no problem with publishing that code ... although there is no nice doc yet
last weekend I worked on gcc compatability, therefore some stuff isnt funny debugged yet, especially CTime::CPUSpeed::getCyclesDifference.
Recording maximum and minimum cpu usage was once a feature but I had to remove it due to restructuring the code ... maybe that's something for the next weekend after the exams
this code has only be tested on a p4 2.66 and a Athlon1Ghz.
and pierre, I think the nose is the best method to profile, although you can use the debugger sometimes as a debugger, that's interesting too
