25th, and the FightAIDS@home site is off the air.

26th, the FA@H site is still down. First full day of crunching and 184,208 results validated for 16.8 years runtime.

Seippel, thanks for kicking it.

Kick it again, Seippel - I've gotten non-FA@H tasks again within the past 6-8 hours.

Maybe demand is greater than the 'not the same priority as before' supply [words to that extent].

So I got a device that has -All- selected and app_config.xml is set to run max 6 MCM and 1 CEP2. Yesterday, no OET, no UGM, no FAHV resulting in an idle core, client loaded with 1 day of MCM and few CEP2 **. If this is fair project weight distribution, I don't know what is, but in practical terms, it's MCM+CEP what my clients get at 90 out of 100. It's almost as if a budget is set for the lower priority projects in terms of # of tasks circulated... if this budget is exhausted, no more of those for the rest of the day.

** Of course, since the buffer is fully loaded with MCM+CEP2, no asking for work for a while, so the lower supply stock can be out again, more idle coring. Since the buffer is filled with 8 core days, but only 7 are crunching, the longer it takes to get through the buffer and fetch new work. What happens at this point is, the client goes fetching work elsewhere from the zero weight set projects. So here is another reason for Sgt.Joe why production is not as high [point 3], not to speak of vacation of course [point 4].

So tell me [again], why do we have app_config?... To see computing time go to waste? With the BOINC governance program, maybe time to rip out the [more control what IS processing] in the next version!