http://setiathome.berkeley.edu/workunit.php?wuid=1899732449

I got invalid, which is pretty rare for me. The stderr doesn't even have a signal count section.

Also concerning me, the three (yes) valid results are all -9s, and one of them is different from the other two.

---

David
Sitting on my butt while others boldly go,
Waiting for a message from a small furry creature from Alpha Centauri.

Yes, that's a known problem, and we've discussed it several times - first on these message boards, and then Keith Myers and I made a concerted effort to track it down at Milkyway.

Two problems conspired against you:

1) the stderr can sometimes be incomplete
2) some SETI workunits require the line "SETI@Home Informational message -9 result_overflow" to be present in stderr before they can be accepted by the validator. Yours wasn't.

There's no perfect solution, but after the work Keith and I did, BOINC versions starting with v7.6.6 have additional protection against incomplete versions of stderr being reported to the servers.

Skip that one, and go straight to the new recommended BOINC v7.6.9 if you don't want to see this again. No guarantees, but the chances are greatly reduced.

Just a FYI, Richard. I have in fact seen just one "invalid" here at SETI since I installed the 7.6.6 BM and now the current 7.6.9. It was invalid for the same reason, missing the full stderr.txt output on an overflow task that was validated for the other 3 wingmen. So, 7.6.9 isn't a "perfect" solution for these occasional reporting failures. I have to say though that the newest BOINC Managers HAVE been a "perfect" solution for the problems I experienced at MilkyWay@Home. I haven't had an invalid there yet, since updating. Thanks again for working with me to develop the solutions, it was muchly appreciated.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

hmmm, interesting. I had expected the client patch to avoid all for all practical purposes (even though I noted during discussions the possibility of failure remained under obscure conditions).

Were there any particular known mitigating circumstances with the 7.6.6 & 9 occurrences ? or was it (in your opinion) just a matter of both sides of the interface needing more work ? [ I tend to the latter, but accept that resistance has been high for 'reasons']

---

"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.

I admit I haven't updated Boinc lately, but it's due to being busy (and/or lazy), not any negative perception of any newer version.

---

David
Sitting on my butt while others boldly go,
Waiting for a message from a small furry creature from Alpha Centauri.