Finishing is hard.

[Cero]

programming java games in jail

dfficulty: nightmare

[theagentd]

@dezzroy

theagentd,

I don’t mean to hijack this thread, but, to elaborate, when you’re in jail, your mind is your only playground.

I wasn’t in jail for computer-related crime, just possession of a laughably small amount of a derivative of a certain popular plant.

We were not allowed to keep anything written while incarcerated after release because the authorities believed we would create and maintain criminal contacts with former cellmates. I had no access to a computer, but used paper and pen to map out rendering algorithms and geometry processing procedures. I had so much time, I even debugged them in my head.

Seriously, look at all of us. Jail, parenthood, employment, poverty, social pressure to be mere consumers, math ineptitude… we persevere!

It’s what we do! 8)

Jeez. IRL is so hard.

@Cero

programming java games in jail

dfficulty: nightmare

I’d say Hardcore Nightmare, because you don’t respawn when you die.

[Antiharpist]

[quote]This is all that matters. Why are you so apologetic? Being bull-headed and driven is the only way to success in this adventure. You’re enjoying what you do, but too concerned with the results. You said it yourself–you enjoy working on this stuff.
[/quote]
Hell yeah dezzroy, that clicks with me. I’m just apologetic because I’ve been out of the internet social scene forever, and don’t want to come off as a dick head. Anything I post isn’t bragging, cause it’s still hot on the plate stuff that I enjoy sharing. I just don’t want to be too egocentric. Staying hungry is a good thing.

If I were locked up, I’m sure I’d doodle out ideas too. I mean I’m not at all intelligent, but it only takes a second to realize a square has 4 sides, and after drawing it out on paper, you realize that you can set the velocity of your 2D rolling objects by something like the following:

//considering animation_action contains frames for a complete rotation of object
double frames_per_pixel = animation_action.getWidth()*4;
frames_per_pixel/=((double)animation_action.getFrameCount());

I don’t know if that’s mathematically correct. I just drew shit on paper and tried it out. It worked.

It’s those little Eureka moments that make me happy and want to share. I do like setting strict timelines, and cranking out stuff, even if it will end up on the cutting room floor. I don’t care if I something I do is the end-all of greatness, but I don’t want to feel embarrassed about how it represents my dreams. In the end, you do it for yourself, and if you aren’t content, nobody will be.

I’ve noticed that sometimes grunt-work is necessary too. It’s not all fun and inspiration. The grinding and leveling up is an awesome experience though. It’s like the end of the World of Warcraft South Park episode, after it’s done, “so what are we going to do now?” “Whattaya mean? We can finally play the game.”

Rich

[Antiharpist]

Today I tried to add Separating Axis Theorem to the collision detection routines, and I failed. It sorta worked, but there were bugs all over, like being able to walk over thin air, and with a rotation of zero degrees not being able to move at all anywhere. If I rotated things only a few degrees, it worked a bit, but it wasn’t tight. I decided to throw in the towel on it for this project. We have so many elements working well, and I just can’t bring myself to tear shit down for something that if I’d maybe had a math class or two wouldn’t have been too tough.

Having your ass kicked is a good thing though. You have to realize when to withdraw from the current battle, knowing it’s not a total loss.

I’m gonna go crack a beer and listen to some Uli Jon Roth on my first day of my week off.

Metal,

Rich

[Antiharpist]

I spent a day, 8am until 9pm, working on the the rotated collision checking yesterday, and I failed. Today, went back on it, 8am, untl 2PM, and then it worked. I was just dirt dumb and rotate one rectangle back to being flat, then the other by the same amount in negative around the axis of the first rectangle. I debugged like hell. I wrote a debug canvas to see shit visually, but still no go. I finally decided to write a separate program to just test, and it didn’t take long to get the routine working well. After that, I made I copy of my main collision detection class, backed up the old one, and then stripped shit down. It turns out it was just some collision escape convenience routines that were killing it. Then, all of the shit all of a suddenly worked great.

Persistence is an alternative to intelligence,

Rich

[theagentd]

I almost always do a small test and then slowly try to integrate it into the rest of the game, especially when working with OpenGL as it tends to just give you a black screen when a single line somewhere is wrong.

[princec]

@Antiharpist

Persistence is an alternative to intelligence,

Amen to that 8)

Cas

[kilvati]

For complex stuff like path-finding and collision detection I first write a bunch of unit tests before I implement the code.

[zendar]

@Antiharpist

After that, I made I copy of my main collision detection class, backed up the old one, and then stripped shit down.

This is a bit unrelated to the thread topic, but here goes:

Sounds like you might not be using version control software. If you are not, I really recommend using some kind of version control. It makes experimenting and messing around with stuff a lot less worrying. If something doesn’t work out, you can always revert to a former version that works.