As long as you are going to relaod anyway, switch to the Command Line
Interface(CLI)! Download the setiathome-3.03.i386-winnt-cmdline.exe
file and of course SetiDriver and SetiSpy too.
Put those 3 files in your empty Seti directory extract the ones needed
and then run SetiDriver. Set the desired cache size to a number
greater than zero and the machine will download a unit from Berkeley
and you will processing about 10% or so faster than the screen saver
version ever thought of!
the cmdline file may have a different number instead of 3.03, that is
the version number and I believe that Seti has a newer one than that.

On Sun, 11 Jan 2004 13:48:07 +0000, Terry Pratchett
<tpratchett@unseen.demon.co.uk> wrote:

In message <mdg10095ao5ovdec1m2loivqr62do92igj@4ax.com>, f/fgeorge 
<f/fgeorge@yourplace.com> writes

Okay the next thing to check are the angles, more specifically the RA
angle. Check in your SetiQueue and see if you aren't receiving lots of
very low or very high angles. An average is around .423 or
thereabouts. Much difference than that neighborhood and the computer
can take a very long time.
You could also be working your way thru workunits that have lots of
Spikes, Triplets, Gaussians, or Pulses. I believe that Spikes and
Pulses will extend the time needed for crunching.

I've been looking at this aspect, but don't see a solution there.  It'd 
mean that the machine in question is *consistently* being thrown big, 
difficult units by SeiQueue while the rest of the network trundles along 
happily.  I don't know how or why this would be done.

I believe that you said that you deleted Seti from your computer and
redownloaded it and reinstalled it? If not I would do that if neither
of the above seem to pan out.

I'm stopping the programme after this unit and will do that.

OOOOH....have you defragmented the drive lately? If yes than that
could have moved Seti to new places, if not that could be part of the
problem especially if the hard drive is almost full.

94% of the 4Ghz drive is empty.

I'm been picking up some ideas from all this
and tomorrow we'll get cracking on it.  Whatever is causing this, 
though, is not interfering with the rest of the machine in any way.

Thanks to all for your help -- I'll report back.