I've got 2 computers here, one a reasonably fast one, and another, a lot slower/older, and I'd like the ability to decide whether to receive repair WU's (i.e. those with a shorter deadline than normal) on a particular device for a particular project. Yes, "in theory", the system should only send WU's out to devices that it thinks can return the WU's in time - but, time, time, time and again, I've received long ARP WU's on my slower computer with shorter deadlines than normal, and I can't recall EVER, my machine returning them within the deadline (and therefore, triggering off another copy being sent out to another device). They're more often than not, either returned "too late" or after the repair WU sent out to replace the repair WU my device received, has already been started - and thus, doesn't then get "server aborted", as it's already "up and running" somewhere else. Thus, wasted CPU cycles spent on crunching them, whereas they could have been used to crunch a normal deadline ARP WU instead.

I've now resorted to manually aborting any shorter deadline WU's that arrive on that particular device - but, there will be times when I'm not able to do this.

Can this be considered as a future enhance ment to the system?