[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Preview of Swarm-hive activities...
From: |
cgl |
Subject: |
Preview of Swarm-hive activities... |
Date: |
Sat, 5 Oct 96 08:17:26 MDT |
Hello folks!
In the last month, we've been undergoing an intense review
of where we have gotten to with Swarm, and where we should
be aiming to go in the near and long-term future.
The result of our internal review was, basicly, that we're
satisfied that, with a few remaining tweaks here and there,
the basic foundation of swarm, the kernel, is essentially
done, and that it is broad enough and provides enough
support for just about everything we want to do from
here on in. We've *really* appreciated and benefitted
from the hard-work that a large number of Beta-users
have undertaken to implement their own models in Swarm,
and the feedback they've given us has been critical in
keeping us from getting lost in the details.
So - the current status is as follows...
The foundation of Swarm is in place, and most of our
attention over the next year will be focused on the
multiple constructs that can and should be built on
that foundation - in a sense, we've been waiting for
Swarm to be ready too! Much of the higher level
stuff that we want to provide will run on the
Swarm kernel as well, and we've been focusing on
getting the foundation firmed up before we
turned to the first, second, and etc. stories....
But...we`ve been plotting and planning and architecting
those higher stories all this time, and now we get
to implement them. Much of this work will take the
form of adding extensively to the libraries, providing
a much larger spectrum of building-block objects
that people can start with and use to build their own
higher level constructs to suit their own application
domains.
In the next week, we'll be sending out a number of
documents detailing our plans for the short and
long term future....
Specifically, we'll be posting (in approximately
this order)
1) A detailed statement of the library objects
we plan to add in the near future (including
visualization tools, analysis tools,
experiment management tools, tools
for interacting with other processes...etc.
2) An extensive Beta-survey to reap the collective
wisdom and experience and hopes and dreams
(and maybe even nightmares!) of the folks
who have done battle with the beta versions
of swarm.
3) An announcement of the first Swarm-Users meeting
to be held here in Santa Fe probably mid-February
next year, during which Beta-users will be able
to meet each other and us, and show off their
models to each other, and for which we will
be preparing a number of tutorials on the
use of swarm and what we consider to be
proper approaches to simulation yielding
comparability and repeatability of results, etc.
4) A restructuring of our Web pages to provide
much more support for user-community interactions
(we will be providing a user-index area where
folks can provide links to their stuff, managed
by a search engine so other users can find who's
doing stuff relevant to their work. We'll also
have a user-contribution area through which folks
can kick-back tools and library objects that
they've developed which might be useful to
other folks out there....and so forth.)
5) Our plans for converting from Beta mode to
V1.0 release mode, and some suggestions and
queries about how we all - developers and users -
can best continue to develop and maintain Swarm
in the long-term future, committed as we are
to public-domain distribution of Swarm and
the "grassroots" community model of development.
So - the relative silence over the last month or so was
actually accompanied by a lot of intense introspection and
planning on our part here. We will now be putting out
incremental releases of the kernel, libs, and apps on
a more frequent basis and, hopefully, be rapidly increasing
the utility and ease-of-use of Swarm.
Stay tuned!
Chris Langton
for the SFI Swarm hive (Roger, Manor, and Glen)
- Preview of Swarm-hive activities...,
cgl <=