[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Dropping agents?
From: |
Barry McMullin |
Subject: |
Dropping agents? |
Date: |
Sat, 28 Sep 1996 11:08:35 +0000 |
Does anybody know of any general problems with *dropping* agents
in the course of a swarm simulation?
I *know* that if there is a probe on an agent, and the agent is
dropped, that will typically (always?) cause a crash. I also
know the swarm team have plans to put in some more sophisticated
memory management machinery to deal with this problem.
But I now have something different in mind...
I have a toy application, derived from heatbugs. I want to
arrange for bugs to (occasionally) disappear - be dropped.
I was alert to the following:
o The bug had better be removed itself from its space.
o The bug had better be removed from the list of bugs maintained
by the model swarm.
o Of course, for the time being, I'd better not have any probes
on any bug that may have to be dropped.
o *Then* it should be safe to drop???
o NB: I do *not* take any special measures to do with "unscheduling"
any activities on the bug - because my understanding is that this is
being driven from the bug list; so I (naively?) assume removing
the bug from that list looks after this....
But: the simulation crashes the first time a drop is tried.
My initial (wild?) conjecture is that the activity scheduler, which
traverses the bug list (sending them each a -step message) gets confused
if the "currently stepping" bug gets dropped from this list. I
have not attempted to study this possibility in any detail yet -
and the real explanation may yet be infinitely simpler. But before
rolling up my sleeves and getting down into the swa(r)mp, I thought
I'd ask if anyone else out there has models with agents being
dynamically created and *dropped*, and, if so, whether you might
have any clues for me?
Thanks,
Barry,
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
| Barry McMullin, Autonomous Systems Group, | address@hidden |
| School of Electronic Engineering, | Voice: +353-1-704-5432 |
| Dublin City University, Dublin 9, IRELAND. | FAX: +353-1-704-5508 |
| http://www.eeng.dcu.ie/~mcmullin/home.html | |
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
- Dropping agents?,
Barry McMullin <=