ICFP 2003 programming contest on 06/28

[Ninja40]

Hello everyone,

For those of you who have nothing to do next week-end and feel they are up to the task, the ICFP 2003 programming competition starts soon.

It's a team competition and the problem is usually quite stimulating (a raytracer or the IA of a competing robot for example, see the previous problems). 1 member teams are also accepted. You have 72 hours to complete the task, in your language(s) of choice.

Official announcement :

"There are just ten days to go to the sixth ICFP Programming Contest!
Starting at midnight on Saturday morning, June 28th, you can prove once again that your favourite programming language is "the programming tool of choice for discriminating hackers!"
Don't miss this opportunity -- check out the contest web site at

http://icfpcontest.org

for details -- and the tiniest little hint of this year's problem.

John Hughes"

[sicarius]

Seems like an interesting contest. Now that I know about it I may take a more serious look next year.

[Strike]

jemfinch told me that he's formed a team for this, and I think they are going to use Common Lisp for this (more for the learning experience than the expectation of winning, from what I heard from him). The #haskell channel on freenode is fielding a team as well, and I've offered myself up as voluntary "cheerleader" since I'm too green with Haskell to really help much. Hopefully this'll help that a lot, though.

I like the ICFP contests a lot. They are good, simple tasks that often breed creative solutions that can be done well in a number of languages. Besides, my all-time favorite language (Python) and my recent favorite language (Haskell) have done fairly well in the past two years :) Last year, a Python program won the Lightning Round (and hence, the Judge's Prize, though not the overall contest win), and two years ago, a Haskell team won the whole thing.

[GnuVince]

06/28? Too bad... party at a friend's of mine with Ismahel et al. We're gonna be WAY drunk, so no contest for me (not that I could do it anyways...)

I am putting my money on O'Caml. It seems like this language was written for the purpose of winning those contests (if you check the past years, O'Caml totally dominated)

[Ninja40]

This year, the problem is nice and simple.

Given a car that can turn L., R. accelerate or slow down, and a set of tracks, find the optimal trajectories for the car (or at least the trajectories closest to the optimal solution). It's an optimization problem, and certainly, some contestants will use AI techniques.