mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   Data (https://www.mersenneforum.org/forumdisplay.php?f=21)
-   -   another false positive? (https://www.mersenneforum.org/showthread.php?t=21102)

ixfd64 2016-03-14 18:50

another false positive?
 
M78253603 has a 0x00000000000000__ residue but is marked as composite. I'm guessing it's really a 0x0000000000000002 residue. Someone look into this?

Madpoo 2016-03-14 18:57

[QUOTE=ixfd64;429104]M78253603 has a 0x00000000000000__ residue but is marked as composite. I'm guessing it's really a 0x0000000000000002 residue. Someone look into this?[/QUOTE]

You guessed right, the masked part is NOT "00". But hey, I can't confirm it was "02" because then you might check in a bogus result with that same residue and "verify" it. :smile:

What's odd is that it wasn't marked suspect... I vote that when a residue comes in like that, it should be marked suspect no matter what errors it reported during the run...clearly the end result is in error no matter what the program thought.

EDIT: The same user has more of those all in the past few weeks... here's their full list of badness:
78253547
78253559
78253603
78253727

It makes me suspicious of their other result with a normal looking residue: M78253627

And if you're in the mood for these types of things, here's a bonus one from a different user last November:
79060367

Prime95 2016-03-14 19:07

Yes, a CudaLucas 0x0000000000000002 residue. Nothing to see here. Madpoo or I need to shove it back in the first-time check queue.

Prime95 2016-03-14 22:11

Result (and 5 others) removed from database. Exponent available for testing once again.

Server (finally) changed to reject any LL result with a residue of 0x0000000000000002


All times are UTC. The time now is 21:06.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2022, Jelsoft Enterprises Ltd.