SIMULATION DIGEST V14 N7

SIMULATION DIGEST V14 N7

Post by Moderator: Paul Fishwi » Wed, 14 Mar 1990 23:32:09



Volume: 14, Issue: 7, Tue Mar 13 09:31:59 EST 1990

+----------------+
| TODAY'S TOPICS |
+----------------+

(1) Software for Simulating Parallel Systems
(2) CALL: Model Based Reasoning
(3) RE: Simulation in LISP
(4) HIRES LISP-based Simulation Package
(5) Request for Info: Simulation Companies

* Moderator: Paul Fishwick, Univ. of Florida
* Send topical mail to: simulat...@bikini.cis.ufl.edu OR
  post to comp.simulation via USENET
* Archives available via FTP to bikini.cis.ufl.edu (128.227.224.1).
  Login as 'ftp', use your last name as the password, change
  directory to pub/simdigest. Do 'type binary' before any file xfers.
* Simulation Tools available by doing above and changing the
  directory to pub/simdigest/tools.

-----------------------------------------------------------------------------

To: simulat...@ufl.edu
Path: ukma!douglas
From: John Douglas Turner <doug...@ms.uky.edu>
Newsgroups: comp.simulation
Subject: Simulation of parallel systems (hardware)
Date: 8 Mar 90 18:49:12 GMT
Reply-To: John Douglas Turner <doug...@ms.uky.edu>
Organization: U of Kentucky, Mathematical Sciences
Sender: doug...@ms.uky.edu

I am interested in the status of software for simulating parallel systems.
Both from low level hardware to simple processor simulation.

john.....

--
John Douglas Turner             doug...@ms.uky.edu   or  doug...@UKMA.BITNET
University of Kentucky                  {rutgers,uunet}!ukma!douglas
902 Patterson Office Tower                      (606)-257-6824
Lexington, Ky.  40502          

------------------------------

Date: Fri, 9 Mar 90 18:12:59 PST
From: Ethan Scarl <et...@atc.boeing.com>
To: AILIST-REQU...@AI.AI.MIT.EDU, Owners-cl...@Xerox.COM,
        arpanet-bboa...@mc.lcs.mit.edu, comp...@ucbvax.berkeley.edu,
        fj-ai%etl...@relay.cs.net, fox%vtcs1.cs.vt....@relay.cs.net,
        jws%ib.rl.ac...@nss.cs.ucl.ac.uk, mcvax!swivax!ot...@uunet.uu.net,
        mod-ki%gmdzi%unido.u...@uunet.uu.net, nl...@rochester.arpa,
        n...@nnsc.nsf.net, simulation-requ...@bikini.cis.ufl.edu
Subject: MBR Workshop at AAAI-90.  Please post as appropriate
Cc: et...@atc.boeing.com

                          * CALL FOR PARTICIPATION *

                           Second AAAI Workshop on

                            MODEL-BASED REASONING

Boston, MA, USA          Monday, July 30, 1990               Sponsored by AAAI  

There are advantages for design, analysis, monitoring, diagnosis, and control
in systematically representing and reasoning about the structure and function
of a system.  This workshop is a forum for defining and exploring issues
associated with reasoning from knowledge of a target system's structure and
function, more recently called Model-Based Reasoning (MBR).  This includes
control, planning, analysis, theory construction, tutoring, and
design/redesign, as well as the more familiar areas of monitoring and
diagnosis.

The workshop seeks papers which offering insight into fundamental problems
associated with MBR, some of which are listed below.  Although analyses of
experiences with MBR may be of interest, this is not a forum for discussing
system implementations or applications.

We encourage papers which contribute to the understanding of issues such as
the following:

* The range of inference mechanisms that exploit explicit structural and
   behavioral knowledge; their power and limitations; their requirements upon
   structural representations and behavioral simulations; factors leading to
   the success or abandonment of MBR
* Target system complexity as a fundamental obstacle; types of complexity; the
   use of structure in highly interconnected systems; the absence of physical
   structure
* Qualitative/quantitative, analog/discrete, and mixed domains; coping with
   computational limits and speed/accuracy tradeoffs; hierarchical and other
   organizational principles; geometric knowledge; feedback or other
   time-dependencies; embedded software
* Principles of model construction; problems and solutions in knowledge
   acquisition; functional/physical model relationships; uncertainty or
   incompleteness in models; mixed heuristic/model knowledge
* Inference of structure from behavior; open systems or environments

To facilitate discussion, acceptance will be limited to 30 papers and 10 or
fewer presentations.  There will be a proceedings of all accepted submissions.
More formal publication will be pursued if there is sufficient interest.
There will be a commentator for each presented paper, and single paragraph
abstracts of each attendee's paper or work in MBR will be distributed prior to
the workshop.

If sufficient interest is expressed by participants, there may be an
evening session for "system" demonstrations on video or slides.

The workshop's organizing committee consists of: Randy Davis, MIT; Johan de
Kleer, Xerox PARC; Richard Doyle, JPL; Dan Dvorak, U of Texas; David Franke, U
of Texas; Ben Kuipers, U of Texas; Ethan Scarl (chair), Boeing Computer
Services; Mark Shirley, Xerox PARC; Dan Weld, U of Washington; and Brian
Williams, Xerox PARC.

The submission format will be short, self-contained papers of no more than
2000 words (the proceedings version of accepted papers may be longer).
Extensions and condensations of work already submitted to AAAI or elsewhere is
permissible, so long as duplications are noted.  Submissions should include
postal addresses, email addresses, and phone numbers for all authors, plus a
stamped self-addressed card if such acknowledgement is desired.  Three copies
of submitted papers should be sent to Ethan Scarl at Boeing Computer Services,
P.O. Box 24346, M/S 7L-64 Seattle, Washington, 98124-0346, (206) 865-3255,
865-2964 [fax].  Papers without illustrations may be sent via email as latex
or plain ASCII to et...@atc.boeing.com.

Important dates:   April 2, 1990           Papers must be received
                   May 21, 1990            Notification of acceptance mailed
                   June 15, 1990           Abstracts due for early distribution
                   July 10, 1990           Revised papers must be received
                   July 30, 1990           Workshop

------------------------------

Date: Sat, 10 Mar 90 22:03:07 MST
From: roberts%studgu...@LANL.GOV (Doug Roberts @ Los Alamos National Laboratory)
To: c...@aio.jesnet.jsc.nasa.gov
Cc: simulat...@bikini.cis.ufl.edu
In-Reply-To: Brian Cox's message of Fri, 2 Mar 90 10:40:24 CST
Subject: Discrete Sim in LISP

> Does anyone know of any discrete event simulators written in
> LISP?  We are developing a discrete event simulation and
> qualitative modeling tool which is currently developed on
> top of KEE and SimKit. We are going to reimplement off of
> KEE and SimKit and would prefer not be tied down to a vendor.
> So any info about public domain discrete event simulation
> code in LISP would be great. Thanks.

Smart decision, getting away from those vendors (Intellicorp). We also
build discrete event simulations on top of KEE: we've been doing it
here for about five years now. We never used Simkit because we built
our own contoller, queue manager, etc. before Intellicorp had such a
product.  That turned out to have been a very smart decision, given
that Simkit is even less well supported than KEE.

By now it may be slightly obvious that we are less than enchanted with
Intellicorp's support. We report bugs, they ignore us. We report more
bugs, ...

We are now investigating alternative OO environments in which to
develop our applications. Among the these are CLOS/X/CLX/??, C++/X/??,
ART, KnowledgeCraft, 386/MSDOS(Shudder)/Actor... Unfortunately, we
haven't the manpower at the moment to devote full-time towards
building an acceptable alternative environment, but we _are_ looking.

--Doug

================================================================
Douglas Roberts                |
Los Alamos National Laboratory |I can resist anything
Box 1663, MS F-609             |  except temptation.
Los Alamos, New Mexico 87545   |  ...
(505)667-4569                  |Oscar Wilde
d...@lanl.gov                  |
================================================================

------------------------------

Date: Tue, 13 Mar 90 09:25:46 -0500
From: Paul Fishwick <fishw...@fish.cis.ufl.edu>
To: simulat...@ufl.edu
Subject: RE: Simulation in LISP
Cc: bad...@cis.upenn.edu

This message relates to the query about LISP based simulation packages.
There are a few packages that I can suggest:

 (1) DEVS Scheme - University of Arizona (Bernard Zeigler).
 (2) ROSS - The Rand Corporation (Phil Klahr and Dave McArthur).
 (3) HIRES - Univ. of Pennsylvania/ Univ. of Florida (Paul Fishwick).

I can say a little about HIRES. Perhaps other readers can speak of
the other LISP/SCHEME based packages. HIRES has the following characteristics:

 (1) All simulations were carried out using a production rule formalism.
     Production rules were, therefore, the "assembly language" of HIRES.
     The implementation language is Common Lisp.
 (2) High level compilers (pre-processors) permitted the creation and
     integration of models based on Petri nets, animation scripts, finite
     state automata, discrete event modeling, continuous (diff. eq.) modeling,
     and expert systems incorporating time.
 (3) Dynamic model switching was implemented to permit simulation models
     to be created as "databases" containing connected models at differing
     levels of abstraction. In this way, it was possible to switch
     abstraction levels "on the fly" according to a conditional
     specification.
 (4) Combined modeling in HIRES was a special case of a much more general
     abstraction network capability, where the process abstraction network
     (i.e. database) was defined as one link connecting two nodes (one
     discrete event node, the other --- continuous).

Concerning the current state of HIRES - I have not worked on this package
for a few years. The human motion simulation group headed by Norm
Badler at the Univ. of Pennsylvania would probably be the best source
for the most recent version. Also, Paul Birkel at Carnegie Mellon Univ.
did some great speed and structural enhancements.

-paul fishwick

------------------------------

To: uunet!comp-simulat...@uunet.UU.NET
Path: intek01!matthew
From: intek01!matt...@uunet.UU.NET (Matthew Kayes)
Newsgroups: comp.simulation
Subject: Names of Simulation Co.
Date: 12 Mar 90 19:28:05 GMT
Organization: Integration Technologies Inc. (Intek), Bellevue WA

I am looking for information on the following simulation products
and/or companies.  If someone would please send me an address or
telephone number, it would be much appriciated.

        Siman/Cimima
        Slam
        Mask
        ProMod (production modeling)
        CimSoft

Thank you for your help.

Matthew Kayes
206-455-9935

------------------------------

END OF SIMULATION DIGEST
************************