> @wijata.com: Please post the Result ids or names, or at least the id of the
> machine you did this on. I'll take a look at this.
result id 1311263, hence the computer was known to have some memory issues in the past.
resid 1286319 and resid 1283956 comes from machines known as working good.
All thouse machines was previously run Folding@Home - no problems at that time, however the computing was surely different.
This wu was assigned to 2 windows systems and 2 linux systems. Looking at the results, the bytecount and checksums for the linux systems are identical. The windows systems are not identical, but obviously within the tolerance range.
Now, the 2 windows systems returned their results first so that set the determination for what is "valid" when the first linux result came in and didn't match. But, if the 2 linux results match exactly, how can it be determined that they are wrong (hence "invalid") and they get 0 credit while the "close enough" windows results are accepted as correct and given credit?
On this particular wu, what would have happened had the 2 linux systems returned their results first? It hardly seems that simply crossing the finish line first should either validate or invalidate science.
All that said, I'm really not overly concerned with points, but I really hate it when my contribution is useless. Thus far, I'm running about 40% invalid on einstein units across 5 linux systems - and that's very frustrating.
lin machine
Fstats.Ha: bytecount 1601439 checksum 76844773
Fstats.Hb: bytecount 1287439 checksum 61769950
(not that result are identical, thus granted zero credits)
Windows machines happen to be first.
We may officialy have conspiracy theory now ;)
No, it happen that is was third (3rd) result - not 4th. 4th result was also from linux and also zero credits. Identical results from different linux machines was marked invalid, while not identical from windows are OK.
And moreover. More and more my linux machines gets zero credits for work.
I don't belive all my linux machines are faulty... Some of them are servers working well for some long time.
Could some developer explain how WU are validated (what is validated)
Or maybe the problem is, that I downloaded 4.24 version (hence i use proxy and there is statement, which made me download 4.24 first)
Maybe it's just a bug in Linux version ?
I know little about the inner workings of BOINC... it's all Voodoo to me =) Having said that here is a simple observation from me...
On Errors:
Running:
AMD 3200+, WinXP Pro, 1 BG DDR Ram in dual mode.
I frequently shut down BOINC (Two to three times in an evening) because I am a PC Gamer and under certain circumstances I need as little overhead as I can accomplish for my PC. E@H seems to be operating A typical when compared to the other projects I am running. It returns bad units, asks for restarts, and never lists credits.
On the credits issue:
If CPDN can trickle, can E@H? ... it would generate a lot of good will if nothing else.
@wijata.com: Please post the
)
@wijata.com: Please post the Result ids or names, or at least the id of the machine you did this on. I'll take a look at this.
BM
BM
> @wijata.com: Please post
)
> @wijata.com: Please post the Result ids or names, or at least the id of the
> machine you did this on. I'll take a look at this.
result id 1311263, hence the computer was known to have some memory issues in the past.
resid 1286319 and resid 1283956 comes from machines known as working good.
All thouse machines was previously run Folding@Home - no problems at that time, however the computing was surely different.
OK, here's another one to add
)
OK, here's another one to add to the mix. I think it's a conspiracy against us Linux users ;~|
396825
This wu was assigned to 2 windows systems and 2 linux systems. Looking at the results, the bytecount and checksums for the linux systems are identical. The windows systems are not identical, but obviously within the tolerance range.
Now, the 2 windows systems returned their results first so that set the determination for what is "valid" when the first linux result came in and didn't match. But, if the 2 linux results match exactly, how can it be determined that they are wrong (hence "invalid") and they get 0 credit while the "close enough" windows results are accepted as correct and given credit?
On this particular wu, what would have happened had the 2 linux systems returned their results first? It hardly seems that simply crossing the finish line first should either validate or invalidate science.
All that said, I'm really not overly concerned with points, but I really hate it when my contribution is useless. Thus far, I'm running about 40% invalid on einstein units across 5 linux systems - and that's very frustrating.
Darren
I'm afraid You are 100
)
I'm afraid You are 100 percent right
Look also here http://einsteinathome.org/workunit/369574
win machine
Fstats.Ha: bytecount 1601439 checksum 76686342
Fstats.Hb: bytecount 1287439 checksum 61644201
win machine
Fstats.Ha: bytecount 1601439 checksum 76686342
Fstats.Hb: bytecount 1287439 checksum 61644194
(note that checksum differs, hence both machines was granted credits)
lin machine
Fstats.Ha: bytecount 1601439 checksum 76844773
Fstats.Hb: bytecount 1287439 checksum 61769950
lin machine
Fstats.Ha: bytecount 1601439 checksum 76844773
Fstats.Hb: bytecount 1287439 checksum 61769950
(not that result are identical, thus granted zero credits)
Windows machines happen to be first.
We may officialy have conspiracy theory now ;)
If i understand right - the
)
If i understand right - the evaluation is done upon first 3 results reported.
It just happened that your linux machine was the last one to report
No, it happen that is was
)
No, it happen that is was third (3rd) result - not 4th. 4th result was also from linux and also zero credits. Identical results from different linux machines was marked invalid, while not identical from windows are OK.
And moreover. More and more my linux machines gets zero credits for work.
I don't belive all my linux machines are faulty... Some of them are servers working well for some long time.
Could some developer explain how WU are validated (what is validated)
Or maybe the problem is, that I downloaded 4.24 version (hence i use proxy and there is statement, which made me download 4.24 first)
Maybe it's just a bug in Linux version ?
hi, i'm
)
hi,
i'm using
http://www.pperry.f2s.com/files/boinc_4.19_pentium4-pc-linux-gnu.tar.bz2
http://www.pperry.f2s.com/files/boinc_4.19_athlon-xp-pc-linux-gnu.tar.bz2
on a laptop and a ws.
all get invalid results
http://einsteinathome.org/task/1358689
http://einsteinathome.org/task/1325488
http://einsteinathome.org/task/1325484
I'm having the same results,
)
I'm having the same results, output looks good but got no credit.
Computer id=45532 WU's are 379527, 375913, 372616 and 368687
> I'm having the same
)
> I'm having the same results, output looks good but got no credit.
>
> Computer id=45532 WU's are 379527, 375913, 372616 and 368687
>
just finished and uploaded
invalid
http://einsteinathome.org/workunit/372656
I know little about the inner
)
I know little about the inner workings of BOINC... it's all Voodoo to me =) Having said that here is a simple observation from me...
On Errors:
Running:
AMD 3200+, WinXP Pro, 1 BG DDR Ram in dual mode.
I frequently shut down BOINC (Two to three times in an evening) because I am a PC Gamer and under certain circumstances I need as little overhead as I can accomplish for my PC. E@H seems to be operating A typical when compared to the other projects I am running. It returns bad units, asks for restarts, and never lists credits.
On the credits issue:
If CPDN can trickle, can E@H? ... it would generate a lot of good will if nothing else.
Thanks
-Kio