RE: Big Science Needs Your Idle Cycles

Nelson Minar (nelson@media.mit.edu)
Mon, 23 Jun 1997 14:18:23 -0400


>>This project is just getting started and already looks First Class!
>>They are looking for volunteers and aliens. Check it out at:
>>http://www.bigscience.com/
>Yes, this project looks very cool indeed.

I agree. The neatest idea is the picture of the globe, who all is
working on the computation at the moment. I'm skeptical of how
successful they'll be, though. Why did people participate in deschall?
Will the motivation be enough to sustain them through lots of other
computations?

>However, it looks like the clients will be in Java?!
>I'm more than a little skeptical about that.

I'm not. Java is not terribly fast (although it's not terribly bad for
Windows). But it has a lot of important features that make distributed
computations a lot easier. In particular, the security and portability
of the virtual machine is a big win.

For my thoughts on that, see my paper "Computational Media for Mobile
Agents" at
http://nelson.www.media.mit.edu/people/nelson/research/dc/

My favourite aspect of the deschall crack has been all the talk about
loosely coordinated distributed computation. Are there any mailing
lists that have that sort of discussion? Should we start one, bring in
the bigscience and distributed.net folks as well?