Archived reservation posts - Page 28

schickel · 2010-10-23, 06:32

Quote:

Originally Posted by Andi47

sequence 10212 is currently at size 147, not 150. Index 2366 is correct. (currently GNFSing the c145 cofactor)

Hmmm...well like I said, I saw some squirrelishness in there. I'll have to kick off a job tonight and see what things look like now.

[Edit] I was reluctant to hit it again too quickly.

Greebley · 2010-10-24, 13:50

I corrected several loss of data issues when the new database first went up. However, a bunch did require aliqueit to be run.

Basically, a term factored by aliqueit was treated as prime by the old database. Therefore the factorization wasn't even in the old db and if aliqueit was run again even on one term, the later sequence would be found.

I did submit earlier versions of the old db to the new to make sure work before the bug made it back in, and looked at a bunch of seq that looked like they might be wrong.

Chances are decent though that some were missed. If you know sequences should be longer, then it is possible even a short run or set of runs of aliqueit might refind them since most factorizations still exist.

It might be worth listing them and we can see if we can regenerate them.

schickel · 2010-10-25, 07:56

Quote:

Originally Posted by Greebley

I corrected several loss of data issues when the new database first went up. However, a bunch did require aliqueit to be run.

Basically, a term factored by aliqueit was treated as prime by the old database. Therefore the factorization wasn't even in the old db and if aliqueit was run again even on one term, the later sequence would be found.

I did submit earlier versions of the old db to the new to make sure work before the bug made it back in, and looked at a bunch of seq that looked like they might be wrong.

Chances are decent though that some were missed. If you know sequences should be longer, then it is possible even a short run or set of runs of aliqueit might refind them since most factorizations still exist.

It might be worth listing them and we can see if we can regenerate them.

Actually, I was just speaking of the current reservation list. There were a couple that didn't line up with the last time I pulled a list. I figure that those were probably a result of the power outage that Syd reported over in the factorDB thread....

fivemack · 2010-10-28, 10:37

86160, after bravely throwing off the 2*3 driver at line 1848, picked up the 2^2*7 driver at line 1866; have taken it to line 1911, fully-ECMed C126 ready for anyone else to pick up, but I'm done with this one.

schickel · 2010-10-30, 04:26

Pulled and posted a complete update.....a couple of nice moves in there:

Code:

 75402    bchaffin         821   105   2^2 * 3^4 * 7
 75402    bchaffin        3581   136   2^4 * 31

and

Code:

392430    unconnected     5637   123   2
392430    unconnected     7206   116   2^2 * 7^2

Andi47 · 2010-10-30, 06:21

Quote:

Originally Posted by schickel

Pulled and posted a complete update

Quote:

Originally Posted by schickel

Code:

Current reservations
 10212    Andi47          2369   150   2 * 3

What makes you always (i.e. at every update) think that sequence 10212 is at 150 digits?

Quote:

Originally Posted by factor database just a minute ago

Code:

Unchecked 	2369 	147 (show) 	3102789883...<147> = 2 · 3 · 58693 · 8810789145...<141>

schickel · 2010-10-30, 06:36

Quote:

Originally Posted by Andi47

What makes you always (i.e. at every update) think that sequence 10212 is at 150 digits?

Well now, that is a very good question. I'll have to see what my program is retieving. The raw output is this line:

Code:

 10212 2369. sz 150 2 * 3 * 58693

What it does is pull the elf file and then check the size of the composite on the last line. Apparently something is being reported peculiarly just for 10212....

[Edit] Check that....it looks like something throwing off the size calulation on everything. I'll have to investigate further. Thanks for catching that.

schickel · 2010-10-30, 07:50

Quote:

Originally Posted by Andi47

What makes you always (i.e. at every update) think that sequence 10212 is at 150 digits?

'Cause I made a rookie mistake the last time I twiddled with the program. After Syd changed the interface, I moved some statements around. Net result: I was taking the position of the '.' in the last line, padding the line, then taking everything from the unpadded position of the '.' to the '=' for the length of the compostite. So I had an 'off by x, x from 2-4' error, depending on the index of the sequence.....

Stupid, stupid, stupid. Sorry, I'll recompile and pull a correct update, but probably not until after work tomorrow.

Frank

gd_barnes · 2010-11-01, 19:44

392430 that shows a downdriver probably needs to be updated in the 1st posting here.

It has advanced nearly 1600 indexes since the last update!

schickel · 2010-11-04, 12:01

Unreserving 67920 for Clifford.

FactorEyes · 2010-11-05, 17:11

Reserving 215778; continuing with 235416.

2010-10-30, 04:26	#302
schickel "Frank <^>" Dec 2004 CDP Janesville 2×1,061 Posts	Pulled and posted a complete update.....a couple of nice moves in there: Code: 75402 bchaffin 821 105 2^2 * 3^4 * 7 75402 bchaffin 3581 136 2^4 * 31 and Code: 392430 unconnected 5637 123 2 392430 unconnected 7206 116 2^2 * 7^2

2010-11-05, 17:11	#308
FactorEyes Oct 2006 vomit_frame_pointer 2³×3²×5 Posts	Reserving 215778 Reserving 215778; continuing with 235416.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Archived reservation posts	lalera	Operazione Doppi Mersennes	52	2013-08-30 14:52
Archived sieve reservations and discussions	MooooMoo	Twin Prime Search	735	2011-12-09 16:33
Archived project section+ some other suggestions	Citrix	Forum Feedback	1	2006-05-03 09:29
Archived data files...	Xyzzy	Data	22	2004-05-25 03:04
older files archived?	tom11784	Lone Mersenne Hunters	0	2004-05-24 12:02

2010-10-24, 13:50	#299
Greebley May 2009 Dedham Massachusetts USA 3·281 Posts	I corrected several loss of data issues when the new database first went up. However, a bunch did require aliqueit to be run. Basically, a term factored by aliqueit was treated as prime by the old database. Therefore the factorization wasn't even in the old db and if aliqueit was run again even on one term, the later sequence would be found. I did submit earlier versions of the old db to the new to make sure work before the bug made it back in, and looked at a bunch of seq that looked like they might be wrong. Chances are decent though that some were missed. If you know sequences should be longer, then it is possible even a short run or set of runs of aliqueit might refind them since most factorizations still exist. It might be worth listing them and we can see if we can regenerate them.

2010-10-28, 10:37	#301
fivemack (loop (#_fork)) Feb 2006 Cambridge, England 1100100011000₂ Posts	86160, after bravely throwing off the 23 driver at line 1848, picked up the 2^27 driver at line 1866; have taken it to line 1911, fully-ECMed C126 ready for anyone else to pick up, but I'm done with this one.

2010-11-01, 19:44	#306
gd_barnes May 2007 Kansas; USA 2·41·127 Posts	392430 that shows a downdriver probably needs to be updated in the 1st posting here. It has advanced nearly 1600 indexes since the last update!

2010-11-04, 12:01	#307
schickel "Frank <^>" Dec 2004 CDP Janesville 84A₁₆ Posts	Unreserving 67920 for Clifford.