On Sun, 20 Mar 2005 15:27:38 GMT, f/fgeorge <ffgeorge@yourplace.com>
wrote:

On Sun, 20 Mar 2005 07:59:36 +0100, Louis Holleman
<louis@holleman.demon.nl> wrote:

At this moment (06.55 UTC) I have still 4 machines crunching their
last bits. They'll run outta work sooner or later today unless they
fix that cable again :-(
It looks like I'll have to set Boinc/Seti to its max caching
capacity.... lately it's been up & down like a yoyo!

That'll fix the problem!!!! NOT LIKELY!!! You set your cache at the
max of 10 days, workunits expire in 14 days. You will only connect to
Berkeley every 10 days, Berkeley goes down and you are stuck with a
ton of unreportable and then expired workunits! Tell me again how
raising your cache is better!!

With all respect, I don't think that's the case. I've been watching
Seti over the past few days and when I set it to "connect every 2
days" it doesn't do just that. It goes fetching WU's for what it
thinks it needs for 2 days and starts crunching. After some time it
decides it will run out of WU's within 2 days and gets some more to
keep it going again for 2 days, etc. Meanwhile it also sends in
results etc. So the statement on the preferences page is a bit false,
it won't just connect every 2 days but it will indeed cache FOR to
days and keeps that cache going whenever it can.
Someone telling me different: go ahead, but it's not what I've seen
over the past period running Boinc/Seti :-)

Louis