[om-list] Re: The Rise of Worse is Better

Mon Dec 16 10:31:42 EST 2002

Mark and Luke

    This is good stuff.  I'm glad we still have our OM group.  I think
something will come of it some day.

    I think I promised to tell you guys more about Soar when I learned more
myself.  I still don't know too much about it, but I'm pretty happy with
what I do know.  It seems like a very good alternative to trying to develop
my MTShell ideas from scratch.  In fact, it's rather similar to MTShell in
some ways: Tcl scripting glue connecting C modules together in a fairly
sophisticated and yet general-purpose AI architecture.  -- and it's a
framework that several organisations are adopting.  It's open source.

    For a video introduction to Soar, follow this link (and try to use
QuickTime to view it, since I could not get RealPlayer to play it):

    http://ritter.ist.psu.edu/papers/soar-mov.mpg

    Below are some other links and one accomplishment.

tomp

~~~~~~~~~~~~~~~~~~~~~~~~~~~~

"SOAR" = State, Operator, And Result.

"Perhaps the largest success for Soar has been flying simulated aircraft in
a hostile environment. Jones et al. (Jones, Laird, Nielsen, Coulter, Kenny,
& Koss, 1999) report how Soar flew all of the US aircraft in the 48 hour
STOW'97 exercise. The simulated pilots talked with each other and ground
control, and carried out 95% of the missions successfully. "  (From
http://acs.ist.psu.edu/soar-faq/soar-faq7-01.html#G2.)

NL-Soar Pages:
-------------------

NL-Soar Home Page:

http://humanities.byu.edu/nlsoar/
http://humanities.byu.edu/nlsoar/homepage.html

NL-Soar FAQ:

http://humanities.byu.edu/nlsoar/nls2002faq.html

NL-Soar Tutorials

http://humanities.byu.edu/nlsoar/nls2002tutorials.html

Installing NL-Soar:

http://humanities.byu.edu/nlsoar/installhowto.html

NL-Soar Glossary:

http://humanities.byu.edu/nlsoar/nls2002glossary.html

Old CMU Description of NL-Soar:

http://www-2.cs.cmu.edu/afs/cs.cmu.edu/project/soar/utc/nl/doc/nl-homepage.h
tml

Soar Pages:
--------------

Soar FAQ:
http://ritter.ist.psu.edu/soar-faq/

Introduction to a collection of published papers spanning the first decade
of Soar research and use.
http://www.isi.edu/soar/papers/soar-papers-book/soar-papers.html

General Soar Home Page.
http://ai.eecs.umich.edu/soar/

General Soar Documentation.
http://ai.eecs.umich.edu/soar/docs.html

Soar Architecture Descriptions:
http://ai.eecs.umich.edu/cogarch0/soar/index.html

Soar Group at University of Southern California:
http://www.isi.edu/soar/soar-homepage.html

Soar Group at University of Michigan:
http://ai.eecs.umich.edu/soar/soar-group.html

Soar Group at Carnegie Mellon University:
http://www-2.cs.cmu.edu/afs/cs/project/soar/public/www/home-page.html

The appendum to a Soar FAQ.
http://www.nottingham.ac.uk/pub/soar/nottingham/soar-lfaq.html

A comparison between Soar and Act-R, concentrating on their theories of
control.
http://www.sahs.uth.tmc.edu.edu/TRJohnson/papers/CogSci97-Control-in-Act-R-a
nd-Soar.html

Act-R Pages:
---------------

http://act-r.psy.cmu.edu/
Act-R home page.

http://act-r.psy.cmu.edu/tutorials/
Act-R home page Tutorials including "Integrated Theory of the Mind".

Other Related Links:
---------------------

Wordnet:
http://www.cogsci.princeton.edu/~wn/

Soar in Commerce:
--------------------

ERS: Explore Reasoning Systems, Inc.
http://www.ers.com/Html/Products.htm

Soar Tech
http://www.soartech.com/

~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Víðar sum quem nihil obstat.
www.Ontolog.Com
~~~~~~~~~~~~~~~~~~~~~~~~~~~~

----- Original Message -----
From: "Mark Butler" <butlerm at middle.net>
To: "One Model List" <om-list at onemodel.org>
Cc: "Luke Call" <lacall at onemodel.org>
Sent: Saturday, 07 December, 2002 22:35
Subject: [om-list] Re: The Rise of Worse is Better

"Worse-is-better" is a pretty provocative name for this phenomenon, but
I agree with the basic principle.  This reminds me of a scripture:

     For which of you, desiring to build a tower, does not first sit
    down and count the cost, whether he has enough to complete it?
    Otherwise, when he has laid a foundation, and is not able to finish,
    all who see it begin to mock him, saying, 'This man began to build,
    and was not able to finish.' - Luke 14:28-29

I don't think that this means we should never build towers, but rather
we should scale the tower to our resources - not just monetary, but
mental and social resources as well.  In software development this often
means incremental refinement rather than trying to accomplish everything
in one big bang (as admirable as that may be).  Software technology is
so incredibly complex that it often seems that we are barely
accomplishing things that we knew how to do thirty years ago - not
because we don't know how, but rather because of the sheer effort and
mutual cooperation involved.

I think the genius of the open source software movement is not in
producing software that is necessarily better in every respect than
closed software - in narrow fields that may never be the case - but
rather in creating a common technological base that anyone can build on.
The interesting thing is building a consensus tends to be a very
conservative process - Unix / C represent consensus not because they are
the leading edge, but rather because they are well understood by the
greatest number of people.

Lots of observers lament that no new programming language can be really
successful unless it has syntax similar to C. But is this a bad thing?
 Would anyone expect a replacement for English to be successful
overnight if it had no grammar or vocabulary in common?  Neither C++,
Java, nor C# are the be-all-and-end-all of programming languages, but
they are successful because they are evolutions rather than revolutions.

Revolutions are nice if you can pull them off, but generally rarely
succeed. This is particularly true in infrastructure software like
operating systems, languages, and databases, though considerably less so
in application software.  You might say that both Unix and American
democracy are classic examples of "worse-is-better".  Making an
improvement to either is not so much a matter of technical superiority,
but rather one of persuading everyone else to go along.

  - Mark

Luke Call wrote:

> Thanks Mark, that is an enjoyable read and one of those things likely
> to get filed away mentally for later pondering or mental connections.
> Getting something working does seem much more successful than getting
> it "perfect" too, and Unix/C people in a way seem like the ultimate
> pragmatists, for some things. And the answer for which is the "right"
> tool to use still seems to be always "it depends".....
>
> Just my first, pedestrian, ruminations....
>
> Luke
>
>

_______________________________________________
om-list mailing list
om-list at onemodel.org
http://www.pairlist.net/mailman/listinfo/om-list