View Single Post
Old 2021-03-31, 11:42   #5
WraithX's Avatar
Mar 2006

2·35 Posts

Originally Posted by CRGreathouse View Post
I believe they each verified this independently back in 2010 (maybe 2009). When I say "independently", I mean that they wrote software implementations independently of each other -- they used the same underlying mathematics, so that's still a potential source of error for those concerned about correctness.

And many people, myself included, checked that there were no BPSW pseudoprimes on the list.
And to make it a little more clear, Jan wrote his code and built his database a long time ago.

Then I coded up the underlying algorithms and Jeff and I ran that code in a coordinated effort to create our own database to compare to Jan's database. Jeff also helped update this code so it could run on his computers back in the day. But, I know I've lost some of my data since then, so we couldn't verify how much of this work we had done.

As Jeff mentioned, we are redoing the verification. I've re-written the code, with a big thanks to Ben Buhrow for providing low level mul-mod and factorization code!, and this is what Jeff and I are currently using to redo the verification. And, to second what Jeff said, the work is ongoing and there's no ETA for when we'll be finished.
WraithX is offline   Reply With Quote