My professor's kid got posted on Slashdot

[Belisarius]

My professor helps in the development of Java, and his kid apparently got tired of hearing people bash Java for being slow, so he ran a bunch of tests that showed Java, when run right, would often beat C++ by a hair. Oddly enough, it eventually got picked up on Slashdot, so we all gave our professor kudos on properly raising an uber-nerd.

Slashdot article

The Java is Faster than C++ and C++ Sucks Unbiased Benchmark

[sde]

cool, .. how old is the 'kid' ?

[Belisarius]

College age, which is 20, plus or minus.

[technobard]

I saw the /. article. I may have to go back and read the details now.

[npa]

there is absolutely no way java can be
faster then c++, in the specialised case.

secondly, the code on the site is not even
well-written, nor does it include the actual
timing code that he has implemented ...

consider this, 'fibo', from the c++ version:

    if (n < 2)
	return(1);
    else // useless else
	return(fib(n-2) + fib(n-1));

and the 'else' is left out of the java version:

	if (n < 2) return(1);
	return( fib(n-2) + fib(n-1) );

, despite this - the original comment stands - there
is absolutely no possibility for java to beat c++ in a
specialised case.

that said, i prefer developing in java.

[Belisarius]

To be fair, I think the entire thing was a bit tongue-in-cheek, he just wanted to show that Java isn't as slow as people kept saying it was. It was a bit of turn-around from C++ programmers mangaling Java code when running tests.