It takes two matching results to validate. Given the long run times of SoB, the long deadlines, and the difficulty some computers have running long LLR tasks, unfortunately it sometimes takes a really long time for another computer to come up with a matching result. Eventually, it will get validated.

I've got one that's almost as old as yours, having started at the beginning of June, and then going through a long string of wingmen who either had errors or just timed out. The first wingman was running SoB on a netbook with an Atom CPU!!! I don't think the wingman running my SoB WU is going to successfully finish it either, so maybe the next one will -- sometime in October, if I'm lucky.

SoB isn't a project with instant gratification! :)


____________
My lucky number is 75898⁵²⁴²⁸⁸+1

Think you're lucky because your running Bust v5.11 as most of the systems do. So if the second system is running the same program, you'll be granted credit. If 2 other systems are going to run Bust v6.05, you won't get any. See the "Bust 5.11 versus 6.05" post just below your post.

Let me just give you a few examples where 6.05 validated against 5.11 just fine:
http://www.primegrid.com/workunit.php?wuid=104266214
http://www.primegrid.com/workunit.php?wuid=109185700
http://www.primegrid.com/workunit.php?wuid=109051021 (in fact, this result also shows that 6.05 vs 6.05 can be deemed inconclusive, whereas 5.11 would validate against 6.05).

So I claim that there is no validation issue between 6.05 vs 5.11.
____________

OK,

So if I look at the last WU, which was validated between a v5.11 and a 6.05 app, the system running for 2.2 mil secs. did this calculation just for fun. I don't think the operator will like it because he had the same outcome as I had: succes, etc., etc. So there must be some kind of discrepancy between the 2 app's.

I need to point out that I've run a bunch of PSP and SoB WUs, and have run into the exact same problem.

I completed my WU, and then waited -- and waited and waited -- for someone else to complete it too. Someone finally did, and, as with you, the results did not validate. Same as you, the versions were different. One was Windows, one was Linux. (I don't remember which I was running since I run both.)

The next computer was running the same version as the other guy -- so I was the one with the not-matching version. Like you, I was a bit worried that there might be a discrepancy in the versions that might cause my perfectly good work to go to waste.

But then I looked at the computer that didn't validate against mine. You know what? It had several WUs that failed validation. So I was fairly confident that the third computer's result would validate against mine.

Sure enough, when it finally finished, that third computer validated against mine, even though it was running the same version as the second (non-valid) computer.

That's why I'm so sure this has nothing to do with the versions.
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

More data is always better than less data so I compiled a list of completed results from all my LLR WUs still in the database. 5.11 & 5.13 are Windows, 6.03 and 6.05 are Linux, and 1.05 is Mac. Only WU's with a valid canonical result are shown. The only SoB task I have left only has one completed result, but PSP is identical except for the values being tested.

Not shown are any results that did not complete processing, including results that timed out, were aborted, or had computation errors during processing.

Results are listed in the order they were returned.

PSP:
127175186 Valid: 5.11, 6.03
127167482 Valid: 5.11, 5.11
127163647 Valid: 5.11, INVALD: 6.03, Valid: 6.03, 6.03
126992980 Valid: 5.11, 5.11, 5.11
126958608 Valid: 5.11, 5.11
124825384 Valid: 5.11, 6.03
124660653 Valid: 5.11, 6.03, 5.11

The Riesel problem is identical to SoB/PSP except that the form of the candidate is k*2^n-1 rather than k*2^n+1.

TRP:
130682090 Valid: 5.11, 5.11
130595394 Valid: 5.11, 5.11
130199819 Valid: 5.11, 5.11
130610698 Valid: 5.11, 5.11
130471520 Valid: 5.11, 5.11
130071455 Valid: 6.05, 5.11
130071463 Valid: 5.11
130096104 Valid: 5.11, 6.05, 1.05

Normally, there's only 2 valid results per WU, but you can get more than two if a result times out, a new result is assigned to another computer, and then the timed-out result gets returned.

Usually, it takes at least two valid results for an LLR WU to be completed, but there are at least two sscenarios I'm aware of that can validate with only a single result. The first (which is what happened with 130071463), is when PrimeGrid is doublechecking a primality test done elsewhere. The second scenario is where Primegrid is selectively requiring only a single valid result when sent to normally reliable hosts.

This is the raw data; feel free to come to any conclusions you wish about what it means. :)

(Note: There's a preponderance of 5.11's because recently I've been running only Windows, so by default at least 50% of the results shown will be 5.11.)
____________
My lucky number is 75898⁵²⁴²⁸⁸+1