Re: [Chicken-users] Advantages of CHICKEN over Gambit-C

[Brandon J. Van Every]

Matthew David Parker wrote:
> Yes, I guess I want to know which one really works better and for what
> types of programs.  I noticed gambit did fibonacci faster than chicken,
> yet it seems to be choppy on my little game.
I'm sure people can stick their finger in the air and tell you various 
things.  They may even have specific experience with various problem 
domains that back up their assertions.  But the bottom line is, what are 
*YOUR* problems?  If you're doing games, and Chicken works and Gambit 
doesn't, there you have it.  Benchmarking always has to be based in 
reality.  It doesn't matter how many theories or "supposed to be's" you 
throw at it.  The only performance you get is the performance you 
actually measure.

I'm nominally a Windows game developer.  I came to Chicken because it's 
performance-oriented.  I saw some proof of this in Shootout benchmark 
scores.  I was previously doing Bigloo, which on those tests was even 
more performance-oriented.  But I had problems getting a working Bigloo 
environment set up.  So I hopped to Chicken, thinking the grass would be 
greener, and had equal but different problems getting a working 
environment set up.  :-)  But I've solved some of those problems, like 
getting it to build reliably with MinGW, and I've established good 
working relationships along the way.  If I didn't have so much job 
pressure at present, I think I'd actually be productive.

I bounced around from language to language, implementation to 
implementation, for a long time before settling on Chicken.  I don't 
recommend the "butterfly" thing.  I have a smattering of knowledge about 
a lot of language, and no real production quality skills in any of 
them.  If something is working, keep going.  If it ain't broke, don't 
fix it.


Cheers,
Brandon Van Every