k=51 or about coordinated prime search

[Kosmaj]

L80 (Benson) just reported that
51*2^832630-1 is prime.

Now, we have been working on k=51 and other 13 Ks in our 5th drive to n=1M. Of course everybody is free to work on whatever K (and form) he/she wants and I mentioned several times that we welcome people working on k<300 without obligation to report primes under RPS. We are only asking that they tell us what K in what range they are working on, so that we can coordinate the search. I'm currently in touch with all people who work on k<300 on their own (without RPS) with exception of Xiao (k=289). And now comes Benson who most likely has been aware of our effort and of our stats page but who doesn't care about telling anybody about his effort. Does anybody know his e-mail address? He hasn't provided any on his bio page on Top-5000.

Regarding our search, I think we just keep on going to about 700k, where we can remove k=51 candidates. We could also extract k=51 numbers now and try to out-race Benson to 1M, but I'm not a fan of this kind of races...

BTW, k=51 has second lowest weight among the 14 Ks from the 5th drive, with about 34.5 candidates per 1000 n's remaining after sieving to 8.25T
in the n=600k-1M range.

[nuggetprime]

Try to ask Prof.Caldwell for his emailaddress.
If this way doesn't work, we should try to out-race him.
But I hope that you'll get the emailadress and Benson will tell you where he's working.

[Cruelty]

Any luck contacting Peter Benson? He has just reported another prime...

[kar_bon]

i mailed Mr. Caldwell directly a few weeks ago, to send Mr. Benson my mailadress to contact me. i'm interested also about his effort especially the start of his tests in the range 1005 < k < 1399 and other k's on our lists.
no reaction yet!

[Kosmaj]

I don't think there is a lot (if anything) we can do. Prof. Caldwell is not supposed to release e-mail addresses of people who registered on Top-5000. We can try to compose a message to him and ask Caldwell to deliver it. That's all but Benson can ignore it.

One advice: if you are working on a k<100, go and check n's between 800k and the exponent where the LLR FFT length is still smaller than 48k. It seems Benson likes to hunt there

Or you can try to spoil his search He now works on k=301, so, if you have resources start your search at 600k.

Or sit back, and just keep on working on your k's, there are enough primes for everybody to discover.

The only bad think is: there are now new gaps (just don't tell to Gary! ) and at n=800k it's no fun filling them...

[kar_bon]

Originally Posted by Kosmaj

"..." We can try to compose a message to him and ask Caldwell to deliver it. "..."

that's what i have done!

Originally Posted by Kosmaj

Or you can try to spoil his search He now works on k=301, so, if you have resources start your search at 600k.

that's a hard work. every week Benson deliver 1 to 4 primes in the range from 1005 < k < 1399 (a range of 200 k's at llr-level of n=400k actually) and now other ranges too (he also reserved k=303 in the range 480k-600k at Primesearch and some other k's!). so he must have 'some' hardware power we dream of, or many cores working for him.

Originally Posted by Kosmaj

Or sit back, and just keep on working on your k's, there are enough primes for everybody to discover.

yeah, infinitely many k's. look here for example, a list of Thomas:
http://www.mersenneforum.org/showpos...57&postcount=8 with low weight k's.

Originally Posted by Kosmaj

The only bad think is: there are now new gaps (just don't tell to Gary! ) and at n=800k it's no fun filling them...

psssst. ok. i'm not telling him

[gd_barnes]

Originally Posted by Kosmaj

The only bad think is: there are now new gaps (just don't tell to Gary! ) and at n=800k it's no fun filling them...

HA HA HA

I'm not filling those high and really wide gaps. n=260K to 520K for k=289 is enough for a while.

Ah, but wait. I have an excellent gap-filling file at a pretty high range of n that I'm about to post in the 5th drive forum that I just finished up with. The main difference is its a very narrow range so didn't take too long. Check it out if you haven't already. Doing that one is what delayed me on finishing k=289 for a little while.


Gary

[VBCurtis]

I didn't think much about this until today; I'll pause k=93 where I am (just over 680k if you wish to update k<300 page) and do the 800 to 8xx range presently.

What a jerk!

-curtis