Re: Congrats & trivia

[glen e. p. ropella]

> Ye Masters of the Hive,

We need to work out the defn of "Hive" at some point.  We
started using it when we started talking about the evolution
of the swarm users' community (about 2 or 3 weeks after I 
came on the project).  We were thinking that there would be
several hives, one here, one at umich, ... basically, anywhere
where there was more than one person using swarm.  Then, 
in a message I posted earlier, I referred to "2/3 of the hive
here at SFI".  But, that's not a correct thing to say because
we really have a hive of at least 4 right now (Chris, Barry,
JJ, and me).  Roger is usually referred to as a member of the
SFI Hive.  But, he actually holds dual citizenship in two 
hives, one here and one at John Deere (albeit a hive of 1, 
maybe).  And, JJ just got here and will be leaving in a little
while.  So, in the tradition of analyzing problems like "how
does one bootstrap a Swarm from a bunch of agents", I'd like
to ask the list:  What should "hive" mean?  Should it be virtual?
Should it be spatial?  Should it be "functional" (i.e. i'm an
anthropologist using Swarm and Bob's an archeologist using 
swarm, so we're in the same hive as distinguishable from the
hive that does simulation of manufacturing processes)?

Etc.  This is not a totally frivolous question. (I know, I know,
if it comes from me, it's frivolity is automatically increased. 
[grin])  But, it has implications as to how we cut the user
community.  We already have a glaring cut between professional
programmers who think multi-agent simulation is cool and 
scientists who think that multi-agent simulation is cool. (given
the assumption that one isn't usually a scientist and a programmer
at the same time)  But, that's not a useful cut for what Swarm
is.  Any thoughts?

> a) For those who are downloading the new release note that the tar file
> unpacks into a directory called swarm. You might want to be careful
> about where you unpack it, I managed to ovewrite old Makefiles and
> source directories when I unpacked my copy. The old binaries were not
> affected (dirty secret: my Linux box has no backup)
> 
> b) The link to necessary software points to the ftp directory. It might
> be useful to know more about the current status of compatability with
> versions of Tk/Tcl and BLT (the ftp directory contains new and old
> versions of Tk/Tcl and I noted that the Swarm Makefile.conf links
> against the new versions). I lost track of where the hive-meisters want
> us to go now.

Both of these are being taken care of (however slowly).  a) we
are definitely going to change the distribution to untar into
a directory with the version number appended.  And b) we're 
trying to create an autoconf configure script that will go
out on a system and check for what's there and what's not.  This
will provide the user with a reporting mechanism that will give
at least a little info on what Swarm expects to be present.

Now, as far as providing b) up front in the web page, I hadn't
intended on doing any more than what's already there.  But, you're
probably right in that there should be some indicator.

glen