[MUD-Dev] Fw: 16K mud server competition !

Cynbe ru Taren cynbe at muq.org
Mon Apr 3 22:11:08 CEST 2000


"Chris Gray"     <cg at ami-cg.GraySage.Edmonton.AB.CA>:

| ["Jon A. Lambert" <jlsysinc at ix.netcom.com>:]
| 
| > >Erwin S. Andreasen is organizing a mud competition. The goal is to code
| > >a mud server in the language of your choice with the highest possible
| > >number
| > >of features, in *only* 16 Kb of source code + data files. This doesn't
| > >include the generated binaries and .o files. A few prizes are to be
| > >earned
| > >( such as programming books, a year of free mud hosting ).
| 
| Wow, that's not much space! I took a quick look at ToyMUD, to see if there
| was any chance of doing simple changes to shrink it down. However, it
| is 80K of code and 26K in the data file. I doubt I could shrink it to
| 16K without losing some major feature or two. The first step would be
| one of those code obfuscators that removes whitespace and makes all
| identifiers one or two characters.
| 
| 16K is enough for a decent MUD, I think, but the result isn't going to
| be anything you'd want to show to your mother! I think a more reasonable
| size count would be one of those program complexity measures, so that
| you could at least read the resulting source code.

Optimal solution is probably a language like J, Haskell or teco
that is more about conciseness than efficiency. :)

16K of J will take you a long way.  For example, check out:

   http://www.ai.mit.edu/extra/icfp-contest/isi.html

This is the 229 lines and 7078 characters (including comments and
development support "dead" code) J program that won honorable
mention in the ICFP '98 Functional Programming Contest for
beating out many compiled C programs many times longer, despite
being an interpreted language.

 Cynbe



_______________________________________________
MUD-Dev mailing list
MUD-Dev at kanga.nu
http://www.kanga.nu/lists/listinfo/mud-dev



More information about the mud-dev-archive mailing list