I don't understand how results that pass validation can be bad.
Some more info please.
One scenario would be if you had an opti version that failed when teamed with the stock version. What happens if it gets teamed up with another host with the same opti version?
I don't understand how results that pass validation can be bad.
Some more info please.
One scenario would be if you had an opti version that failed when teamed with the stock version. What happens if it gets teamed up with another host with the same opti version?
Alinator
all mine have validated against computers using the stock app
Please, don't use any patched versions of Einstein@Home app.
The bad results caused some problems in the database.
Thanks!
edit: Probably these units will be send out again. ( about 5000 WUs at moment )
Hi Akos, how severe is this? Do we need to produce instructions for people to strip off all patches and get back to the 4.02 stock application? Will we be able to move forward with patches again later? Mike
I would think all you need to do is delete the app and/or app_info files and then update the project.
Alinator
I'm not sure it is quite that simple. If people have changed to reporting as a 4.10 application, they may need to set no new work, drain their cache or abort all WUs, stop BOINC, remove the app_info.xml file, start BOINC, reset the project etc, otherwise they may trash a load more WUs. A full set of instructions can be produced, but it is as much work going back safely as it was going to the patched application in the first place. There are a number of possible states people can be in. Mike
I guess the most important thing **right now** is to set No New Work/No New Tasks on Einstein if you are running a patched version to stop anything else downloading until this get resolved and there is a good plan in place. Mike
I'm not sure it is quite that simple. If people have changed to reporting as a 4.10 application, they may need to set no new work, drain their cache or abort all WUs, stop BOINC, remove the app_info.xml file, start BOINC, reset the project etc, otherwise they may trash a load more WUs. A full set of instructions can be produced, but it is as much work going back safely as it was going to the patched application in the first place. There are a number of possible states people can be in. Mike
I am suspending E@H and may detach. The employment of Akos as a consultant seems to have been completely mishandled. We have an only half-optimised stock app and another round of chaos and insurrection - exactly what I thought 'they' were trying to avoid.
I don't understand how results that pass validation can be bad.
Some more info please.
One scenario would be if you had an opti version that failed when teamed with the stock version. What happens if it gets teamed up with another host with the same opti version?
Alinator
all mine have validated against computers using the stock app
Be that as it may, the objective is to enable EAH to maintain continuity for the results in the MSD, therefore since these are experimental versions, stable or otherwise, it's best to honor Akos' request until he has a chance to evaluate all the ramifications. Remember, the science app is *not* GPL, AFAIK.
I am suspending E@H and may detach. The employment of Akos as a consultant seems to have been completely mishandled. We have an only half-optimised stock app and another round of chaos and insurrection - exactly what I thought 'they' were trying to avoid.
Well, that may be a little harsh. :-)
I wonder if UMW has the resources to spare to start a beta project like SAH does for pushing the envelope. ;-)
I am suspending E@H and may detach. The employment of Akos as a consultant seems to have been completely mishandled. We have an only half-optimised stock app and another round of chaos and insurrection - exactly what I thought 'they' were trying to avoid.
Well, that may be a little harsh. :-)
I wonder if UMW has the resources to spare to start a beta project like SAH does for pushing the envelope. ;-)
Alinator
That would be an ideal solution in the longer term, I agree. Meanwhile, it would be useful to have some more information from Akos or Bernd about what we need to do to protect the project most effectively - is it a calm retreat, is it an immediate abort of all WUs, is it just the dodgy versions of the application that are an issue. Please tell us and we can get out of this with minimum damage to the project. Mike
RE: I don't understand how
)
One scenario would be if you had an opti version that failed when teamed with the stock version. What happens if it gets teamed up with another host with the same opti version?
Alinator
RE: RE: I don't
)
all mine have validated against computers using the stock app
98SE XP2500+ @ 2.1 GHz Boinc v5.8.8
RE: RE: RE: RE: Hi
)
I guess the most important thing **right now** is to set No New Work/No New Tasks on Einstein if you are running a patched version to stop anything else downloading until this get resolved and there is a good plan in place. Mike
RE: I'm not sure it is
)
Agreed, that's why I PS'ed.
Alinator
I aborted all WUs I had - but
)
I aborted all WUs I had - but I cannot access all computers now :-/
Two PCs @ work I will not be able to access before wednesday :-(
Same here, I aborted all
)
Same here, I aborted all mine, too.
I am suspending E@H and may
)
I am suspending E@H and may detach. The employment of Akos as a consultant seems to have been completely mishandled. We have an only half-optimised stock app and another round of chaos and insurrection - exactly what I thought 'they' were trying to avoid.
Dead men don't get the baby washed. HTH
RE: RE: RE: I don't
)
Be that as it may, the objective is to enable EAH to maintain continuity for the results in the MSD, therefore since these are experimental versions, stable or otherwise, it's best to honor Akos' request until he has a chance to evaluate all the ramifications. Remember, the science app is *not* GPL, AFAIK.
Just my opinion.
Alinator
RE: I am suspending E@H and
)
Well, that may be a little harsh. :-)
I wonder if UMW has the resources to spare to start a beta project like SAH does for pushing the envelope. ;-)
Alinator
RE: RE: I am suspending
)
That would be an ideal solution in the longer term, I agree. Meanwhile, it would be useful to have some more information from Akos or Bernd about what we need to do to protect the project most effectively - is it a calm retreat, is it an immediate abort of all WUs, is it just the dodgy versions of the application that are an issue. Please tell us and we can get out of this with minimum damage to the project. Mike