[QUOTE=Prime95;592084]Dropbox, email, whatever works for you. The files will be in the same folder as prime95.exe.[/QUOTE]

Will do on Monday. In the mean time, I think I want to restart all the tests on that computer. After I send the files, I will delete Prime95 with all its files and re-download the program. Which version of the program should I download?

The official version is fine for running PRP tests.

[QUOTE=Prime95;592099]The official version is fine for running PRP tests.[/QUOTE]

Got it. I sent the files and then deleted and re-downloaded 30.3 b6 and restarted the exponent as well.

I've looked at the files and think either

1) A corrupt save file caused the iteration counter to go off by one (switched to non-Gerbicz error checking)
2) The PRP test was continued from an older version of prime95 that did not support Gerbicz error checking.

I think I have a fix in place to catch case number 1.

[QUOTE=Prime95;592563]I've looked at the files and think either

1) A corrupt save file caused the iteration counter to go off by one (switched to non-Gerbicz error checking)
2) The PRP test was continued from an older version of prime95 that did not support Gerbicz error checking.

I think I have a fix in place to catch case number 1.[/QUOTE]
If it's just a corrupted file, I think I may just continue with the restart. It's so early in the test and the computer was being iffy to begin with. Honestly, after 2 issues with the same exponent, if it helps to run just to check for a bug, I am totally willing to do that. Otherwise, I'll stick to the restart.

btw, just stumbled about this thread, remember it as a very old bug, already found in version 25.8 (from 2008) under Windows XP 32-bit ...
so has nothing to do with GEC, corrupted files, special entries from undoc.txt ...