[DISC] Planning? How do you solve problems and wirte your apps and games?

[hooligan2001]

Does anyone do any planning on their game? Do any one use VTOC diagrams, sub problem table's and Algorithms?

When i first started to code in flash i developed a bad habbit of just skip planning and just code my app's and game's. And now that i got taught it was bad to do that and taught and the correct way to go about it the quality of my work and how nice it runs has improved a great deal. At first when i was getting taught about correct planning i just thought it's just extra work that you don't need. But when you start to do the stages in planning and testing your algorithm you will find that all you will need to do is convert you algorithm into code it will cancel all logical errors if checked properly. And make the code go from noodle code to nice structured code with less variables and no logical errors.

What i do to solve game problems is, work out sub problems and then define my inputs and what i want and then work out how i am going to use those inputs and get my outputs from them. I then use this written diagram to write my algorithm. Test that by using test data to find logical errors and then simply code. It sound's like alot of work but it will save you time. Because noodle code is hard to fix.

So i was just wondering how everyone goes about it. Go straight to coding or do the planning.

[DancingOctopus]

I usually plan most of it in my head before I start. Occasionally I'll write a few notes down, but generally.. It's all up there. /\/\

[pellepiano]

Im hopeless. I get so carried away when I get a game idea, that I just go for it. Im also afraid that seeing all the work in front of me, at the start, will make me not start the project at all.

[ihoss.com]

I make weird drawings on pieces of paper that nobody understands, and they are so simplified that when I try to turn them into functions I give up.

Most of my planning is done in bed. I think better when I lie down and stare at teh ceiling, so I got myself a beanbag that I could just sit in and plan. Then when I have an idea I write it down so I don't forget it. Thats about it.

[Kakihara]

i do much the same as ihoss and tonypa, i find i lose hours of sleep because i am still in code mode when i get into bed, i usually do some planning, i open notepad and jot down the basic premise of the game and how you score and die, then i write most of it out in psuedo-code then just go at it from there, i got to admit i dont really look at the planning documents when building the game apart from using it to remember ideas for power ups and enemies because i store most of the ideas in my head.

i suppose i use the planning stuff so i have something to show prospective employers to show them how i work.

[DancingOctopus]

Originally posted by ihoss.com
I make weird drawings on pieces of paper that nobody understands, and they are so simplified that when I try to turn them into functions I give up.

Most of my planning is done in bed. I think better when I lie down and stare at teh ceiling, so I got myself a beanbag that I could just sit in and plan. Then when I have an idea I write it down so I don't forget it. Thats about it.

Oh man, I simply can't think straight like that. I'm a walker. I have to be moving around, seeing different things, I think if I focus on anything too long I'll get distracted (ya ya, ADHD RITALIN KID ) I once took a 5km walk, came up with a solution, turned around and walked back. By the time I sat down again at the computer, I'd forgotten what my marvelous revelation was (luckily I have a good memory so it eventually came back to me .) I took up pacing out on the veranda after that.

One thing I almost never do is plan code. I'm probably at a stage where I know pretty much what flash is capable of, even if I don't know exactly how I can always research later. Most of my planning goes into the logic and structure of the game/app.

[hooligan2001]

Yeah tell me about that laying in bed thing. Ill be at the computer for hours trying to work something out or the best way to do something. I end up giving up and laying in my bed then start to fall asleep the first thing that comes is the answer. So i always keep a jotter pad next to my bed to jot them down so when i wake up i can fix the prob. I was told that you have more control over your brain when you are asleep because it doesn't have to send message's to most of your body sense's and intructions. Like a pc. One app open its ok 10 apps open and it wont save Or something like that.

[WilloughbyJackson]

I'm a preplanning fiend...

First, I can think of an idea anytime, anywhere, so I usually have a little notebook with me. The idea usually has to do with the rules for the game, or the character designs, or the plotline but usually not a combination of them.

Second, I start to draw out characters, write the details for the rules of the game. Winning condition and losing conditions, objectives, how it gets more difficult, and usually a bit of code theory. Sometimes I then program a prototype game.

Then I put it in a folder and file it away until I get time to develop it.

The End...

-pXw

[yakov]

i've been taught planning. alot. if your project is greater in terms of time or team, than what most coffeebreak flash games are - we've been taught it most definitely need planning. And even if you're just making a tetris, it's a damn good idea to do a designdocument. i'm always spending alot of time writing down ideas into structured documents before trying to get a game into art and function. I usually end up with quite a lot of documents, so when something finally gets to development I've got quite a wide range of ideas to build from.

[Squize]

I must be odd, 'cause I never plan.

The most I'll do is spend 30 mins with an a4 pad making some notes, but in saying that I'll usually get bored and just go for a fag half way through.

For the sN stuff, I don't even do that much. Never have.

As to the walking thing, it's been proved that that's one of the best ways to get inspiration, your brain solves the problem in the "background".
Also lying in bed is a sure fire winner, as just before sleep you hit that theta state which aids creativity ( I know you're all going to think I'm a smelly new ager, but listening to theta encoded music chills you out a treat and gets the ideas flowing ).

Squize.

[flashMasterHand]

Well, I guess I could say I am a combination of the two extremes.

I usually try to plan out everything down to the smallest detail, and sometimes, I just end of frustrated and give up. But many times, I try to plan it out, I and I can never think of what will be needed. I always end up missing a function here, or a variable there. So what I usually end up doing is some basic outlining, most of the game idea is in my head, but sort out the major tasks. Then I just start coding somewhere, usually at the heart of the project, the main functions. Once I am done with that, I look through my code. And I suddenly realize, oh, I need to write that function out, and this variable needs to be initialized, etc. So I write a sort of To Do list out, and go from there.

I am also one of those bed people. I sometimes come up with the greatest ideas in bed, if I don't fall alseep first .

Interesting topic, by the way.

Dave

[swcAndrew]

I am the kind of person who always wants to plan it out but become too excitied and just jumps right into the code. I feel if I try and get into much detail I wouldn't be able to finish the project. But when those times come and you feel you have got a great idea, forget it I won't even be able to sleep because I will be so damn awake coming up with an idea everytime I close my eyes. Although sometimes so I make sure I will finish project, everyday I will make a to-do list with all the things I would like to finish during that day.

[metanet]

i think that the term "game programming" spans such a vast multitude of tasks that when people give answers to questions like this one, the differing answers usually reflect different interpretations of the question..

concerning the stuff the OP was suggesting, that sounds to me like pretty hardcore compsci stuff -- proving that your code is correct works alright for small tiny things, but when you have a complicated realtime system it gets very hard, if not impossible, to formally prove that things are "correct".

having said that i'm sure that if i at least tried a bit harder in that direction i'd spend less time tearing my hair out.


in terms of planning, i don't beleive anyone who says that they don't plan. the only time you can just sit down and crank out a game is if you've already made a very similar one, so that hard problems of design and breaking the game down into solveable tasks is premade and sitting in your head.

otherwise.. well, you're some type of god and i'm jealous of your crazy skills (squize...)

for stuff like the game as a whole, i TRY to plan it on paper, just roughly decompose the game into a few components that are each responsible for different htings, but at some point you have to learn by trial and error -- come up with a design, implement it, and then change is as necessary. this way your next design will be better.

john carmack could never have made, say, q3a without the experience of making wolfenstein, doom, quake, etc.. if you look at the design for those games you can see that they're essentially all analogous, each new one is a refinement of the old one that improves incrementally. i think this incremental approach is the best model to follow.

for stuff like geometry, etc.. i can't see how i'd get by without good old graph paper

the book "code complete" has some really interesting information about this aspect of programming and i'd urge the OP to track down a copy, it's a great book.

raigan

[Squize]

"i'm jealous of your crazy skills"

Ha, if only

The games I do for work seem to be pretty straight forward, but with only one ( Usually insanely ) hard aspect, and by fluke more than skill it just seems to fall into place ( Don't get me wrong, nearly every game has given me that "Christ, I can't actually do this, I'm screwed!" moment, but every game has to be done so there's not a lot I can do about it aside from sit down and get it working ).

The personal stuff is always pretty straight forward too, I don't have a N or a mario kart engine in me, just some nice sprites moving about.
Plus everything I make for myself is a remake of an 8-bit or arcade game, so game play is something I don't have to worry about too much, I already have a feel for how I want it to play, then it's just a case of tweaking it til I like it.

The one nod of the hat to planning I guess is making a "To Do" list, but that's really just near the end and is more a case of making sure nothing falls off the radar rather than anything drastic.

Squize.

[token 3]

I'm with Squize on this, I never plan anything before hand, I'll sit, code a little, if there’s something that I’m not 100 percent clear on (not very often; I'm not trying to be big-headed, I'll usually pre-planning in my head as I go), I'll stick a comment in and program around it, an solve it later

However, figuring problems out when lying in bed is incredibly frustrating, once I have worked out how to do it, I can't just jot it down and code it the next day, I have to get up and do it there and then, or I never get to sleep

[Psyfish]

I usually draw out my maps and characters on paper. If I make it tile based I like to use graph paper. I also plan games in my head while I take walks or when I go to bed. I don't write out the code because I'm new at flash so I would make a lot of mistakes. Sometimes I will write one or two lines of code beside a map that I will use but not much more than that.

[Incrue]

Originally posted by metanet

the book "code complete" has some really interesting information about this aspect of programming and i'd urge the OP to track down a copy, it's a great book.

raigan

And were can we find this book?
I also would like to know, from those who plan things, How to keep your mind organized