----- Original Message ----- From: "Steve Harris" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Thursday, October 24, 2002 8:49 PM Subject: Re: Mersenne: On v18 factoring
> Daran - you ask why highest and not lowest? The discussion started > regarding old machines running v18 which are no longer in the care, > custody & control of an active GIMPS participant, AND which are > asking for factoring assignments which they cannot handle. Whatever > assignment is given to them, there is no telling how long until (or even > if) they will finish it... That's true of any client, not just runaway v18s. > ...We would not want to give them something that would hold up a > milestone a year or so down the road... If it got to the point where a milestone was being blocked, then someone else would poach it. I'd rather that happen to a forgotten client than to a slow but active participant. Also any server code-change is going to take the path of least resistance. The server is programmed to hand out the smallest exponent available. To start handing out larger exponents would involve more work than just changing the assignment type, and would probably introduce more bugs. [...] > You make a good point about P-1 completed assignments, but on further > reflection I don't think that is necessary. There aren't that many > available and certainly not at the higher end of the current range. They > will more than likely be P-1 tested when double-checked. That's not true. Most *new* DC assignments (currently > 8500000) are not P-1 complete as they were originally LLed by v18 or earlier clients. Many recycled DC assignments (mostly 7000000-8500000) were also never P-1ed by the client that let them expire. I specialise in P-1ing these 'neglected children'. However, my above remarks about 'path of least resistance' applies. There are probably more important server changes pending. I've been wondering if it would be possible to compile a list of P-1 incomplete exponents currently assigned to v18 or earlier clients. If so, then I would consider giving these a P-1. This, it could be argued, would be a form of poaching, in so far as if I were successfully to factorise, then the 'owner' would get a 'exponent already complete' error, which might cause some upset. OTOH, the project gains a factor that wouldn't otherwise have been found, and people still using v18 aren't likely to be particularly attentive. I'd like the views of list members concerning the ethics of this. > Steve Daran G. _________________________________________________________________________ Unsubscribe & list info -- http://www.ndatech.com/mersenne/signup.htm Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
