Back to the Future of Swarm

[glen e. p. ropella]

Ken Cline writes:
 > I thought this might be premature, but in light of the
 > discussions about where to have SwarmFest '98...

Not at all.  We need to start bandying about ideas.

 > I would like to start up the "Future of Swarm" discussion.
 > 
 > At the very least maybe Chris, or any Hive member, would
 > post a description of all the options that are considered
 > serious possibilities.  This could be for the benefit of
 > those who missed SwarmFest, as well as those of us still
 > trying to sort out the implications of each of the four
 > options mentioned during the workshop.

Well, I'll give a first rough estimate and maybe Chris or Roger will
follow with some clarifications.  And, as usual, I'll tell it with all
the bias associated with my personal view of the situation.  [grin]

We run out of funding September '97.  And since SFI is a research
institution, keeping Swarm on as simply a software development effort
doesn't make sense.  So, it is reasonable that Swarm find another
avenue for its continued development.  However, SFI might remain
interested both in the end product "Swarm" and whatever research is
done either using it or within it.  That means that projects running
at SFI will probably be using Swarm and there may be certain avenues
of research into computing that would go on at the institute and be
fed back into Swarm.

Now, there are several paths we could take.  I'll try to list them and
give hints about how they might be organized; but, I'm sure i'll leave
out alot, so feel free to correct me or add onto what I say.

1. Move Swarm to another educational institution.  This would involve
trying to find some university or research institute willing to help
provide resources in return for any grant or contract money Swarm
might be able to muster.  The institution could be anywhere in the
world.

2. Form a non-profit organization for the development and maintenance
of Swarm.  There are several shapes an entity like this could take.
One of the favorites is similar to the X Consortium, where financial
entities pay to belong to the consortium in return for control over
the development of the software package.  Another one might be
something like the FSF.

3. Form a commercial company that would sell simulation consulting,
application software (that runs on top of Swarm), training, etc.  The
ideal relationship between the company and Swarm would be that the
company would have a vested interest in keeping the public package as
up to date as possible in order to optimize the company's
effectiveness for it's customers.

4. A hybrid approach of options 1-3.  For instance, we could form a
non-profit org to handle the free software package and the user
community and a sister company that focused on application building.
Or pass the software to an edu institution for r&d and have the same
type of sister company.


The *important* thing to remember here is that the goal is to keep the
free software that we're calling Swarm free and as cutting edge as
possible.  That's the point.  Even in the formation of the company,
a significant part  of the profit from that company should go to
developing and maintaining Swarm.  In other models where something like
this occurred, this was the hardest thing to ensure.

There are several pros and cons to each scheme; but, I'll let somebody
else talk now. [grin]

glen