My AMD Athlon 1.4GHz system is 50% S@H and 50% CPDN. It trickles to CPDN roughly once each 2.5-3.0 days....and it is turned off each night.

For quite a while now the Recent Average Credit has been falling....and is now down to 1.33.....which is odd, given this system is a consistent performer.

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/show_host_detail.php?hostid=56344

What\'s happening?

---

I think it\'s because I\'m using an \"average trickle date\" and this workunit has been going for a few months, so the \"average date\" is far back enough that the exponential decay in the RAC algorithm makes the RAC pretty small! I\'m going to have to rework this calculation so that it favors more recent trickles (CPDN doesn\'t really fit into the BOINC/SETI RAC paradigm).

at the urging of John Keck (and many others! :-) I am redoing the RAC calc which should hopefully be more fair & more in line with other boinc projects --- basically the RAC for a result will be the last trickle value (the incremental value, i.e. 141 for the slab model) if the trickle was received in the last day (so that you\'re not \"penalized\" if you trickle just after the stats were recalc\'d). Then it will exponentially decay from there. So we should be seeing more sensible numbers soon.

Then it will exponentially decay from there. So we should be seeing more sensible numbers soon.

Thanks very much. The RAC is now 29.<summat>...which makes more sense.

Steve

---