I love the thought that the WCG is 25% more productive now! All just after a few days. I suspect though that it is going to take another week for pending validations to settle down to their new normal levels for us crunchers. I was typically seeing about 2-8 hours of WUs in Pending Validations, generally either 1 FAAH or 1-3 HPF. Now I am up to 28.5 hours of WU, 2 FAAH and 5 HPF. This is 239 claimed BOINC credits or 1673 points. That's right on the 8-8.5 credits per CPU hour I normally see on my machine so that's still in line. However, 3 of the 7 WU's in PV have already been sent out to a fourth machine (all 7 have two of the original three at PV Status). I expect the "dominance" of HPF in PV is of no statistical significance. With this change to send only to 3 users initially, there is even less reason to manage your WUs. Aborting a WU that already has three valid results as I have read some do should no longer be worth bothering with if even possible. If you are not one of the initial three, then your the fourth for a delayed WU. If you're one of the initial three, processing WU's in the order received is probably going to be the best approach. If you have WU's that get sent to a fourth user because you haven't processed them yet, then you probably need to cut back on the value of your Connect to Server setting in your profile.

No point with the UD agent as it doesn't use the same validation process - it validates immediately.

It awards points immediately. I don't know if there is some kind of penalty mechanism for when an invalid result is identified, but in the UD system, points and validation are separated.

Hey Dagorath

This is what happened from taking your advice

Marine jokes aside.... on the UD validation (a separate thing from points), i would not be surprised if the same rules are in place....i.e. Kevin (knreed of WCG), sending out groups of 3 to the UD agents and giving them the 4 or 7 days to come back, then the 4th copy being send etc....that's entirely something only the WCG can tell us. It would devy logic not to. The 23 / 25% performance improvement would thus be across the board.