swarm-support
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

preview of upcoming changes


From: Nelson Minar
Subject: preview of upcoming changes
Date: Thu, 16 Nov 95 12:22:23 MST

After some much needed vacation, and lots of discussion about what to
do next, we're getting ready to put out a new version of Swarm in the
near future (one to two weeks). Here's roughly what's planned for that
next version:

Gcc 2.7: Swarm will require gcc 2.7 - there are some incompatible
  changes in the Objective C runtime that require a switch. We're
  currently developing with gcc 2.7.0, but 2.7.1 is out on
  ftp://prep.ai.mit.edu/pub/gnu. The Objective C changes are minor,
  so Swarm should work with both versions (modulo changes to the C
  compiler on your particular machine.)

Activity library: small revisions, a few messages names are changing,
  SwarmPlan is being simplified.

New Swarm Object: a "Swarm" as an object is getting higher status.
  A Swarm is used to encapsulate a running schedule structure with
  and extra agent code (agent as Swarm with schedule). This will be
  useful for multiple Swarm simulations.
 
Bug fixes: the goal is all the problems our beta users have
  encountered in the last release will be fixed in this one. I don't
  have a clear idea what all those problems were, but I'll go through
  my mail archives and also look at surveys.

Makefiles: the makefile structure is going to be made a bit more
  flexible: allow you to choose which swarm libs to link into an app,
  change the way include files are installed for more simplicity.

Clean up probe: the Probe facility is being expanded to also allow
  message calls for probe. The "DataStuff" hacks in heatbugs are being
  generalized, so you can (begin to) do data collection on simulations
  in a principled fashion.


Some of the upcoming things after this version are:

Experiment with multiple Swarms
Use multiple Swarms to drive simulation startup and control
More docs and code in defobj and collections
Profile the code, find the worst bottlenecks
More data analysis support

-- 
                                __                      
address@hidden              \/              http://www.santafe.edu/~nelson/
PGP key 9D719FAD   Fingerprint 3B 9B 8E 58 1C 90 57 3E  B7 99 ED 13 65 2E 0B 24


reply via email to

[Prev in Thread] Current Thread [Next in Thread]