mersenne.ca - Page 72

James Heinrich · 2022-01-13, 19:39

There is a new report showing TF results that claim no-factor for a certain range when there is a factor in that range:
https://www.mersenne.ca/tfmissed.php

The NF run may have happened before or after the factor discovery. In about 1/3 of the cases the NF run happens within an hour of factor discovery, and might possibly be a poorly-recorded effort to show that the bitrange was completed even after finding the factor (this is not confirmed, maybe one of the users listed wants to weigh in on how they got so many of these kind of results).

SethTro · 2022-01-13, 20:22

Quote:

Originally Posted by James Heinrich

might possibly be a poorly-recorded effort to show that the bitrange was completed even after finding the factor (this is not confirmed, maybe one of the users listed wants to weigh in on how they got so many of these kind of results).

Exactly this; One reason mfaktc doesn't produce a NF result if it finds a composite factor (and the server rejects the composite factor result) so I manually produced many NF results for ranges I had completed but would otherwise be ignored

mathwiz · 2022-01-14, 17:11

This is a bit of a nitpick, but can this page: https://www.mersenne.ca/userfactors/any/1/bits (and related) have an "snfs" category? Factors like those found by Joppe Bos, Thorsten Kleinjung, Arjen Lenstra, Peter Montgomery were found by SNFS, not ECM.

https://eprint.iacr.org/2014/653.pdf

James Heinrich · 2022-01-15, 16:31

Quote:

Originally Posted by mathwiz

Factors like those found by Joppe Bos, Thorsten Kleinjung, Arjen Lenstra, Peter Montgomery were found by SNFS, not ECM.

If you want to give me an explicit list of the factors you mean (by exponent-bits is fine to save space) I can edit those few to show snfs.

nordi · 2022-01-15, 19:57

Quote:

Originally Posted by James Heinrich

If you want to give me an explicit list of the factors you mean (by exponent-bits is fine to save space) I can edit those few to show snfs.

M991: mersenneforum.org users, see https://www.mersenneforum.org/showth...290#post404290
M1061: Greg Childers, see M1061.pdf

Might also be good to give credit to finders instead of listing them as "anonymous"

nordi · 2022-01-15, 20:14

Quote:

Originally Posted by James Heinrich

If you want to give me an explicit list of the factors you mean (by exponent-bits is fine to save space) I can edit those few to show snfs.

And for the paper mathwiz was referencing, the exponents are
1009,1093,1109,1117,1123,1129,1151,1153,1171,1193
as listed on pages 16 to 17.

James Heinrich · 2022-01-15, 21:19

Quote:

Originally Posted by nordi

And for the paper mathwiz was referencing, the exponents are
1009,1093,1109,1117,1123,1129,1151,1153,1171,1193
as listed on pages 16 to 17.

Thanks, it wasn't obvious to me what the linked PDF was talking about. I have added those.

Quote:

Originally Posted by nordi

M991: mersenneforum.org users, see https://www.mersenneforum.org/showth...290#post404290
M1061: Greg Childers, see M1061.pdf

And those as well.

Quote:

Originally Posted by nordi

Might also be good to give credit to finders instead of listing them as "anonymous"

I can only credit what I know. If you see something that's not credited as it should be, please tell me.

Zhangrc · 2022-01-20, 07:05

Suggestion:
Show "MISSED FACTOR!" in P-1 results too, as in https://www.mersenne.ca/exponent/1014197 (shown in TF).
Some are not really "MISSED FACTORs", because P-1 can be run on the cofactor. But if it's a factor not previously found, it's a missed factor.
Recently I found 4 missed factors in 13.2M (missed by the 2001 anonymous P-1 results).

LaurV · 2022-01-20, 07:49

Kind of difficult to make, from the "organizational" point of view, one table shows who attempted the factorization, therefore missed the factor, the other one shows who found the factor. So, if you want to modify the second table, the text should be more like "missed by ..." and shows the name of who missed it. Honestly, this would be too much trouble. Let the things as they are, the addition of the nfs class of factors is already a very good improvement.

My two cents.

VBCurtis · 2022-01-20, 15:43

Quote:

Originally Posted by Zhangrc

But if it's a factor not previously found, it's a missed factor.
Recently I found 4 missed factors in 13.2M (missed by the 2001 anonymous P-1 results).

I suspect you misunderstand how P-1 bounds work. Just because you find one doesn't mean the previous P-1 run missed it the way that TF report says "missed factor". Can you give us an example of an exponent where you think the original P-1 work missed the factor, and the bounds for the previous P-1? It's more likely you found factors with bigger bounds than all 4 of your finds would have been found by the old bounds.

James Heinrich · 2022-01-20, 15:46

Quote:

Originally Posted by VBCurtis

Can you give us an example of an exponent where you think the original P-1 work missed the factor, and the bounds for the previous P-1?

https://www.mersenne.ca/p1missed.php

2022-01-20, 07:05	#789
Zhangrc "University student" May 2021 Beijing, China 2×5³ Posts	Suggestion: Show "MISSED FACTOR!" in P-1 results too, as in https://www.mersenne.ca/exponent/1014197 (shown in TF). Some are not really "MISSED FACTORs", because P-1 can be run on the cofactor. But if it's a factor not previously found, it's a missed factor. Recently I found 4 missed factors in 13.2M (missed by the 2001 anonymous P-1 results). Last fiddled with by Zhangrc on 2022-01-20 at 07:05

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Small inconsistencies between mersenne.org and mersenne.ca factor databases	GP2	mersenne.ca	44	2016-06-19 19:29
mersenne.ca (ex mersenne-aries.sili.net)	LaurV	mersenne.ca	8	2013-11-25 21:01
Gaussian-Mersenne & Eisenstein-Mersenne primes	siegert81	Math	2	2011-09-19 17:36
Mersenne Wiki: Improving the mersenne primes web site by FOSS methods	optim	PrimeNet	13	2004-07-09 13:51

2022-01-13, 19:39	#782
James Heinrich "James Heinrich" May 2004 ex-Northern Ontario 2·17·109 Posts	There is a new report showing TF results that claim no-factor for a certain range when there is a factor in that range: https://www.mersenne.ca/tfmissed.php The NF run may have happened before or after the factor discovery. In about 1/3 of the cases the NF run happens within an hour of factor discovery, and might possibly be a poorly-recorded effort to show that the bitrange was completed even after finding the factor (this is not confirmed, maybe one of the users listed wants to weigh in on how they got so many of these kind of results).

2022-01-14, 17:11	#784
mathwiz Mar 2019 404₈ Posts	This is a bit of a nitpick, but can this page: https://www.mersenne.ca/userfactors/any/1/bits (and related) have an "snfs" category? Factors like those found by Joppe Bos, Thorsten Kleinjung, Arjen Lenstra, Peter Montgomery were found by SNFS, not ECM. https://eprint.iacr.org/2014/653.pdf

2022-01-20, 07:49	#790
LaurV Romulan Interpreter "name field" Jun 2011 Thailand 7×1,423 Posts	Kind of difficult to make, from the "organizational" point of view, one table shows who attempted the factorization, therefore missed the factor, the other one shows who found the factor. So, if you want to modify the second table, the text should be more like "missed by ..." and shows the name of who missed it. Honestly, this would be too much trouble. Let the things as they are, the addition of the nfs class of factors is already a very good improvement. My two cents.