Presumably you are running BOINC. Bring up BOINC Manager, click on the Messages tab, and you will see it starts up listing information. The Device ID, identifying the computer, is on the 7th line for my Windows installation of BOINC.

Guys, the full announcement on this beta program is a bit stealthed, but it sits 'read only' on the member news section....here a copy and bolded a few keys....personally i would have added a condition of some form of competency requirement on using/managing BOINC....sorry!

Would a (temporary) suffix to the forum section title - BOINC Agent Only help? If the title is non negotiable, the description of this forum section up front on the index could use a few (temporary) insertions of the BOINC word.

UNiRAC,

The beta program is an attempt to fix one existing problem and to identify future problems before they get released into the larger community. I am not sure why your P3 stopped crunching, mine is still going strong. If you post details of your problem in an appropriate thread I am sure someone will help you find the answer.

Please join the beta test team when your P3 is up and crunching again.

When we launch a beta test we will do so with a limited set of workunits which have tight deadlines. During this time we will be monitoring the behavior of the application to make sure that it is not causing high error rates. If we see and we have sufficient data to research the problem, then we will shut it down.

Invalid results will be granted credit automatically (as we do for all of our applications). However, we will manually award credit for results that are returned as error only as appropriate and if the errors are in excess of what we see in our normal production applications. Also, the mechanisms that will limit the number of results will need to be left in place. The reason for this is the following:

The credit awarded for beta testing is counted as credit just like any other application. There is a very small number of folks that are participating soley for the purposes of competition who would be willing to exploit any opening to gain more credit. If we automatically award credit for errors, then this may open up a avenue for them to cheat and in doing so artificially increase the number errors returned - thus damaging the usefulness of beta testing. As a result we will only award credit for errors if those errors are due to a systemic application problem and credit will only be awarded manully after a problem has been found.

We also have to leave in place the mechanism that automatically stops a computer from receiving more work if they return too many errors. The reason for this is that there is a certain small percentage of computers that we call 'cyclers'. These computers return every result as an error and can very quickly go through many workunits. If they are allowed to go through the workuntis unchecked, then they can cause serious headaches for the server management (for example, last friday a intel-mac cycler hit the servers and caused 30 workunits to be marked as intel-mac workunits within an hour - if we were running a 200 workunit beta test and we allowed a computer to run without contraints then all of the 200 workunits could get marked as intel-mac workunits thus denying other computers the ability to download work. We can reset this but testing time would be lost). We will be watching the progress and if we observe that the application is experiencing errors - we will restore members download limits so that they can fetch work.

These steps have to be manually becuase the computers that can cause harm can do so very quickly - so the defense against those computers needs to be automated.

I hope that the community understands the need to do things this way. We do value the additional support that the beta testers have committed to providing.

I would like to thank you all again. I just checked and we now have 729 members who have signed up to be beta testers. Thanks!

Hello,

Even if a huge number of the appropriate errors occured, would you grant credit for each of them? It would be a tough task very much:(

Please keep in mind that in beta testing there are some people who devote their machines for testing, not so much for credits. Of course it would be better for us to be granted credits than not. However, if granting is too hard task for you, credits needn't to be granted although an error occured on the work units, since in testing errors can occur at high rates.

Of my experience, if participants have been all sure that an error can happen at any rates, they don't tend to blame for developers for it. For example, RALPH@Home (Rosetta ALPHA project), testing some new implements of Rosetta@Home to improve it and keep it from being in trouble in testing, applys the same system as other projects: sending work units to computers, computing them and sending them back to the central server. Because the work units are being tested, some errors can occurs. However, the participants have never blamed developers for the error or even requested developers to grant credits for it; rather they gladly post comments on the message board around the error they faced. I think they might be "proud " of participating the project, since the alpha project determines how rosetta@home will be developed. Surely the reason why they don't blame developers may be that one work unit requires just one or two hours to be completed. If the work unit required ten hours or more, they might require some credits. But, since the implements tested have been added so far and they realise that their participating were rewarded, nobody complains about neither occuring errors nor not granting credits.

Whether credits would be granted or not isn't the right problem in testing. If one want credits more, he/she just shouldn't participate in the beta test and contribute to the other projects. If we regard the test as worthy to join in regardless credits, then they will do so