Posted 21 May 2015 by
DoctorNow
Then another program has grabbed all the power.
Maybe Firefox. It happens from time to time that one of my addons suddenly grows in cpu-usage and then occupies a whole core. This immediately affects the BOINC processes and slows them down. I experienced the more addons the more this can happen, for whatever reason.
Posted 2 Jan 2015 by
DoctorNow
How about this for a badge design? Very cool, very, very cool....
I rather would see a complete other badge to prevent any confusions to which projects they refer, but I guess that's a bit hard to make.
Maybe it would be enough to remove the 10 and set an "SR" in the lower left corner, but I'm not so sure at the moment...
Posted 2 Jan 2015 by
DoctorNow
Can we make one eventually
No, "we" can't. Doesn't look like you readed this thread here, the reasons why it isn't possible are in there.
Posted 1 Jan 2015 by
DoctorNow
So ... we still don't have an answer ... what does ARM stand for? (Tin hat on)
It's called Advanced RISC Machines. Take a look up on Wikipedia or Google if you want to know even more.
Sidenote: the abbration readed as a whole in German means poor or as synonym weak. Another hint for the "uselessness" of the thing nowadays. ;-)
Happy New Year to everyone btw.! ;-)
Posted 28 Dec 2014 by
DoctorNow
2126*396^65889+1 found by DoctorNow
Now that's nice. :-)
Is that one for the Top 5000?
Posted 23 Dec 2014 by
DoctorNow
Ok, too bad...
Posted 23 Dec 2014 by
DoctorNow
Tasks are finished now but as I wanted to take a look what primes they were I only see a short version of the output and no prime at all - the Windows outputs are quite longer with more information.
Example: 686362
Is that intended? At least they validated...
Posted 23 Dec 2014 by
DoctorNow
Yeah, I know. Since I don't have AVX machines it takes longer anyway.
I did not look at the estimates, but because the tasks were so fast at 100% the remaining time disappeared anayway.
Posted 23 Dec 2014 by
DoctorNow
Hm, progress behaviour of the Linux tasks is a bit odd in comparison to the Windoze ones:
They went quite fast to 99%, after some minutes they were already at 100%, but they are still crunching over an hour now at 100% - range 150-175 as far as I see from the wu-name. Looks like they take longer than the same wus on Windows.
Posted 23 Dec 2014 by
DoctorNow
Linux 64 also seems to run good, example http://srbase.myfirewall.org/sr5/result.php?resultid=680381
First thought it won't because of the short runtime but they validated.
Posted 21 Dec 2014 by
DoctorNow
Rebirther wrote:
The R3 is a really big base with millions of primes.
This much? Didn't know that. Well, wonder if I have one then, did not check any result yet, but since that's so many WUs it would take a while to check oneself unfortunately. That's why I asked...
Werinbert wrote:
You should at least show the number of primes found by each user on their accounts page.
Well, if there are millions of primes this would become a big list in the account page - while we have currently 102 users every user can get thousands of primes, theoretically spoken. Would be a bit much...
Edit 1:
This needs some programming in boinc forum code and assimilator using db but Iam only using output files and I have no skills in that, sorry.
This makes me a bit speechless now, I wonder...
Edit 2:
Great, just noticed that all WUs are already purged... Hmpf! ;-)
Posted 21 Dec 2014 by
DoctorNow
86 primes found in the last run
And when and where can we expect a list of the finders?
Would it be possible to get a list of the own found prime numbers in the own account?
Posted 9 Dec 2014 by
DoctorNow
The master file is downloaded, but when I look into the /var/lib/projects/sr../ folder, no application downloaded.
As long as there are no WUs available, an app isn't downloaded as well as far as I know. And there is currently no work.
SRBase: Scheduler request failed: HTTP file not found
SRBase: Sending...
SRBase: Scheduler completed: Got 0 new tasks
I guess this is happening to everyone, also Windows users like me.
First try always results in an http error, second try works.
Must be some misconfiguration serverside I guess...
Posted 2 Dec 2014 by
DoctorNow
Hm, doesn't work under my Virtual openSuse 11.4 (process exited with code 1-error).
But maybe it's too old in either way - didn't update it for a while due to not using it anymore...
Posted 28 Nov 2014 by
DoctorNow
Hm, I don't know how exactly PrimeGrid did solve that, but as far as I observed the fixed credits there for some subprojects do adjust (themselves?) to the length of the WUs - maybe this n value plays a role in the credit adjustment then.
Posted 28 Nov 2014 by
DoctorNow
Rebirther wrote:
Credits are based on FLOPs at the moment
I guess this does mean that CreditNew is in action and does this on it's own, right?
Because I just get these low scores:
28 Nov 2014, 10:27:29 UTC 28 Nov 2014, 12:54:48 UTC Fertig und Bestätigt 1,998.54 1,866.91 5.02
28 Nov 2014, 10:27:29 UTC 28 Nov 2014, 12:59:08 UTC Fertig und Bestätigt 2,180.58 2,034.38 5.56
28 Nov 2014, 10:27:36 UTC 28 Nov 2014, 13:28:34 UTC Fertig und Bestätigt 2,059.13 1,866.67 5.38
Would be nice if you can figure out some fixed credits like on PrimeGrid sooner or later... ;-)
Right now it's only a hint because I don't care much about the credits anyway.