I'm going to go on a rant.

[Tesselode]

I've been doing on and off programming endeavors for like 5 years now. And I still can't make a platformer.

A PLATFORMER.

WHY IS IT SO HARD.

There's always something shaky and wrong about my platformer engines, and those always cause some nasty problem. Then I delete them because they're not working, which is kind of stupid because now I'll never know how to fix the really odd glitches that arose. Why are platformers so finicky AARARGHGHGHG

I'm posting in support and development because I need moral support. Somebody help me. *sob*

[veethree]

http://www.youtube.com/watch?v=7uUlOAyQsn4

[Jasoco]

I've been coding on and off for 20 years and only fairly recently was able to create a buggy-ass platformer. They're pretty simple but also difficult. You'll get it eventually. Just don't stoop to using Box2D if you want realistic physics. Best to code your own for something like Mario or Sonic.

[master both]

A platformer isn't the most difficult part, is the freaking collision, i still can't manage to resolve diagonal collisions.

[adnzzzzZ]

You need to figure out what problems you have when creating platformers and actively try to fix them. Coding something over and over isn't going to be any better if you keep repeating the same mistakes. I've found that using box2d is super simple once you get used it (so weird bugs aren't that common since it usually handles those for you (on the collision side, at least, which seems to be big source of bugs)).

[Kingdaro]

Using box2d (or love.physics) for platforming purposes is a little clunky. I'd prefer to use smaller, general-purpose libraries like HardonCollider or bump.lua.

[Robin]

My Advice For The Soul is: make it smaller.

Every time you find yourself out of your depth, or getting completely stuck, make the scope of your project smaller. Make things easier for yourself. And continue doing that, until you actually finish a game. Only then start thinking about making the scope of your next project larger again.

We have a tendency to want too much, too soon. I know I do. Even after 10 years of programming (I know, not as long as Jasoco, but he's, like, old!), I still have that tendency.

[kikito]

The question is psychological. Why are platformers so comples, if they look so simple?

The answer is right there: they look very simple. That doesn't mean that they are simple.

Platformers depict stuff that comes to us by instict; we don't have to think about it: Run. Jump. Fall. Crouch. We have those since we are born.

The problem that you can't program a computer only with your instincts. You must dig deeper. You must understand some Very Abstract and Very Weird and Unintuitive Things (like "There are infinite numbers between 0 and 1" and "Time is continuous, but our machines work in discrete intervals"). And then express what your instincts tell you in base of those abstract terms. Your mammal instincts can't help you there, because it's a different problem than simply jumping (or watching someone jump). My cat knows how to jump and run, but can't program a game.

In other words: it looks easier than it is, because the recognition part is "hardware accelerated", but unfortunately we are not born with "programming instincts" or a "logic GPU". We must do it "the hard way", and learn stuff.

The good news is that brains are very versatile. If you practice enough, you can teach your brain to do almost anything; even programming platformers. Once you have programmed 10 or 20, it will feel as natural as jumping or running

[veethree]

I'm still stuck at the collision. If i used a library for the collision i'd have made like 12 platformers by now probably. But i just prefer trying to figure things out for myself.

[T-Bone]

My best advice is to not code alone. Show your code to other programmers. It doesn't matter much if it's here on the forums or friends irl. Discuss the problems you face and how you can solve them.

Writing a platformer is tricky, but don't give up! It's definitely possible even for a novice programmer.