mersenneforum.org 2*b^n-1, b<=2048
 Register FAQ Search Today's Posts Mark Forums Read

2019-08-30, 01:25   #12
sweety439

Nov 2016

13·173 Posts

Quote:
 Originally Posted by LaurV 2*1595^175532-1 is prime! (562188 decimal digits) 18 to go Reported to UTM, waiting verification.
Which 18 bases remain?

And did you complete to n=200K?

Also, can update a text file like this?
Attached Files
 Riesel k=2.txt (7.2 KB, 35 views)

Last fiddled with by sweety439 on 2019-08-30 at 01:25

2019-08-30, 05:25   #13
Batalov

"Serge"
Mar 2008
Phi(4,2^7658614+1)/2

2·33·132 Posts

Quote:
 Originally Posted by paulunderwood https://primes.utm.edu/primes/download.php will get you to the complete top5000 list.
That's a snapshot of the top5000 list (plus special categories), but a (rather open) secret is that the numbers that ever were in the Top5K never leave.

Use the search page https://primes.utm.edu/primes/search.php
Enter ^2*[[:digit:]]{1,}^[[:digit:]]{1,}-1 in Mathematical Description
check the * all (includes unverified primes)
and set 2000 Maximum number of primes to output
and press the Search button

That's what I would do.

 2019-08-30, 21:42 #14 Batalov     "Serge" Mar 2008 Phi(4,2^7658614+1)/2 23A616 Posts In the thread title, is it n<=2048 or is it b<=2048? 1067 1079 1152 1193 1217 1228 1283 1316 1593 1646 1670 1760 1762 1829 1901 1909 1957 1997 2012 ...some are probably eliminated
2019-08-31, 01:09   #15
Dr Sardonicus

Feb 2017
Nowhere

23×3×5×29 Posts

Quote:
 Originally Posted by Batalov In the thread title, is it n<=2048 or is it b<=2048?
The smart-aleck answer would be "Yes."

Quote:
 Originally Posted by LaurV 2*2018^145464-1 is prime!
is that it should be b <= 2048.

 2019-08-31, 04:07 #16 LaurV Romulan Interpreter     Jun 2011 Thailand 2×3×31×47 Posts Grrr, typo, thanks for signaling it. Trying to fix it, our new supermod rights would allow changing thread titles, however we never did that before and we know it is tricky, as "modifitting" first post can lead to deletion of the whole thread... (we read that somewhere ) Edit: fixed. We are at job today, our working Sat (here we work on alternate Sat, one yes, one not, I mean, we go to the job one Sat and do nice things, and we stay home and argue with MF people - pun intended - on the other Sat, specification to avoid the interpretation we go to the office all Sat's but one we do something and one we do nothing ), but when we arrive home tonight we may post the remaining bases with their reached n. We have a nice Excel file which we were convinced we posted in the past, but as many people insist here, we think our memory is flawed.. Last fiddled with by LaurV on 2019-08-31 at 04:14
 2019-08-31, 21:14 #17 Batalov     "Serge" Mar 2008 Phi(4,2^7658614+1)/2 2×33×132 Posts Quick free hint. If you are attacking 1152, then remove all odd n values. (Why? N^2 - 1 is composite for N>>2 )
 2019-09-01, 04:05 #18 LaurV Romulan Interpreter     Jun 2011 Thailand 2·3·31·47 Posts 1152 is tested to n=200k, and indeed, without me ever thinking to it, all the odd values were removed in the past (probably by srsieve or srsieve1, which we used), because there are no odd values in the cllr log file lresults.zip Last fiddled with by LaurV on 2019-09-01 at 04:07 Reason: attach
 2019-09-01, 05:03 #19 LaurV Romulan Interpreter     Jun 2011 Thailand 210468 Posts Also to note that for this particular "project", the debate "$$n\ge 1$$ vs. $$n>1$$" (i.e. $$n\ne 1$$ ) is relatively easy to set. There are 563 candidates that are eliminated (in the original case) because 2*b-1 is prime, and when you test them with 2<=n<=100, only 31 remain. With n to 1000, only 10 remain, then with n to 10k, only 3 remain. The 7 which are largest below 10k, sorted by the base, are (note that these are all prime for n=1 too, and not for other n smaller than shown one): 2*591^1191-1 2*954^1477-1 2*976^1925-1 2*1332^1068-1 2*1417^3861-1 2*1717^6360-1 2*1867^6172-1 Daring to go higher will spot 2*1297^19840-1, and a bit over 50k you find 2*1102^52177-1. The last rebel came out at 2*1360^74689-1, and the dispute is solved, there are no additional bases open below b=2048, when n is restricted to be greater than 1. You can easily see that if you play with yellow cell L2 in the attached excel file. From which, of course, other "lists" can be done, or "text files" exported very easy, to make Sweety happy... (caution, data is grouped to hide the non-important rows and columns, you have to expand them manually, clicking the "+" signs on the left, or clicking the group leaders (the little 1, 2 numbers which appear in the upper left). Do not pay attention to the sandbox page, that is related to how the windows are arranged on my screen, not of your interest. Some stuff in those sheets are manually maintained, some automatically. Forgive us if the manual stuff is not to date... prime-riesel-k2-b2.zip Last fiddled with by LaurV on 2019-09-01 at 05:32 Reason: grrr, no way to convince that code box to properly align them. Deleted.
2019-09-01, 05:32   #20
Batalov

"Serge"
Mar 2008
Phi(4,2^7658614+1)/2

2·33·132 Posts

Quote:
 Originally Posted by LaurV 1152 is tested to n=200k, and indeed, without me ever thinking to it, all the odd values were removed in the past (probably by srsieve or srsieve1, which we used), because there are no odd values in the cllr log file
No they weren't; e.g. (just a few random lines out of a few hundred of these)
Quote:
 Originally Posted by LaurV 2*1152^194841-1 is not prime. ...Time : 3841.382 sec. 2*1152^194301-1 is not prime. ...Time : 3550.637 sec. 2*1152^199745-1 is not prime. ...Time : 1971.451 sec.
...but every 2*11522*n+1 - 1 = 482*11522*n - 1 = N2 - 1

 2019-09-01, 05:37 #21 LaurV Romulan Interpreter     Jun 2011 Thailand 2·3·31·47 Posts Nice, well spotted, and good to know for my next step. Thanks. As I said, I was never thinking about that. About 50-80 hours lost.