"Verified" bad results?

[ixfd64]

Have a look at M1373881: http://mersenne.org/M1373881

The first LL result appears to be wrong but is still marked as verified. What gives?

[Prime95]

Originally Posted by ixfd64

Have a look at M1373881: http://mersenne.org/M1373881

The first LL result appears to be wrong but is still marked as verified. What gives?

Myrman ran his own program which had some bugs in printing the residue. His residues were granted a little extra leniency.

[bloodIce]

I think that after noticing a discrepancy one should always attempt factoring. In this case it was easy to find a factor: 39615513750093690993572647. Does not solve the problem with the misreported residue, but it is clear that the number is composite if a factor is found.

[Madpoo]

Originally Posted by Prime95

Myrman ran his own program which had some bugs in printing the residue. His residues were granted a little extra leniency.

Those Myrman residues caused me to do some workarounds when I was analyzing the smaller exponents. LOL

In this case, and some others, I ran my own test just to confirm, since they were kind of weird. I wanted to make sure they really matched even though George was aware the residue was calculated a little differently but still matched, in theory. There shouldn't be any problems with the Myrman results at this point, I never found anything off kilter.

[Mark Rose]

I suppose it's kind of evil, but you could overwrite the bad-but-correct residues with the proper residues.

[0PolarBearsHere]

I'd be interested to know what the difference in residue calculation is.
If it's going to become more widespread and we are likely to see more of them, it'd be good to have some indication of which residues are which type, so we know if a mismatch is actually a mismatch, or a match in a halloween outfit.

[S485122]

Originally Posted by 0PolarBearsHere

If it's going to become more widespread and we are likely to see more of them...

This is old stuff, look at the exponent :

Originally Posted by ixfd64

Have a look at M1373881: http://mersenne.org/M1373881

Also note the tense George used : ran not runs and had not has.

Originally Posted by Prime95

Myrman ran his own program which had some bugs in printing the residue.

Jacob

[Madpoo]

Originally Posted by S485122

This is old stuff, look at the exponent...

Yes, those results from that user are all quite old and no longer a concern.

[0PolarBearsHere]

Originally Posted by Madpoo

Yes, those results from that user are all quite old and no longer a concern.

Is it likely/possible that similar issues will be appear in the future, or will it be a case of "if you write your own software make sure your residues are going to match".