Basics
Project information:  
Create account
Your account
Teams
Download BOINC
Add-ons
Community
Participant profiles
Message boards
Questions and answers
Donations/Sponsors
Live Games
Statistics
Top countries
Top participants
Top computers
Top teams
Server Status
Other statistics
|
1) Message boards : :
Number crunching :
Invalid
Posted 1525 days ago by svenni96
This problem is known and is also handled in another Therad, http://www.chess960athome.org/alpha/forum_thread.php?id=29
It seem that the application has sometimes problems to finish a wu. So there will be difference in the resultfiles and you did not get any credits.
But the WUs here are very small and fast to compute. You will only waste a small amount of processor time for that invalid WUs.
And its still ALPHA.
|
2) Message boards : :
Number crunching :
new daily quota
Posted 1531 days ago by svenni96
Right now for me. Just reached the 120 wu.
|
3) Message boards : :
Number crunching :
Some very first test w/u with a Chess960 application
Posted 1532 days ago by svenni96
EDIT. Forget that question. I found a solution to my problem myself.
|
4) Message boards : :
Number crunching :
daily quota
Posted 1540 days ago by svenni96
Well, thatīs a good idea.
In the meantime I have to cheat. ;)
|
5) Message boards : :
Number crunching :
daily quota
Posted 1540 days ago by svenni96
Is it possible to increase it once more?
Have reached the daily quato with one of my computers. :)
|
6) Message boards : :
Number crunching :
Some very first test w/u with a Chess960 application
Posted 1540 days ago by svenni96
I think there is no problem but it shows that the validator works fine.
When no credits will be granted in the most cases the result isn`t correct in all details.
When your result received to the server the validator starts to compare it with another received result for the same workunit. If there are diffences (rounding difference because of AMD and Intel processors,a missing value because of restart or suspending the wu) both results get "Validate state = Checked, but no consensus yet". And, if only two WUs was send out, a third one will be send out.
After receiving that third result, the validator starts again checking the new result against the two old. If it is identical with ons of these, that incorrect result get the state = Invalid and no credits will be granted.
So, it seem that your four results are not correct.
I have very often that problem when crunching Leiden-WUs.
|
|
