6 Jun
2008
6 Jun
'08
8:04 p.m.
Frank Mori Hess:
I'm not quite sure what you're driving at with the above 2 paragraphs. Are you satisfied with your pseudocode scheduler's performance, or no?
I'm not sure, as it's hard to measure the performance of a pseudocode. :-) To see how it fares, we need a benchmark based on a workload that is representative for your intended use cases... one that cannot be trivially rewritten to perform better using an alternative concurrency model. ;-)