| User | Post |
Kyoufu Kawa
Posts: 1113/2481 |
Para, I use CA tech for maze solving. For each corridor cell, count the number of adjoining wall cells. If => 3, turn it into a wall. Keep going until it doesn't change any more. |
Zem
Posts: 654/1107 |
Originally posted by Parasyte
Well, that's how I view things when trying something completely new. I'm sure many programmers can wholeheartedly agree with this method.
Ooh... I do. My problem is while I'm thinking it through I'll get to step 3, then think "Ooh I have a great way of doing this" and start writing things with half a plan. Then when it turns out things need to be changed I never want to do it. |
Parasyte
Posts: 237/514 |
The best approach to writing any type of algorithm which you are unfamiliar with is just breaking it down into smaller portions and decide exactly how to do each one seperately. For example, if you were to design a 2D maze-solving (or 'path finding') algorithm, you would have to think along the lines of, "OK, I have a 2D maze with many walls and paths, but with only one proper path. The object is to get from Point-A to Point-B. So I must: 1) Start at Point-A. 2) Move in a straight line until I am given a choice of which direction to take. 3) Choose a direction, and record this decision. 4) Follow this path until another choice comes up. 5) Repeatedly make decisions until a dead-end or infinite loop are reached. 6) Backtrack to the last decision made and try the next direction. 7) Repeat until all possible choices have been made, or until Point-B is reached."
After you understand what your algorithm will do, you must decide how it will accomplish it. In this case, you could use recursion for each choice made, or keep a queued or stacked 'log book' of previous choices. It really does not matter HOW it works, just as long as it DOES work.
Well, that's how I view things when trying something completely new. I'm sure many programmers can wholeheartedly agree with this method. |
Ramsus
Posts: 17/162 |
I never really coded anything really cool as far as most programs go (I can't work on a project for more than a day or two it seems like), but I understand the frustration you're going through. Coming up with algorithms for game logic or models without any guide or code to learn from feels like wondering in the jungle all of the time. I put up with it for a while, then quit because I felt that I couldn't be a real programmer. Don't give up though, because you'll eventually get it.
Recently I've become interested in programming again, which is why I've quit programming temporarily to study concrete math and then computer algorithms first. I'm hoping to expand the way I think about problems in order to improve my ability to find solutions (in other words, write my own algorithms). I prefer the problem solving and modeling aspects of computer programming and math anyway.
If I were further along myself, I'd offer to help out, but in the meantime I hope you find the help you need. GameDev.net has a lot of articles, and you can find a lot of open source games to look at and study. Learning a few time and project management skills and sticking to them might help too, since then you'll be able to make deadlines and stick to projects longer. |
kitty hacker
Posts: 9/14 |
I call coding minion if you help me with graphics Sensei |
HyperLamer
Posts: 2959/8210 |
Originally posted by Zem
Comment the fuck out of your code.
That's always a good thing to do, but once I really get into something I tend to forget to do it, and end up having to come back and add them later.  |
Kyoufu Kawa
Posts: 1090/2481 |
Emil... I hate to post this where everyone can read it but I don't think HAM is a very good thing to use.
Real GBA coders avoid library intros. |
Ailure
Posts: 7277/11162 |
I think I try following this http://www.jharbour.com/gameboy/default.aspx tutorial, considering that the other tutorial didn't look as proffessional.
Heh, I wish I could get that as a book... |
Squash Monster
Posts: 508/677 |
I know how you feel, I have the engine, graphics engine, and the harder half of the level generator for my current game project sitting around on my hardrive and it hasn't been touched for at least a week. I kinda lost momentum when I started a recode and an attempt at making a better 3D graphics engine thingy.
I find the most difficult part to be dealing with gravity, making things not fall through objects when they go too fast and such. I typically store all coordinates in two variables, one for tiles and one for pixels off of that tile, becouse, along with capping the speed at one tile per second, it makes it possible to ignore a lot of the weird collision stuff.
Does any specific part give you trouble?
If it'd help, I could whip up a platformer engine pretty quickly if you're alright with Java (well, I guess I could put up with VB6).
|
Parasyte
Posts: 206/514 |
Oh! Open your eyes? Hehehe |
windwaker
Posts: 994/1797 |
Originally posted by Parasyte
Don't sweat it.
Even good programmers have trouble completing anything. Just look at me; (No intention of sounding like an elitist asshole with a huge ego) I am a damn good hacker and programmer [to which I will provide detailed evidence, if required for some very odd reason].
I don't believe you, prove it ;o. |
Ailure
Posts: 7272/11162 |
Zem: Well, that's something I should consider. Especially when I only use comments to disable bits of code... 
I did look through a few tutorials, apparently it looked easy enough for me and with some training... |
Kyoufu Kawa
Posts: 1082/2481 |
Spazzed the GameDev boards a bit. Found one DOS game that looked about right for platform collission detection, XRaiders.
Originally posted by Zem
Comment the fuck out of your code.
Guess what I did when I managed to write a replacement for SPrintF?
|
Zem
Posts: 563/1107 |
I have a suggestion.
I don't know if this is feasible with any manner of assembly, but I've found that if I'm working on something big, I will ALWAYS forget how it works. This is what they tell you to do in books and classes, and I always ignored it, but lately I've been trying it and it's helped a lot:
Comment the fuck out of your code.
Seriously. In this latest php toy I'm working on I have comments on practically every fucking line. It's made everything way, way easier. Maybe it's just me, but try it out. =( |
Kyoufu Kawa
Posts: 1079/2481 |
Well, thanks I guess.
So far I've found...
Lord Draga's platform demo
Looks reasonably helpful, even if it uses an 8x8 player on 8x8 tiles. Problem: it's only the binary.
Some guy's jumping demo
Utter crap, both in source and result. Also not at all what I needed, just a stickman that must jump to dodge a certain Jan's head. No actual platforms included and generally very poor and unusable.
I shall now return and search on...
Edit/Update: Following SlimeSmile's advice, I pinpointed some books on Amazon. I am currently downloading Game Architecture and Design - A New Edition from ye olde peertopeer network. |
Ailure
Posts: 7235/11162 |
And if you need help with GBA code, I could reach a hand.
Yes i'm competly clueless on it, but guess what type of proccessor I want to learn programming for?
...yes ARM. Besides I do want to try programming for another platform than PC. I know basic Java so I probably wouldn't have that big problem with C...
If there is anything I lack, it's gotta be artistic talent. If I did program something I would make so much sucky placeholder graphics it wouldn't be funny...
But consider that I have a few school projects right now, but I try to at least learn the basics during the weekends. |
HyperLamer
Posts: 2912/8210 |
You're not alone. I have tons of programs I never finished for all sorts of reasons from not being able to find something I need to just getting bored of it.
And BTW, copying code isn't really a bad thing as long as you understand what it does and how it works. |
Kyoufu Kawa
Posts: 1076/2481 |
Well, it's not motivation. It's more like raw frustration due to semi-constant failure.
At least when I made my first ASM hack, a new intro screen for Pokemon R/S, I managed to completely finish it well within 24 hours, not counting sleep/food.
For the record, I use ProDG, an official product. It seems to be mostly based on GCC.
Also, by lack of a functioning string manipulation library, I succesfully wrote my own number-to-string routine a couple days ago. So far, it only goes up to 255 but it works like a charm. This, however, is -not- related to game engines.
PS: Your name reminds me of dancing games. |
Parasyte
Posts: 183/514 |
Don't sweat it.
Even good programmers have trouble completing anything. Just look at me; (No intention of sounding like an elitist asshole with a huge ego) I am a damn good hacker and programmer [to which I will provide detailed evidence, if required for some very odd reason]. However, I find it very, very difficult to complete any project. Particularly, I've been unable to complete ANY game project, let alone get any of them to a point where they are actually playable. There are numerous reasons for this, which I won't go into detail about. The main problem comes from a general lack of motivation. If I work on a project too long (with regard to certain, specific exceptions) I will get bored with the project and throw it aside for another. That's just the way I am.
I am not sure exactly what it is that you are having trouble with, but it sounds more like a self-esteem issue (if you don't mind my saying so) rather than my ADD-like symptoms. For that I am afraid I cannot provide any advice. What I can provide, though, is minimal programming support if you really need it. But, I am a bit strict on what programming languages/compilers/development environments I use. If you stay away from languages like BASIC, Java, etc. and most any Microsoft compiler, I would be glad to provide assistance. I won't do a lot of the work for you, but maybe a bit. I'm generally interested in providing suggestions and such to get you moving. If you really want to program, it's best to just go for it, rather than beating around the bush. And if someone is able to push in the right ways, it really does help.
Yeah, that should be plenty of rambling. |
Kyoufu Kawa
Posts: 1075/2481 |
Originally posted by labmaster
If it's games with sources you're after, there's a heap at gbadev - http://www.gbadev.org/demos.php
I know that. But I've stolen enough code already. |
| This is a long thread. Click here to view it. |
