RE: Big Science Needs Your Idle Cycles **SETI UPDATE**

Nelson Minar (nelson@media.mit.edu)
Mon, 23 Jun 1997 23:45:11 -0400


>>The simple answer is that Java is only going to be used for the control
>>and visualization.

That's a reasonable compromise. It does require that everyone trust
the author of the SETI search code, though, and that the authors be
willing to port it to all the architectures around. Similar problems
we faced in deschall, I might add, and they weren't insurmountable.

>I'm keeping the rc5 running but when SETI launches, I'm there dude.

One question about the SETI thing - what reward do people get for
participating? I think one reason that deschall worked, at least for
me, was that it was *fun* to participate. I checked the stats every
day. I loved seeing the milestones, I liked the competition between
sites, I got a warm fuzzy feeling everytime I could turn an idle
machine over to it. All that because we had a measurable goal.

The SETI project doesn't have that. There's not much we're going to
see in progress and there's no guarantee that the haystack even has a
needle in it. Any ideas how they're addressing the question of interest?

(Another topic of discussion for a general mailing list on loosely
coordinated distributed computations. Hrm.)

PS - for another take on SETI type activities, read the novel "His
Master's Voice" for Stanislaw Lem. It addresses the question of
supposing we did find an alien signal, how would we understand it?