Happy New Year /prog/

What are your planning for 2014 /prog/?

I am planning to finally settle and get a Ruby or SysAdmin job.

Not becoming institutionized for another year.

To finish a large non-trivial project by myself.

>>2
We missed you, tdavis-kun.

>>3
This.
I've been trying to make an AI for a text game I made myself without knowing absolutely anything about the field. Using Markov chains is giving me astoundingly good results, but I wish I had the time to finish it.

>>4
Post sample code implementations pls. Dont care what language.

>>2

i lol'd

>>5
I'd rather not, it's a FIOC prototype and it's very specific to the game. The (not even close to complete) game is an online multiplayer ncurses adventure about living in a city where you can either get a job/start a small business like a normal member of society or shoot people and rob banks like a nigger. The AI is meant to simulate the actions of hundreds of citizens taking into account factors like the person's field of vision, the time they've been awake or how much money they've spent in the last few hours.

The prototype is SLOW AS BALLS but I was expecting it to be much worse, being FIOC and such. And it's not that impressive, it's just a bunch of polynomials pulled out of my ass that simulate each person's ``mind states'', hooked up to the Markov chain that decides if the person will execute some specific action.

I'm planning to rewrite the actual AI in C since the game is also written in C. I'll release it when at least some of the actual code is up and running, just like the DistBB guy did.

^{This is for you, Mentifex. (´･ω･`)}

>>7
If you want some basis for the maths around these types of things, the field of econometrics and intermediate microeconomics actually does provide equations for things such as chance of crime depending on factors like average pay, chance of getting caught (maybe a factor of size of policy force / security budget etc?), payoff for crime, and so on.

Sounds interesting, anyway.

>>7

ncurses games are garbage dude! srsly, dont waste your time with that.

>>9
Keep talking like a retard. That sure will get somebody to take you ``srsly''.

>>10

you are an india-nigger.

>>8
Thank you for the heads up. I suppose they don't work that well in the real world though.

>>9
If the AI works as intended, I don't care if I'm the only player in the world.

ncurses games are fun and flexible in many ways.

>>12
They work better than you would imagine, even in the real world.

>>8
I looked up a bit and found just some probability, linear functions and some logarithms. Are humans and the real world this boring? Or do I have too much fun coming up with random functions and refining them?

>>13
That's interesting, because math rarely works in the real world. (I know this is an oversimplification, I'm kidding.)
I'd wish you could fix this city's transportation issues with queueing theory but sadly things don't work that way.

>>14
Most econometrics is retroactive, so that's the simple linear equations you are seeing. You need to look at the forecasting and complex systems modeling side.

That gets more into things like autoregressive models, Gauss-Markov processes, OLS Asymptotics, multicollinearity, heteroscedasticity, etc.

The best thing to do is to look for some online public econometric data about a population that's close to what your city's initial state is like and mess around with the models that forecast its various states.

Markov chains work well at first but hit a limit fairly quickly, I suggest looking into something like ART-2 (Adaptive Resonance Theory 2) as a start for simple forecasting, or Hidden Markov Models which are more complicated but yield better results in most cases.

>>11
What's up with using something-nigger for all races that aren't white? Nigger is for those from Africa only.

I'm a caucasianigger.

>>17
You're from the Caucasus?

>>16
Shut up, nigger.

Making an ncurses file manager in Perl. (Does Perl even have ncurses bindings?)

It would be a good excuse for learning Perl and I'd rather not deal with the bullshit of Unicode libraries. I just want everything to be UTF-8 by default, because that's what my system uses and I have tons of files with weeb filenames.

... Perl handles UTF-8 out of the box, right? Right? ;_;

>>20
If it doesn't, I'm afraid I'll have to use FIOC 3. Or is there any other relevant programming language with ncurses bindings and UTF-8 by default, or at least a sane alternative that doesn't require millions of encode/decodes or len/utf8len?

>>20,21
Perl handles *curses through Curses (see CPAN), although there's probably a newer binding that leverages a NoSQL database to store your object-oriented frame states or something. As for unicode, it's just fine, but I'd give perldoc unitut and perldoc uniintro a read before doing anything serious.

>>20
Perl does handle utf 8 out of the box.