Newsgroups: comp.ai.alife
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!gatech!swrinde!emory!mind.org!runic!thantos
From: thantos@runic.mind.org (Alexander Williams)
Subject: Re: Tierra Working Group Report
Organization: Runic Writings UUCP Link: Convoco Hasturam
Message-ID: <D8qtJF.w6@runic.mind.org>
References: <3ojf3i$235@gap.cco.caltech.edu> <3omves$4pc@gap.cco.caltech.edu> <jaboweryD8JnK6.JI5@netcom.com> <3p46e9$1mn@gap.cco.caltech.edu>
Date: Wed, 17 May 1995 22:09:13 GMT
Lines: 66

In an arcane scroll, Alexander Williams quotes the holy scripturist
C. Titus Brown, replying to the mystic words as written, saying:

>Remember, in net.tierra the puddles may have *completely* different environments
>in which solutions discovered in connected environments may have NO bearing
>whatsoever to the current environment.  This is usually not completely
>true in nature, where there is usually some correspondence between neighboring
>areas.  (I suppose one could make the argument that this is because life made
>it that way, and if that happened in tierra it would be *wonderful*...)

    I'm not sure this would be an extremely useful thing to have in
great abundance in the "Great Sea."  In a sense, they all have to
have reasonably similar traits, in using the same instruction set,
etc...  The major differences would seem to be in the size of the
population and any differences in the Reaper that may have been
hacked.  The former mimics the ways of nature pretty firmly, seeing
the population as a function of the "horsepower" available to the
Tierrans (food, in a sense).  The latter, the changes in the Reaper,
I don't see as altering the entire sense of the environment, but
rather behaving like different "biomes," within an environment (as
defined by instruction set).
   If I turn the "chlorophyll" option on in my Reaper then hook up
with the Great Sea, my Tierrans will tend to be less fit on
neighboring "islands" than others, possibly, but if they impliment
Reapers which favour reproduction mine doesn't, for some reason, I
may evolve Tierrans that grow in my environ, absorb energy through
"chlorophyll" and then shed the "chlorophyll" tail, migrate next
door, reproduce, then head back to mature and spawn.  I don't see
that so much as different "environments" in the grand sense, more
like the difference in "grasslands" (where photons fall widely and
provide lots of power) and "underwater" (where there's not much
light but plenty of nutrients that can be used in reproduction).

>>Some concept of a 2-d "terrain" needs to be incorporated such that
>>"puddles" with small cartesian distances are exponentially more 
>>likely to exchange genetic information than more remote.  The
>>parameter on this exponential decay will turn out to be as important as 
>>the mutation rate.
>
>One could almost think that I'd asked you to say that...
>
>That's exactly what we've done with tierra.  Take a look at Avida via WWW:
>http://www.krl.caltech.edu/avida/.  Unfortunately, due to the somewhat
>nasty computational requirements of these systems, we haven't gotten
>a truly huge grid going yet.  We've got someone working on it, though,
>(Johan Chu) and we've (very recently ;) gotten up to ~250,000 organisms.

   Another approach to the problem might be found in one of my
earlier posts, consider the relative "lag time" between transmission
and reception as well as the "bandwidth" of the link to represent
distance.  My little island at the end of a looooong, slow, narrow
UUCP link might be considered a considerable difference away while
the local island on the next workstation no longer away than a quick
TCP/IP ping might be considered to be a neighboring island whose
shores touch your own.  If you need the bandwith of the connection
narrow between the neighboring islands, it might be rationalized as
either a very rocky, dangerous trip or as longer.
   An advantage of this approach is that with the right docs, you
can use pathalias to create the routing config file.  Gotta' love
code reuse...

-- 
thantos@runic.mind.org (Alexander Williams)     | PGP 2.6 key avail
  Should we shed our mental pants and compare   | DF 22 16 CE CA 7F
  the size of our consciousnesses?              | 98 47 13 EE 8E EC
      Jan Sand to Marvin Minsky                 | 9C 2D 9B 9B
