debug contest

[Hextic]

There should be a contest where a player has to fix some intentionally broken source code (which as absolutely shot with errors renging from the barey visible to blatant "connnekts();" calls. When the program is working, is connects to a server that raises a flag as that player being the winner. And it does not accept telnet client connections as legit, so it should work, and be kinda fun.

[TheCotMan]

There should be a contest where a player has to fix some intentionally broken source code (which as absolutely shot with errors renging from the barey visible to blatant "connnekts();" calls. When the program is working, is connects to a server that raises a flag as that player being the winner. And it does not accept telnet client connections as legit, so it should work, and be kinda fun.

That could be interesting. I like this idea.
:-)

You would need to make sure the programmer knows what it is supposed to do, but it sound like you thought of that.

[Hextic]

Yeah, there would probably be a huge /* comment block at the top of the source file explaining exactly what the program has to do in order to be considered to be working. It also gives some other information, such as that it is supposed to be connecting to (insert ip address here) on port (insert port here) with (insert either TCP or UDP here). That way, a person will hopefully notice that all the socket stuff is okay, the encryption of the message is good, but it is trying to send via a UDP port instead of TCP.

[bascule]

You're describing part of the Capture the Flag event, which has been done for several years. Here's a description:

http://www.cse.ogi.edu/~crispin/discex3_autonomix_defcon.pdf

And here's the page of the (former?) organizers of the event for the past few years:

http://www.ghettohackers.net/rootfu/

[Tierra]

On another note, I've seen games played before where someone will post up code to do a specific task, and from the time it's posted, until say 20 minutes later, whoever managed to take that peice of code, and either 1) make it into the smallest lines of code, or 2) made it run the most efficient (based on which was decided to be the winning factor before the code was posted) won the round.

There could be a coding competition that touches on Python, Perl, PHP, C, Haskell, ASM, Whitespace, Brainfuck (competition could get interesting with these last two), etc, etc. And maybe an ultimate round that combined a few languages to complete one task. (The used languages would need to be announced before hand so only those that know all of the above could compete).

[Vyrus]

This sounds like a fun game to me (and i am the most SUCKY coder imaginable) " and no i cant spell lol", plus it would be a good learning experiance for those wanting to know more about a spicific language... /me points to himself :P

[TheCotMan]

This sounds like a fun game to me (and i am the most SUCKY coder imaginable) " and no i cant spell lol", plus it would be a good learning experiance for those wanting to know more about a spicific language... /me points to himself :P

Also, if the API is described well enough, some may choose to replace a multi-file CPP or C project with a few lines of Perl. The Mythical Man-Month teaches us that there is often greater cost in supporting an existing project than there is in building the project to be supported.

[Voltage Spike]

Also, if the API is described well enough, some may choose to replace a multi-file CPP or C project with a few lines of Perl.


#include "unistd.h"

int main (int argc, char* argv[]) {
char* perlArgs[argc + 2];
perlArgs[0] = "myPerlProg";
for (int i = 0; i < argc; i++) {
perlArgs[i + 1] = argc[i];
}
perlArgs[argc + 1] = 0;
return execv ("/usr/bin/perl", perlArgs);
}