Source code for Doom 3

[Mickelukas]

Hi all,

If you're interested in an OpenGL game that used to be very nice looking compared to other games the source code for Doom 3 has now been made public under GPL
https://github.com/TTimo/doom3.gpl

[Mickelukas]

Double post, please remove.

[Cero]

yes - as announced by Carmack @ this years QuakeCon
I understand he had to rewrite some code due to stupid licensing problems, but here it is
I like it - although, its not like I can read it and understand it or stuff =D

Games published by our own members! Check 'em out!

[Mickelukas]

He had to add four lines and change two lines because of a patent of Creative...

Mike

[ReBirth]

I havent been understanding opengl...

but this is one of my fav game and FPS that freaked me out for first time!

[gimbal]

I'm always in awe of people that can take Carmack's code and actually make it do something else. There is just so much code in there (even the Wolfenstein3D code was huge), how would you even begin to comprehend it all?

[lastaid]

Doom3 was the first game that genuinely scared me. I remember what people did to the id tech 1 - 3 after it was open sourced
http://bytonic.de/html/jake2.html
http://sourceforge.net/projects/xreal/

would be intersting what happens to doom 3, although i don't think i will come to java anytime soon ^^

[Cero]

I'm always in awe of people that can take Carmack's code and actually make it do something else. There is just so much code in there (even the Wolfenstein3D code was huge), how would you even begin to comprehend it all?

yeah try going back to even your own code after a year or so
its hard

[gimbal]

I'm always in awe of people that can take Carmack's code and actually make it do something else. There is just so much code in there (even the Wolfenstein3D code was huge), how would you even begin to comprehend it all?

yeah try going back to even your own code after a year or so
its hard

Especially if you're still in that "less than 10 years Java programming experience" phase where after a year you go back to code that horrifies you

[sproingie]

I don't even like C++, and avoid it wherever possible, but I find Carmack's code to be incredibly readable.  The absence of idioms like virtual inheritance or template metaprogramming leads to some very straightforward code, and quite well-commented code at that.  You want giant code hairballs, go read the source to libstdc++ sometime.

This isn't to say that I don't like the existence of clever idioms in languages (my language of choice these days is Scala, which is stuffed top to bottom with clever tricks), it's just that I appreciate them most when they make things clearer in the end.

Games published by our own members! Check 'em out!

[impaler]

What better way to learn than by reading the source code of a finished product?

[gouessej]

What better way to learn than by reading the source code of a finished product?

Making a finished product.

[princec]

Hehe that's ironic coming from you Julien 

Cas

[ReBirth]

What better way to learn than by reading the source code of a finished product?

try to make one by yourself and fix every errors. When errors are out, try something naughty

[ra4king]

Hehe that's ironic coming from you Julien  

Cas

Bahahahahaha

[gouessej]

Hehe that's ironic coming from you Julien 

Cas

Lol I wasn't ironic, I'm still far from having succeeding in making a finished product but I still think trying to do it is a nice way to learn a lot of things.

[gimbal]

What better way to learn than by reading the source code of a finished product?

You can learn programming this way, you are generally not going to learn the concepts behind what is programmed though, especially when dealing with something as complex as a 3D game. I can't speak for other people, but I certainly wouldn't be able to figure out how for example occlusion culling works just by looking at the code of a 3D FPS game, I need an article or book to explain the theory to me in detail (with many pictures) before I can even begin to understand code.

[princec]

Trying is the most important thing! Without trying no-one ever succeeds.

Cas

[R.D.]

Trying is the most important thing! Without trying no-one ever succeeds.

Cas

I would go a little bit further and say, failing is even more important.

[Roquen]

The old "axe" about learning more from failures than successes is pretty true.  My take on this stuff is that CS is applied mathematics.  And the only way to truly learn either is to get off your a$$ and work problems.

[princec]

Pedant mode on but... I think I was saying, if you don't even try, you can't even fail either.

Cas

[Roquen]

Sure.  And your position of telling people getting off their bums and coding stuff (even if in the stupidest way possible) rather than getting mired down in minutia is good advice.  Only with a moderate amount of experience can you start to get a feel for when complexity is truly needed.

[princec]

I don't think I'll ever figure it out, myself. I've been doing it for over 30 years now and I still can't get the balance right.

Cas

[Roquen]

My thinking is that you're an expert programmer if you fit into either of these conditions:
1) You've forgotten more about CS than you currently know.
2) You realize that you'll never know all that much.

[gimbal]

I don't think I'll ever figure it out, myself. I've been doing it for over 30 years now and I still can't get the balance right.

Cas

And nobody probably ever will, there are simply too many conflicting interests to ever get to a point where you can follow the mighty rule of "thou shalt engineer software". Conflicts of interest, changing requirements, money, deadlines- they all kill proper software engineering. And those murderers are always lurking, no matter what project you do.