Re: Distributed networks?

["Robert J. Bradbury" <bradbury@aeiveos.com>]

I agree, sounds like someone needs to document the data
requirements/camera/night.

I just checked my line (the slowest DSL rate the Qwest
will provide).  It gives me ~25KB sustained transfer rate.
That works out to something like 7.5 hours per CD.

Both satellite and cable are much faster (though the latency
over satellite is higher we aren't very concerned with that).
I've setup a satellite (for TV) before and it isn't really difficult.

The only "hitch" I can imagine is if the data link provider and/or
ones ISP constrain or charge for high amounts of data traffic on
a sustained basis.

But according to /. this morning the FBI was raiding people who
were "distributing" anywhere from 1 to 100 GB of data (songs
and videos) presumably through P2P networks.  So people have
clearly figured out how to make this work.

Robert

On Thu, 26 Aug 2004, Dirk Terrell wrote:

> On Thu, 26 Aug 2004 10:10:09 -0700 (PDT), Robert J. Bradbury wrote:
>
> >
> >Something that might work to our advantage...
>
> I've done some work with distributed systems, having done the OS/2
> version of SETI@Home and written one from scratch using Java and
> Fortran to do solar system dynamics simulations. That approach works
> well with problems that are CPU bound (i.e. small data files but lots
> of crunching on them). It's less obvious how well it would work for
> something like TASS which would have to push large amounts of data
> around. Asymmetric connections like the DSL and cable that most people
> have would be ok (since downloads would be large and uploads much
> smaller), but Tom would need some serious bandwidth on his end to keep
> all of the clients fed.
>
> Dirk
>
>
>