On 1/14/10, David Anderson <[email protected]> wrote: > Right now there's a problem: there's no clear definition of credit. > Projects with GPU apps have to come up with ad-hoc formulas for credit, > and all projects have to fiddle around with "normalization" factors. > > This creates an unhealthy environment in which credit-motivated > volunteers pressure project admins to grant more credit, > and volunteers select projects based on their credit policy. > > This will go away (or at least subside considerably) > when we implement the new credit system, > which eliminates fudge factors; > all projects will grant credit in exactly the same way. > > In my view, credit-based competition is good for volunteer computing, > but only if credit is an objective measure.
Credits *can't* be an objective measure. There is no objective way to compare one hour of protein folding with one hour of prime checking. I have yet to see how the "new credit system" will help in this scenario, which in my opinion is an inherent problem and not solvable by changing how many credits are granted: http://www.boinc-wiki.info/User:Nicolas/Credit_scenarios#CPUs_and_GPUs -- Nicolas _______________________________________________ boinc_dev mailing list [email protected] http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev To unsubscribe, visit the above URL and (near bottom of page) enter your email address.
