The git repository is listed on The UCC git page as "progcomp2012.git"
Direct Link Here
We will use the same mailing list as last year (progcomp).
There is a #progcomp irc channel on the ucc irc server (irc.ucc.asn.au) where you can ask questions or help with setting things up.
I am in the process of learning how to set up a VM for this competition. Please be patient.
This site explains what Stratego is.
Short Version: It is a board game in which all pieces have a value that is, initially, unknown to the opponent player. The objective is to destroy all enemy pieces, or capture the enemy "Flag". Pieces with higher values destroy pieces with lower values. There are several special pieces/rules.
Create an AI to play Stratego.
Programs are written independently and interface through stdin/stdout with a manager program, which queries them on setup and moves.
The manager program provides the protocol for two seperate AI to play a game of stratego. It has the imaginative name of 'stratego', but I will probably refer to it as 'the manager program' or 'stratego' with absolutely no consistency.
It also aims to assist with AI design by providing options for graphical or terminal output and saving/reading games from files
Human players are also supported, although the interface is minimal, as this feature is meant for testing.
If you just want to play a game, without having to write your own AI, try Probe
For the sake of simplicity and keeping things in one place, the protocol is now entirely described in the manual page of the manager program. All updates to the protocol will be reflected in that file.
Warning: The accuracy of the above file depends on how recently I pulled it from git. To ensure you get the latest version, find it under "web/doc/manager_manual.txt" in the git repository
Another Warning: AI programs must unbuffer stdin and stdout themselves. This is explained in the manual page, but I figured no one would read it. It is fairly simple to unbuffer stdin/stdout in C/C++ and python, I have not investigated other languages yet.
WARNING: Work in progress
It is currently planned to store all the AIs on a virtual machine, and periodically run a script to play a round robin
The scores for each round and the total scores will be recorded from the start to the end of the competition.
The competition will run over a period of weeks (depending on general enthusiasm...), and competitors will be able to alter their programs except during the periods in which the script is running.
The following categories will be used when determining the final winners:
WARNING: Work in progress
The following sample programs are currently available (and in a working state - refer to the git repository):
|asmodeus||Python||Scored||Path finding, known combat results, piece values|
It is planned to implement the equivelants of these samples in C++ and Python at least, possibly other languages later.
We (I?) are now accepting test submissions.
You must submit the full source code, and build instructions or makefile(s) for your program.
Also include the name of the executable or script, the name of the AI, your name, and optionally a description of your AI and its tactics.
Please email [email protected] if you have a submission.
The competition will run in 2012. The exact dates have not been determined. I will consider whether it is worth waiting until freshers have a chance to make submissions March/April, or running the competition earlier in January/February.
If you want to help set up the competition, email [email protected]
There isn't much left to do, but all help is appreciated. Maybe you can suggest something, and volunteer to implement it!
The most useful thing you can do is actually write an AI, and test the manager program. Please report any bugs (in the manager program, not in your AI!) to [email protected] (you can feel free to fix them, but won't be able to push unless you email matches).
Last webpage update: 20/12/11