prolog - but its encouraging that those brilliant tools are waiting in
the wings to take things farther....  I don't need them yet, and
that's kinda nice for my sanity.

THEN I want to mention that their are decisionless Boids that are
effectively reviews - they don't need to make any decisions - they
just tell you what somebody thinks about something.  So there are
different classes of Boids (think of Ants - that helps me)....

BUT then my point is that mostly, right now I'm laying a foundation of
"System" boids that lay a foundation for the "free form" boids.   This
will vary / I know this from experience already - greatly from domain
to domain.  But system boids are authored by programmers / aand are
ahhh must be "systematic" - as in one per Drug or so.  I have played
with the idea just weeks ago of "SuperBoids" the spawn themselves -
not too much of a trick BUT I think the concept of One Boid per
Assertion set leaves a concrete set of code where something can be
different if/when the time comes.  that seems important to me.  BUT
clearly, the Boids need not be authored one at a time, I just don't
want to spend the time just now to write that code before I have more
"testing" done on the other parts.....

My point is that I am developing "HElperClasses" for sub-domain issues
to do certain things - another point getting lost a little in this -
but they are just the foundation of it all, not the essence of it
all.

More later, I have to do.....


The SwarmShepherd,


thunk