[eastbay]

[ QUOTE ]
I don't know about his friend, but if he has to do it, he has to learn the language and development time will be huge. Also, development time is human time, which in some rare cases is more valuable than computer time. Plus, it's almost certainly a one-off thing with no practical purpose or deadline. Also, if you do have a purpose to reuse it a lot it isn't too hard to rewrite and/or a lot of easier to learn and develop languages can handle extensions in C for the crunching (I know this program is mostly crunching).

Ok, Ok, I'm just a PHP/Perl hack, I admit it.

[/ QUOTE ]

All valid points.

On the other hand...

I know a little bit about doing simulation work. If a code can turn around a result in an hour vs overnight, the difference in utility is vast. When it takes a day to turn around a result, it's too easy to forget exactly what your last run was supposed to show, and just give up. It's a human factors thing. Nobody has the patience to get good results out of a slow code.

So there's a strong nonlinear utility to fast execution times for these kinds of sim tools.

eastbay