Mersenne: RE: Mersenne Digest V1 #463 464
A collective block of Q A: Is it possible for two GIMPS accounts to be merged into one new account, such that the work completed by the two accounts is credited to the new account ? Or is it possible to take the work done by one account and transfer it to another ? Yes. Change your userid in Prime95, connect to the server. PrimeNet will take care of transferring credit and assignments automatically. Alternatively, email me [EMAIL PROTECTED]. Anyone aware of any work in Mac-GIMPs land? As Will indirectly pointed out, PrimeNet only supports MacLucas for manual testing at the moment. I continue to offer support for anyone who needs help with the server's protocols and other requirements. Takes some work, but it's worth it. [sorry, replying to digests is probably unfriendly, but I can't face hordes of email. Do people mind if I just reply to interesting messages from the digest two-daily or so?] I reply to digests. Like sending email from light-hours away... Still getting a 2250 error with one of my networked PCs. All the rest work fine. Any ideas (yes I have a primenet.ini) Error 2250 is the RPC protocol can't connect error. Select HTTP in the PrimeNet options to use the primenet.ini file. A group of students here at Virginia Tech have formed a group here to promote distributed computing. The Computer Science department is interested in participating, but are hesitant to join GIMPs (or distributed.net). One of their questions is "What other companies or schools are participating?" I have read the link for the University of Albany. If there is anyone else out there representing a school or company, or who has permission to use someone else's computer systems, I'd greatly appreciate hearing from you. A few days ago I did a count of the colleges and universities contributing CPU time to GIMPS on PrimeNet, by email domain. There are 267, including almost every big name you can think of, plus about a hundred high and elementary schools. Commercial addresses are harder to count because they include ISPs, but they are in the thousands. There are a few dozen government accounts, including the US Navy and Air Force. What (if any) are the concerns with having an account's password and user ID posted on a web page? Would someone be able to change the "Your Name" and "Your email address" fields with them? Someone else could monkey with your account. So if my school set up a web page to encourage people to join our team, could someone come along and usurp our work? Not if you at least once set LockUserInfo=1 in the prime.ini file. To change your account info or transfer anything around after this, you need to contact me [EMAIL PROTECTED]. The challenge account prototyped distributed teams for GIMPS, and originally had a few hundred contributors (they have since started their own accounts, like netconx). Also, what would happen if someone changed (or deleted) the existing UserID and password in the middle of a LL test? The server would return 'assigned to someone else' errors until the test completed, at which time the server would accept and credit the result to the new user, then email the assigned account owner that the test was done. The secure-yet-still-automated way to do this is to have a public html form that takes the new persons data and wraps it up and sends you an email containing another url that you can click on from inside your e-mail system which will process the user's membership. One working out of the system, then many one-click approvals, rather than simply having the applicants e-mail you and needing to do boring data entry, and you still have control unlike if you just put the password out. Have teams been implemented? Theyt could follow the practice of "web rings" and have an "inner circle" who have approval power; applications to join can get sent to all approvers, or round-robined to each in turn, or a designates one or two. There is no limit to possible subtletl when it comes to dreaming up a user interface. There are a few hundred teams. The account locking mechanism works quite well, so most team leaders simply pre-configure the software with the selected account settings, add the locking flag, and let people copy the files and start the program. If they somehow change things, they at most take the one or two exponents assigned to that computer. Would it be possible to have the next version sending all expected completion dates in one time, instead of one by one? Or are there objections to do so? (copied in part from a private reply to Henk) A speed-up tip: If you are using RPC, you might try HTTP. Another speed boost for v17 RPC/HTTP and v15/v16 HTTP is to cache the entropia.com DNS lookup locally in your HOSTS file, using IP 207.104.25.155. Please see the FAQ page at http://entropia.com/ips/faq.html for more. Best regards, scott
Re: Mersenne: Net abuse
...as I was saying... they have been obnoxious net neighbors for years although this capricious and inconsistent blocking of arbitrary email addresses is a new one on me. I know numerous ISP and private email servers who automatically block any SMTP from a known 'open relay' server. There are several 'services' online which monitor said open relays, and provide automatic blocking lists for these sites. http://maps.vix.com/rbl is one such service. -jrp
Re: Mersenne: Password safety
you an email containing another url that you can click on from inside your e-mail system which will process the user's membership. Click? On e-mail? Note - not everybody is reading e-mail from their home PC using Netscrape or Internet Exploder. I use ELM - many others use PINE - neither of them is graphical, nor tied to the Mickeysoft OLE way of things... If you do e-mail confirmation, the best (and easiest) way is to send the user a message with a unique code, and then have the user reply to it. Possibly with added information to make the transaction even more secure.
Re: Mersenne: Net abuse
At 08:54 PM 11/12/98 -0800, you wrote: I know numerous ISP and private email servers who automatically block any SMTP from a known 'open relay' server. Usa.net and globalserve.net are not open relays. -- .*. "Clouds are not spheres, mountains are not cones, coastlines are not -()circles, and bark is not smooth, nor does lightning travel in a `*' straight line."- -- B. Mandelbrot |http://surf.to/pgd.net _ | Paul Derbyshire [EMAIL PROTECTED] Programmer Humanist|ICQ: 10423848|
Mersenne: interesting theorem
I ran across an interesting statement on the top of a math paper that I was helping my sister with. It said that every even number greater than 4 is the sum of two primes. I am curious if this has been proven and if anyone knows where I could find more info about this. Thanks. -- Visit my new and improved home page at http://www.fireantproductions.com/cannona ICQ #: 22773363
Re: Mersenne: interesting theorem
Ok, let's see. 3x2=6 But what about 8? It's factors are 1 and 8 and 2 and 4. That doesn't work too well. Jon I ran across an interesting statement on the top of a math paper that I was helping my sister with. It said that every even number greater than 4 is the sum of two primes. I am curious if this has been proven and if anyone knows where I could find more info about this. Thanks.
Re: Mersenne: interesting theorem
I'm sure that it probably has been done, however, I am considering writing a program to see if I can disprove this theorem. It is my belief that it can be disproven because the higher you go, the less primes you have. Anyway, writing the program will be challenging and give me something to do. -- Visit my new and improved home page at http://www.fireantproductions.com/cannona ICQ #: 22773363 On Fri, 13 Nov 1998 [EMAIL PROTECTED] wrote: In a message dated 11/13/98 6:55:53 PM Eastern Standard Time, [EMAIL PROTECTED] writes: I ran across an interesting statement on the top of a math paper that I was helping my sister with. It said that every even number greater than 4 is the sum of two primes. I am curious if this has been proven and if anyone knows where I could find more info about this. Thanks. I, too, had heard this when in the seventh grade (we did a lot of work with theorems and conjectures). As I recall, this particular one was not or could not be proven. It sure is interesting to consider and who knows in what ways this simple theorem could be useful for us. -Joel
Mersenne: Newbie Questions
If this is documented somewhere forgive me. I have looked. In the individual account status report, it is not obvious to me how the P90 CPU hrs/day is calculated. I have calculated that I should have about 400 but my report shows 111. All machines are running at full tilt with no screen savers. Some of my machines are still on their first assignment. Does a PC with a long task check in every so often or does it wait until the end of the task? Is this why my count is low? Why when my setup has 5 days work indicated does my machine get a month long task. Just wondering. I'll probably have more questions Steve Gardner [EMAIL PROTECTED] Check out 84000 computer products at www.pcavenue.com
Re: Mersenne: interesting theorem
Try http://www.utm.edu/research/primes/notes/conjectures/ for a list of interesting conjectures including this one. Steve Gardner [EMAIL PROTECTED] Check out 84000 computer products at www.pcavenue.com -Original Message- From: Aaron Cannon [EMAIL PROTECTED] To: [EMAIL PROTECTED] [EMAIL PROTECTED] Date: Friday, November 13, 1998 9:11 PM Subject: Mersenne: interesting theorem I ran across an interesting statement on the top of a math paper that I was helping my sister with. It said that every even number greater than 4 is the sum of two primes. I am curious if this has been proven and if anyone knows where I could find more info about this. Thanks. -- Visit my new and improved home page at http://www.fireantproductions.com/cannona ICQ #: 22773363
Re: Mersenne: interesting theorem
At 8:51 PM -0500 11/13/98, [EMAIL PROTECTED] wrote: No, not factors. The different addends of the number. 4 = 2 + 2 (2 is prime) 6 = 3 + 3 (3 is prime) 8 = 3 + 5 (both prime) 10 = 5+5 12 = 5+7 14=7+7 16=5+11 18=7+11 Well, you get the idea. Another interesting thing to note which is part of the theorem is that it can only be done in one way, and exactly one way. So what ever the two primes are that add to 250...few secs later 163 and 87, no other pair of primes will add to 250 according to the theorem. Joel Not quite. 10= 5 + 5 = 3 + 7, 14 = 7 + 7 = 3 + 11, 34 = 17 + 17 = 31 + 3 = 29 + 5 = 23 + 11. There are many others... Anyway, it's not a theorem, it's a conjecture (Goldbach's conjecture). It's never been proven, nor has anyone found a counterexample. mark snyder mark snyder fitchburg state college
Re: Mersenne: GIMPS Exposure
In a message dated 11/13/98 9:55:37 PM Eastern Standard Time, [EMAIL PROTECTED] writes: Maybe there is a web service out there that does archive newspapers. Anyone know of one? Yes, www.elibrary.com archives certain papers accessible for a fee. -Joel