This project has certainly introduced a higher error rate then we typically experience. We will take a look at the discussion above. There is certainly some merit in the argument that errors that occur during beta testing should be treated differently then errors that occur in production.

We will have to consider this while considering the following issues:

* The mechanism should reward members appropriately for their contribution
* The mechanism needs to be part of the BOINC framework so that it works automatically
* The mechanism should discourage cheating and overclocked computers that return erroneous results.

we will not be able to make any changes quickly. With a bit of luck there will be three beta's this week (I am about to start beta testing version 6.30 for CEP1) and we will have a small test for FightAIDS@Home and Dengue Fever (small change to some settings in header files to allow for some additional parameters in the workunits).

I just picked up 2 of the new Beta WU's for CEP. Estimated run time is 17.5h for each one.

Are the new Beta's really going to take this long? My prior Beta's only took about 6h each.

As the shareholders reports always include "past performance is no guarantee for the future" . BOINC just takes the estimated flops in the header, for beta I guess manually assigned, and takes the latest benchmark values to compute a TTC. With substantial varying flops in several projects, to name HFCC and CEP and the change in mean run time target from 6 hours to 7 hours for most WCG projects, it will be another week or longer before the duration correction factor stabilizes, if it's allowed to do so.

the last set of beta's I received from cep, they were estimated at 34hrs, but ended up taking 18hrs

the 2 beta's I received today are showing 34hr's to complete but I am betting on them only lasting 18hrs

I got 2 hours ago BETA for CEP. Estimated duration almost 45 hours. Huge... I hope it wont be inconclusive after I crunch through