In other words, David is "pre-emptively poaching" these numbers, to prevent
them from getting taken by quitters, in thinking he can clear them before
others.... So, he's a poacher himself, it stands to reason on this side
of the aisle...
To me there's no difference between hoarding and poaching -- both are
detrimental to the effort.
You claim that David is doing a good job completing the exponents much
faster. I think they'd get done much faster if we just created a "cleanup
pool" of machines, perhaps a few dozen. I'd donate a P-100 or two to the
effort. v20 of Prime95 could allow a machine to join "the cleanup
crusade", which would have the following differences from standard
participation:
1) The smallest exponents could ONLY be assigned to the cleanup crusade.
If there were 100 cleanup boxes, then the smallest 400 exponents ought
to be held back for them.
2) The TTL (expiry time) on a cleanup box ought to be MUCH shorter than
60 days. Perhaps as low as 7 days. If you don't have a frequent
connection, don't be a cleanup hitter.
3) Once operational, Dave could release most of his queue, and flood the
crusade with work. We'd clear his 260 day backlog in about a month,
and would continue to crank out milestones even faster than random
poaching would allow.
If we did that, and didn't do ANYTHING else, other than ensure that when
small exponents under CURRENT operation expire, they go to the cleanup pool
rather than to the general pool. Questionable results and errors can also
go to the cleanup pool, as well.
Comments, flames, etc, welcome.... ;-)
At 08:09 PM 2/4/00 +0100, you wrote:
>Not so long ago David send a mail to this list on how he's doing this.
>He said:
>
> And now the hard and real important part:
>
> monitor your Account report EVERY DAY (I can't stress this enough, because
> EVERY DAY 4-5 new exponent are going to be assigned to you, if you are
> able to monitor your account on the weekend modify the part* schedule and
> Time setting, to fit your need) and keep ONLY 60 days of work for your
> computer, use the release form on PrimeNet to release high exponent that
> can't complete in 60 Days. and reorder your assignment so that the
> smallest exponent are finished first.
>
>So he takes only 60 days of work for each pc.
>
>
> >Most of Dave's assignments have gone untouched for 30 - 90 days.
>
>That those exponents have gone untouched is because he is scheduling the
>lowest
>exponents first and they are finished within 3 or 4 while a lot of other
>exponents
>take months. I saw some exponents that would expire in a few days and then
>the data
>gets updated and it takes another 60 days before the exponents expire while
>the LL
>test wasn't even started on those exponents.
>
>I think David is doing a good job in completing the milestones MUCH faster.
>He never poaches an exponent but lets the exponents asing by the server.
>Although i agree with you that he could release some of the higher exponents
>for so
>that others could test them.
>
> >If Dave gets poached, I won't shed a tear.
>
>Nobody should poach PERIOD
>
> >Any defense you'd like to offer for holding 9 months' work, I'll listen to,
>
> >but I doubt you'll come up with anything convincing.
>
>Like he said befor No more than 60 thays for each machine
>
>Sander
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
