[plt-scheme] Experience Using Mz(Dr)Scheme for Numerical Work

From: Will M. Farr (farr at MIT.EDU)
Date: Mon Apr 18 02:04:20 EDT 2005

> Why not use mzscheme for incremental development and bigloo for the 
> long computations?
> I've got a little simpleminded utility called 'staste' that can be 
> used to tailor code for specific schemes.
> Best regards,
> Hans

I have considered this, but I'd prefer not to have to transfer my code 
between systems before conducting a serious run.  Once you have a good 
stellar dynamics code (a n-body integrator), there's a lot of stuff you 
can do with it, but most of this involves patching other physics codes 
into it---stellar evolution (some of the stars explode now, rather than 
just moving according to Newton's Law) is one example.  I would like to 
be able to experiment with some real physics without having to 
continuously transfer the code from one scheme system to another, 
recompile, etc.  (And the transfer would probably be non-trivial, 
because I would be exercising the object systems of both, the typing 
system of bigloo, the FFI in both, etc....)  I'm sure that such an 
approach would work for many people, however, and it is something I 
will seriously consider.  If I do take it up, I may be back asking 
about staste....


Posted on the users mailing list.