It has been addressed: WCG is self-balancing. WCG techs decide how much power is put into each project. Let's say they decide that 50% should be spent on HCC; 15% FAAH; 15% HPF2; 15% DDDT; 5% ACAH.

In the grand scheme of things, it doesn't make an ounce of difference if you and 20 other teams all stop crunching HCC; the HCC work will continue to get 50% of the processing because the majority of users here select all projects. The server dishing out the WUs (I forget what they call it) will automatically re-balance the workloads so the percentages stay the same.

Have to be careful here, but they're all called 'Daemons'. There's a Start Here FAQ that includes a schematic of the 10 or so. The centre-piece is the Scheduler Daemon which continuously checks what work goes out, what comes back and to what platforms it is going and coming from. Considering 5 projects on 6 Homogeneous Redundancy groups + 2 UD projects and all balanced so that everything syncs up such that the batches of each project complete in a timely and coordinated fashion is a marvel. But it is as you explained with the change coming that DDDT is being jacked up again to get its equal share, which it is not getting presently due to WU volume incurred bottlenecking. That I understand is overcome and is countered in part by making DDDT longer (more work in a Task). Effectively Roger The Dodger can put it soon back into his profile without clogging his ADSL/Cable. (He's doing 7x more CPU time as courine, but 14 times more jobs per the member comparison statistics ;>).

Here, try this. To simplify matters, lets take 2 systems only:

System #1 is a single core faster than 1GHz and not PC100. Lets say its a 2.4 GHz P4 with 1gig of memory.

System#2 is a multi core of any speed, but is in all other ways the same as System #1

The self-balancing system you tout does not see that system 2 would be better for non-HCC work, as there is work in-progress to resolve the current bug fix regarding the HCC project.

Right now, HCC work units are being sent to machines that would be better elsewhere. Conversely, the opposite is true also. This thread was, and I mean WAS, to address this issue. The problem that is not a game at all is why it hasn't.

So now this thread has been completely sent off track and delayed during a time it would be most needed. Why? When the matter is civically as well as technically this obvious.

There is more to this, but please understand, I am hardly the enemy. For if I was, I wouldn’t take the time to correct the problem.

It works to also to move the single core public into HCC in return. This way it would keep both sides happy.

The message is to move away from HCC if it hurt in any way, as well as bring about the civil awareness to move single cores to HCC.

If I run using the self-balancing system, I would loose about 10% in my score. Roger would also. The only person in my team that can run HCC effectively is Cargod. So have “fung-shwayed” my group to recover the lost efficiency.