On Fri, 4 Nov 2011 09:45:58 -0600 Ian Mallett <geometr...@gmail.com> wrote:
> Good C++ is always faster than Java code. > However, crappy C++ is slower than crappy Java code. The flaw here is that the underlying assumption is that execution time is THE criteria for evaluating code quality. Contrarily, there are often instances where one might prefer slower performing code if more readable / better structured / easier to maintain. It might be true that the C++ implementation of any algorithm is faster than its Java counterpart, though. As well as it is true that one could write an optimised-for-speed java implementation of something that outperform a non optimised C++ version. Both scenarios have nothing to do with "crappiness" of a given implementation though. :o /mac