I don't think that's possible.

WCG uses its normal user base (or rather the subset of it that opted to participate) for beta tests. Only known computers are used for this which are considered "reliable" based on recent faultless workunit crunching in the normal WCG projects. Therefore you would never be considered for a BETA unit even if you managed to manipulate the checkboxes in such a way (which I doubt you can, unless there's a bug in the site somewhere).

To be eligable for beta workunits you need to participate in normal WCG projects aswell.

As a suggestion here (and please, anyone jump in if I'm wrong on this), why not run a project with short WU's - as, I'm led to believe, that if you've got the 'Beta' box checked and a very small work cache (i.e., there's no point if you've got 5 days of work cached up - there's no chance of you ever getting any Beta WU's in that scenario), BOINC checks to see if there's any Beta WU's available before checking your other selected project(s) - which ever they may be.

Therefore, your Macs have a chance to build up their 'reliability index', so that if/when there's some CEP Beta work for that platform, your machines are ready to accept it...

In summary: In order to participate in the CEP beta, one has to crunch non-CEP projects. Something is wrong with that set-up, and needs to be fixed.

The World Community Grid beta test pool is so large that there is no shortage of potential test computers, and the pool is suitably diverse.

So, nobody needs to go to exceptional lengths to participate in the beta project.

The problem is that people want to participate (trying to get that badge, I would guess).

The bottom line is there isn't enough beta work to go around. This is a good thing from WCG's point of view, so don't expect it to change.

The FAQ says 21 hours return time history. Theoretically even if you set WCG to a relative low resource weight of say 5%, but active, it would fetch /return work within that limit, on a little buffer.

The time frame for someone dropping off the reliability list is not known, but from information on this forum it seems weekend crunchers do maintain the "reliability" rating.

_Sunnyboy_ scored maximum points on his assessment and continues for the bonus questions.

1) What is known about the client/server contact deferral time and increments when no work is available for a platform?

2) How likely is a "Beta only" client going to hit the servers right when that 1 hour Mac window is open to speed off those ~1000 test units?

3) Is there an incentive plan behind keeping devices "reliable" in an environment where a beta batch is snapped up in no time?

Some things WCG does not publish. I would not either to include looks into the feeders how much there is and most certainly not the Beta task feeder. Red lined messages in the log that no work is available are a good reference.

sort of 2 Euro cents Sunday evening thoughts, before coffee.

PS: A device crunching 45 hours on a CEP job like my laptop just did probably would cause an extradition, but given it's dual and me making sure to never do more than 1 CEP job at the time, up to 20% throughput improvement, it keeps chugging results at 7 hours on the other cores.