Prime Gap News

[SethTro]

Thanks D.Stevens for turned in a big batch of multiples of 6199# and 9973#

This moves him/you into the top 10 contributors!

I personally enjoyed seeing "129008 32479#/2310 - 128966" (merit=4) replaced by "421733*9973#/208110-65834" (merit=13)

and "145542 10^19655-31389 (PDeGeest)" (merit=3.2) replaced by "1143911*9973#/208110-98744" (merit=14.7) as well as a few other low merit records (there seem to only be ~100 in the DB under <160,000)

I'm going to add you to the trusted list which means it won't take 48 hours to verify your next batch of submissions :)

[henryzz]

Originally Posted by SethTro

I'm going to add you to the trusted list which means it won't take 48 hours to verify your next batch of submissions :)

What does that change? It is still worth double-checking the gaps to some extent. At least the endpoints. Why do some of the commits have failed checks? I assume these have been sorted,

I added a few to the low merit list. Still haven't found that elusive 1M gap. 969864 was close.

[SethTro]

Originally Posted by henryzz

What does that change? It is still worth double-checking the gaps to some extent. At least the endpoints. Why do some of the commits have failed checks? I assume these have been sorted,

I added a few to the low merit list. Still haven't found that elusive 1M gap. 969864 was close.

For gaps with merit < 22, that are expected to take more than 3 minutes to verify (endpoints + internal) I (READ: the primegaps server) verify both endpoints but only a portion of the interval composites. the amount verified and if it reclassifies them as "C??" depends arbitrarily on if you have submitted a lot of records before (and I think you should be "trusted")

[ATH]

Originally Posted by SethTro

I updated two tables on the lists site

"Top 20 gaps by merit" is now "Top 20 largest gaps ordered by merit"
https://primegap-list-project.github...gaps-by-merit/

I am not a part of this project, so I am not complaining just suggesting.

In my opinion it is confusing to sort the very largest gaps by merit, so you have to search the list and find the very largest gap at rank 4, and 2nd largest at the bottom at rank 20.

I think the list used to be called "Largest gaps with merit>10" or something like that and was sorted by gapsize.

EDIT: Nevermind, I just noticed that list still exists: https://primegap-list-project.github...merit-over-10/
(But I still do not understand the point of sorting the very largest gaps by merit).

[SethTro]

Originally Posted by ATH

I am not a part of this project, so I am not complaining just suggesting.

In my opinion it is confusing to sort the very largest gaps by merit, so you have to search the list and find the very largest gap at rank 4, and 2nd largest at the bottom at rank 20.

I think the list used to be called "Largest gaps with merit>10" or something like that and was sorted by gapsize.

EDIT: Nevermind, I just noticed that list still exists: https://primegap-list-project.github...merit-over-10/
(But I still do not understand the point of sorting the very largest gaps by merit).

Thanks for the comment; I was focused on trying to understand what the list original had intended instead of asking "does this make sense?", So I changed it so there are four lists: top 20 with merit > 30, with merit > 20, with merit > 10, and overall top 20.

I changed the name so it's just top20 gaps by size now.

[SethTro]

The top 20 overall merits list has this new addition "974879377*2221#/4626930 -49418 to +28932". It's a gap of 78350 with merit 36.2822 the first large merit of 2021.

I've spent significant effort on 2221#/4626930 evaluated ~500 million values from m=1 to 1,400,000,000. This is my best by >2 merit! Overall I've found 2002 records with an average merit of 24.5!

The distribution of merit over ~432million entries looks like this

1 10405193
2 10961102
3 6918671
...
10 297378
11 226169
12 170160
13 124756
...
20 5984
21 3527
22 2029
23 1092
24 560
25 337
26 161
27 78
28 53
29 18
30 13
31 10
32 2
33 3
36 1