Hacker News new | past | comments | ask | show | jobs | submit login

I don't know how to say this without sounding like a troll, so I'll just come out and say it: the people who have responded to your question so far sound like they know a lot about Lisp or functional programming and rather little about games.

You came to HackerNews, where all the functional programming boosters hang out (myself included) and asked: can I use functional languages for X? What, were you expecting an unbiased answer? To HackerNews, functional programming can do X optimally, forall X.

All languages have tradeoffs, though it sometimes seems like we have yet to face up to the tradeoffs in functional programming. But in truth, functional programming kinda sucks for games, and the more functional it gets (that is, as it approaches Haskell), the more it sucks. Because game logic is all about the one thing that pure functional programming disdains with all its heart: global internal state and side effects. The purer you get the more you have to jump through hoops to manage this kind of stuff. So IMHO you're really asking: can I learn a functional programming language by coding something which functional programming is fairly hostile? And people here are responding: sure, you'll love it! It's perfect for you!

Also note that real game programming is also often about speed and soft real-time guarantees. Sure, games often have a high-level scripting language in which certain gameplay elements are written. Lua, for example, is one popular choice. Variations on Javascript are another. But if you're looking to write a game in a programming language, and need to eek out that extra bit of cycles to do something nifty, most of the languages here are bad bad choices. This is probably not something you care about since you're just looking to learn, which I applaud. But it does tell me that boosterism is in effect when people recommend Lua and Clojure and ABCL and SISC for game programming that they've not considered, and they don't bother to tell you that these languages are quite slow, for an environment (games) where speed really matters.

So if you're trying to learn a functional language by doing a game, I'd pick the least hostile functional language that has highly portable graphics and event support and which is decently fast. Ordinarily I'd say CommonLisp, but ABCL's not there yet. One language in this category is Kawa, a JVM scheme which can be optimized to run at a decent fraction of Java. PLT might be another choice: it's not super fast but has very strong and portable libraries.




I agree with you ("Hey self-selecting community of lispers: Is Lisp a good language for X?" is going to get a predictable reply), but half of the original question was about learning functional programming. Separating the two halves of the question would have probably got a more useful response.

Scheme is a good language for learning functional programming. (Haskell is as well, but I would argue that several techniques the poster is likely to learn from Haskell are difficult to use in non-lazy languages, and telling them apart requires experience they probably do not possess. Also, Haskell is very poorly suited to games, and (* dodges rocks* ) seems to me like a less practical language overall .)

Also: Lua is fast (certainly fast enough for prototyping), and has a clear migration path to C / C++ for libraries and the parts of the game that need the raw speed. Something on the JVM (including Java, of course) may better suit the author, since they are probably already familiar with many of the relevant libraries, but I don't have much experience there.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: