What about them ??
Is it really bad for the project ?
Can you send them personal message or something..?
Yes I’ve noticed this too. I have a couple of WUs pending where my co-cruncher is still using a patched optimisation. He has 81 results, all of his validations have failed to validate and have been granted zero credits, but this does not stop him. He is still allowed to download the maximum 32 WUs per day. Could the scheduler stop sending WUs to these people until they have removed the patched file as it did when we moved to the new S5 WUs? dAVE.
Only way to easely stop this is bring out a new version of the einstein@home program with a higher version number.
then all older versions and patched versions will be overwriten and the problem solved.
I have a couple of WUs pending where my co-cruncher is still using a patched optimisation. He has 81 results, all of his validations have failed to validate and have been granted zero credits, but this does not stop him.
Hm, he is using E@H S5R1 4.02 SSE3-0710 TEST
Thats a version known for producing invalid results..... Really bad that people use such bad versions and give everyone a reason to blame people who want to use "good" optimized versions like 0712........
edit: any very strange his WU Quota doesn't go down....
Version Stable producing bad results too ? (E@H S5R1 4.02 SSE-0007 STABLE)
Because users using it get always credits: http://einsteinathome.org/task/36001685
Version Stable producing bad results too ? (E@H S5R1 4.02 SSE-0007 STABLE)
Because users using it get always credits: http://einsteinathome.org/task/36001685
I confess I don't see the problem. Crunchers were asked to stop using *all* the Akos patched versions, Stable or otherwise. See http://einsteinathome.org/goto/comment/40178 and other posts. There are no versions other than the official project versions that were deemed acceptable for the moment. Just because a WU is still getting credits doesn't avoid this, the validator is obviously struggling to separate approved results from non-approved results. If or when it learns how to do it and and if it cancels all credit for non-official results back to the original STOP request people who have not complied with these requests have only themselves to blame, I think. Mike
It is very easy to cancel credit for optimised results, isn't it ?
Why einstein's staff don't do this ?
That's why we can believe it's not really a problem to use optimisation and are still using it...
Why to give credit to users using Akosf opti ? because they won't stop !
It is very easy to cancel credit for optimised results, isn't it ?
Why einstein's staff don't do this ?
That's why we can believe it's not really a problem to use optimisation and are still using it...
Why to give credit to users using Akosf opti ? because they won't stop !
It's only a handful of people working for the project. They are busy trying to comb through the database (read Bernard's message). I hope they back all that credit out, and prove to those who are using it, that it's bad.
They could fix some of this quickly by pushing out a new version, with a quick code change that will not allow the older version to crunch. It's been suggested, and it will probably happen, just as to when? Bruce has been gone for a while, and has just returned, or returning this week. So they have been without one of the main developers, etc.
Remember that this project it partially in the US (University of Wisconsin - Milwaukee, WI USA) and they are on summer break for school. So, school is out, which also means limited availability. This last week was Independence Day (the national birthday/holiday) giving more people off.
Ok I am not a computer god. So Please tell me Exactly what I should do to make sure that I do not have the wrong app installed. Or if it is easer Exactly how to determine if I have the correct app installed.
Mine says 5.4.9
Start of BOINC application 'projects/einstein.phys.uwm.edu/einstein_S5R1_4.02_windows_intelx86.exe'.
Is this acceptable. I do not want to contribute bad work units
Ok I am not a computer god. So Please tell me Exactly what I should do to make sure that I do not have the wrong app installed. Or if it is easer Exactly how to determine if I have the correct app installed.
Mine says 5.4.9
Start of BOINC application 'projects/einstein.phys.uwm.edu/einstein_S5R1_4.02_windows_intelx86.exe'.
Is this acceptable. I do not want to contribute bad work units
You're good. What you check is the text of a workunit result. The akos apps add a line into the output indicating which version of his modifications they are.
RE: Lot of users are still
)
Yes I’ve noticed this too. I have a couple of WUs pending where my co-cruncher is still using a patched optimisation. He has 81 results, all of his validations have failed to validate and have been granted zero credits, but this does not stop him. He is still allowed to download the maximum 32 WUs per day. Could the scheduler stop sending WUs to these people until they have removed the patched file as it did when we moved to the new S5 WUs? dAVE.
Only way to easely stop this
)
Only way to easely stop this is bring out a new version of the einstein@home program with a higher version number.
then all older versions and patched versions will be overwriten and the problem solved.
but only time will tell what will happen
RE: I have a couple of WUs
)
Hm, he is using E@H S5R1 4.02 SSE3-0710 TEST
Thats a version known for producing invalid results..... Really bad that people use such bad versions and give everyone a reason to blame people who want to use "good" optimized versions like 0712........
edit: any very strange his WU Quota doesn't go down....
message from Bernard stating
)
message from Bernard stating people should stop using it.
Version Stable producing bad
)
Version Stable producing bad results too ? (E@H S5R1 4.02 SSE-0007 STABLE)
Because users using it get always credits:
http://einsteinathome.org/task/36001685
RE: Version Stable
)
I confess I don't see the problem. Crunchers were asked to stop using *all* the Akos patched versions, Stable or otherwise. See http://einsteinathome.org/goto/comment/40178 and other posts. There are no versions other than the official project versions that were deemed acceptable for the moment. Just because a WU is still getting credits doesn't avoid this, the validator is obviously struggling to separate approved results from non-approved results. If or when it learns how to do it and and if it cancels all credit for non-official results back to the original STOP request people who have not complied with these requests have only themselves to blame, I think. Mike
It is very easy to cancel
)
It is very easy to cancel credit for optimised results, isn't it ?
Why einstein's staff don't do this ?
That's why we can believe it's not really a problem to use optimisation and are still using it...
Why to give credit to users using Akosf opti ? because they won't stop !
RE: It is very easy to
)
It's only a handful of people working for the project. They are busy trying to comb through the database (read Bernard's message). I hope they back all that credit out, and prove to those who are using it, that it's bad.
They could fix some of this quickly by pushing out a new version, with a quick code change that will not allow the older version to crunch. It's been suggested, and it will probably happen, just as to when? Bruce has been gone for a while, and has just returned, or returning this week. So they have been without one of the main developers, etc.
Remember that this project it partially in the US (University of Wisconsin - Milwaukee, WI USA) and they are on summer break for school. So, school is out, which also means limited availability. This last week was Independence Day (the national birthday/holiday) giving more people off.
Ok I am not a computer god.
)
Ok I am not a computer god. So Please tell me Exactly what I should do to make sure that I do not have the wrong app installed. Or if it is easer Exactly how to determine if I have the correct app installed.
Mine says 5.4.9
Start of BOINC application 'projects/einstein.phys.uwm.edu/einstein_S5R1_4.02_windows_intelx86.exe'.
Is this acceptable. I do not want to contribute bad work units
RE: Ok I am not a computer
)
You're good. What you check is the text of a workunit result. The akos apps add a line into the output indicating which version of his modifications they are.