This is awesome. I used SimpleCV in Python to detect the walls, much like you did. But I used it to transform the screen to make the platforms fall vertically:
For anyone interested in alternate bot approaches:
While the author made this "pixel bot" to pair their Super Hexagon with ambitions to learn OpenCV, pixel bots are almost always inferior to memory-based bots. As mentioned elsewhere in these comments, code injection should use caves and jumps, not overwrite existing code.
With a memory-based bot you often do not even need to write to the game's memory. You can use the ReadProcessMemory API call for data gathering and perform inputs with SendInput (emulate keyboard input with LLKHF_INJECTED set) or PostMessage (no LLKHF_INJECTED, also works in backgrounded windows).
The disadvantage to memory bots is that they're often more complicated at scale (beyond the point where a pixel bot would likely fail entirely), particularly with games which are updated. In cases where a game client is updated, signatures similar to virus definitions can be provided to find the particular memory address of a sought value regardless of version. By scanning the application's source code for a unique bytecode pattern known to contain the target address, you can then extract the address and read the value. This is commonly known as signature or pattern scanning.
There is another variant of bots for client/server programs where a "headless" (no display) client can interact with the server via network communication alone. Not as much fun to watch but also a very effective way of interacting with games, though it does not apply to standalone clients like this one.
I don't have any resources in particular but I've personally made them for years. In general you read memory and take action (via keyboard input) based upon what you read from memory.
For example in a game where you fish, you would wait until a fish is on hook by watching the representative value in memory and pressing a key to catch a fish when that value indicates you have a fish on-hook.
In Super Hexagon you would watch the values indicating where walls were placed, which is likely an array of enumerable values indicating a shape (each shape has at least one opening). So a 1 might indicate a wall only at north, a 3 might indicate every odd side of the hexagon is a wall which must be avoided, etc.
Finding values is obviously easier when the value is displayed on-screen, like ammo in first-person shooters. It takes some time to become familiar with exploring memory like this and you certainly wouldn't want to write your own memory scanning primitives to do it, start with a tool like Cheat Engine (http://www.cheatengine.org/), ArtMoney (http://www.artmoney.ru/), or one of the many other memory editors available (https://www.google.com/search?q=memory+editor).
If you'd like more details please be specific about what you'd like explained!
Ever since my 8-year old son beat the game multiple times on level 6, I have been working on getting there as well. The best I have made has been 19 seconds on level 6 - though I have beat the first five levels, level 6 is truly insane.
Watching my son play is quite the experience. It seems my reflexes are not what they once were. It's an amazing game.
I want to think that writing my own bot would count as beating the game, but I doubt my son would say it counts. So, here's to the next few thousand attempts at level 6. "Again!"
What's this "hard mode" you speak of? As far as I know, there are only 6 levels, the hardest being "hexagonest" in hyper mode, which is the "hardestestest" level. I'm playing it on PC via steam.
I've beaten that but I'm keen for a harder challenge!
I would assume by "hard mode" they meant the last 3 difficulties you unlock by beating the first 3. And the challenge at that point is chasing even higher times and doing leaderboard chasing (though good luck cracking the top 200 at this point, my 250+ is only 317th last I looked).
i saw a 6 year old play Temple Run on an iPad and i was amazed! Someone jokingly said that he doesn't trust kids; they pretend not to know how to speak, then they make transactions and build empires on their iPad.
For anyone who hasn't tried this game, it's pretty awesome! I was one of the top players on one of the level back then (http://www.twitch.tv/david_u/c/1985844).
I guess OP could have just done pattern recognition since there are not so many patterns in the end.
Also I'm impressed, here: https://www.youtube.com/watch?v=OyVcokAUj3E at 0:14 you can see that the screen is suddenly rotating. This is a super hard moment to pass and I can see that as a very disturbing thing for a bot as well.
(I initially posted this on the Reddit thread yesterday, but I thought I'd repeat it here since there are many programmer types who might attempt this here.)
From the article:
> Of course, because we replaced the function with our own, we have to actually swap the buffers by explicitly calling the original glutSwapBuffer() function. I did this by undoing the hook, calling the original function and hooking it again.
Background qualification: I do this sort of debug launcher + DLL injection thing a lot. For one example with source, see my tool header magic ( http://www.romhacking.net/utilities/796/ )
As someone with experience, please take it from me: this is a really bad idea. If you rewrite functions to call the originals, it becomes possible for another thread to call that function in the middle of your rewrite, and end up executing illegal instructions. Not only is this theoretically possible, it routinely happens all the time. Even when you think for sure that your program is not multi-threaded. Even something like hooking a Windows file size function, you'll learn that the open file dialog on Windows Vista+ is threaded, and will bomb out on you with alarming frequency.
The proper way to do this is to create "trampolines" instead. The idea is that you copy the first few instructions from the original routine, and put them somewhere else. And at the end, you jump back into the original function after your patched jump. The address of where you put the original instructions is your pointer to the "original function" that you can use if you want.
So basically, something like this:
originalFunction:
opcode1
opcode2
opcode3
Becomes:
originalFunction:
jmp detourFunction
nop <padding>
resumeAddress:
opcode3
trampolineFunction:
opcode1
opcode2
jmp resumeAddress
detourFunction:
<do your function hooking magic here; modify arguments; whatever>
call trampolineFunction <only if you want to>
<do whatever you want with the results from the original function call>
Now, I know this sounds horrifyingly complicated: you probably think you'll need a full-on x86/amd64 disassembler+assembler. And it's theoretically possible to make a near-impossible function to trampoline. But keep in mind that 99% of APIs are built in higher level languages, and as such, have very consistent function prologues (usually saving the stack frame.) So in practice, you can write trampoline installers using one or two "signatures" and hook nearly every function you'd ever want to.
Hmm, I'm a bit surprised that some sort of ad-hoc AI is used, instead of extracting the topology of the level and running a pathfinding algorithm on it.
Once you remove all the human-confusing effects, the levels are actually very simple, being just 4-6 columns. Since the minimum thickness of obstacles is quite high and they are aligned with each other, you could even represent the level as a graph, and run your pathfinding without having to manage the geometry explicitly.
But then again, I'm the guy who thought about implementing a bot when the game came out, instead of actually doing it.
I attempted the same thing[1] but between the fact that I had no idea what I was doing and having to screencap every frame (instead of hooking directly into the game's buffer), it didn't work too well. I think at best it tended to last about 10 seconds on level 1.
Not having managed to get past 30 sec on level 6, I applaud this bot. Using DLL injection was a very good idea too, but I wish the algorithm could play the post-finish level for a bit longer.
https://github.com/shaunlebron/super-hexagon-unwrapper