How to get started

Quake 3 Source is out

The Quake 3 source is out now too, you can get it a number of places by googling for it:

http://www.fileaholic.com/idgames.d/quake3/official/source/

Just thought I would mention that...

[Edit] Guess this is just the game source, my mistake...

 

I think they meant the Quake 3 engine source, not the game source, the game source have been out for years.

 

Yeah, the engine source was supposed to come out at the end of last year, but John Carmack made an announcment that a recent Quake3-based deal had recently gone through, so the release of Quake3 is delayed until some nondefinite future point in time.

Having said that, there are a few Quake2-based engines out there that have evolved to the point where they basically support almost everything that Quake3 does. Check out Qbism - it's pretty decent for what it is. From my limited experience with it, it does have some bugs (that I've come across just in some code inspection) and doesn't do many things so optimally, but it could be a good place to start for people looking to do Quake3-quality games.

 

forget engines, forget source code

If you don't know much about games or programming them, then looking at source code (especially Id's) is only a lesson in wasting time.

IMHO you need to start at the 1000-mile high view and work your way downwards to the details of source codes and engines.

How does a computer game run? What is involved in a basic game? What is a game loop? etc..

Solve these "large" questions, then you can work your way down to the engines, then into 2D/3D API's (ie. OpenGL, Direct3D, DirectDraw, etc)...and then even further into different languages etc..

Of course, that's only if you learn things better that way. If you'd rather just learn from reading source code, then by all means, grab the Q2 stuff...and tylenol.

hth,

 

Well, destron, have you started trying out any of the packages mentioned above yet? Programming is about doing - start doing something, and see whether you like it. I would also recommend trying out the free demo version of Blitz.

 

Honestly, you should make a simple tetris game before you try to duplicate Quake II. Maybe you can, but you can learn a lot about your framework, package, how to work with artists, and yourself, by making a simple 2D game using proven gameplay.

Best regards;
Emmanuel

 

Destron,

If I remember correctly, the Blitz trial can only be started a certain number of times (30 or so, I think). And it can't compile EXEs - you just run projects through the IDE.

I can recommend BlitzPlus, which is the 2D version of Blitz3D and is slightly cheaper. There's nothing very complicated in there at all - no 3D stuff. just basic 2D commands, and it's all centred around the BASIC language (which is easy to pick up). It comes with examples of some very simple games. All in all, a great tool for beginners yet it can create professional 2D games. Registered users can chat on the forums too, which can save newbies a lot of time and effort.

My test PC is an old P500, and BlitzPlus works pretty quickly on it. The minimum spec for BlitzPlus is a really old version of Direct X, so it's compatible with most Windows PCs (old and new).

Just start off creating some simple games (Space Invaders, Pac-Man, that kinda thing) to learn the language and different techniques and then you'll be able to eventually progress to writing your own game.

 

Blitz*max* demo is a 30 day trial (I have it at the moment - 27 days remaining to finish my game.. and it looks possible in this lol *joke* - I don't know what kind of functionality blitzplus has compared to that (whether is can use OO, user defined types, member functions etc) but Bmax looks like you can do a lot with it at a lower level (if you choose to).

so maybe check that out too - It looks easy enough (if you steer clear of pointers etc) but reminds me of C++ in places, which for me is a good thing because it doesn't make me feel as restricted or boxed in as I might otherwise. It looks to all intents and purposes like a very well written abstracted framework on top of C++ that can be used in the same was as BASIC.

I would reccommend running in STRICT mode though (apart from for quick tests) there is lot to be said for doing things "the right way" 3 months down the line when bugs (that advanced languages were designed to rule out) otherwise may have started to creep in.

 

More than 5 years and more like 100 years..

 

That's what I was thinking.. look at Halflife 2 - 6 years and a multiman team... and they still can't make it non-linear and last more than 10 hours

 

That's no excuse. Frankly, in the world of computers, if something is taking too long, it's almost always possible to have the computer do some of the work for you. For example, it could have taken a lot less development time writing a utility to import images/textures of buildings, cut them up into wallface, window, door, etc, and construct randomly generated buildings (in 3D) based on that information. I promise it's a lot easier to click "No" "No" "No" "No" "Ok" "No" than design a building, have the art director say "nah, it's no good, make another one" and make another one. Placing those larger entities in the world would then become a separate process (which it should be) and not part of the level design. Background art (such as non-interactive buildings) are becoming the "draw every star by hand" bit, and it's unnecessary in this day and age.

 

not really the point I was making.. more that if one man was making a game to rival the big retail efforts.. it would take an amazingly long time to get anywhere near a decent result..(ie - from scratch a lot more than 5 years not accounting for the resulting game being compromised in seperate areas of art, ai, physics, sound etc) mainly based on art assets and content rather than the core code (the physics for example which valve "borrowed" anyway) and other code which once done is reusable.. CONTENT production is the biggest and most labour intensive problem facing modern game creation.. because it is becomming more and more the only "Unique" thing about a game.. which is why UnrealEngine3 is doing so well, devs want the details taken care of so they can concentrate on the game and providing the masses of scripts, characters, models, textures and sounds.

I don't want to labour the point as I know your "5 years" comment was just a quip.. and I get your point

But.. put it this way.. as essentially a one man "team", I would never even attempt to make a Doom3 (from scratch - Total Conversion "mods" don't count) regardless of how many tetris clones I had got under my belt.. there is never a good time to set out making one without a team, and expect to get it done to a competitive standard before technology has moved on. And that's on top of probably not being able to sell it - as those gamers interested in those genres generally want the best looking and best tech.

<edit - should clarify I mean technology led content fest retail games here - not *all* retail games such as pirates, Tycoon games, GBA games etc /edit>

 

Thats insulting to me, and others who have slaved away in front of a monitor for hours for not a lot of money comparitively.

You don't seem to have a clue how much effort really goes into making model after model, and no you cannot automate this. Unless you imagine games use all the same models with randomly coloured textures? in fact your comments amaze me...

 

I appologize for insulting you and others whom you speak for.

On the contrary. I'm well aware of how much effort it takes to make model after model. In fact, my post was about how that effort is unjustly duplicated as projects become more and more ambitious.

I am surprised at this question. Unless you imagine computers being limited to processing only textures and incapable of 3 dimensional manipulation and generation. I don't suggest a computer could design Frank Lloyd Wright architectures one after another. I do, however, know it's well within the capabilities of a computer to take the basic information about the construction of a building (walls, windows, doors, roof, etc) and construct it's own, using textures pulled from the array of building images.

Most of the buildings in Half-Life 2 didn't need an artists attention to every little detail. Only buildings that would come under close scrutiny would need that level of involvement.

While I'm sincerely appologetic for insulting you, I assume you don't write verticies by hand in a flatfile database for your models. That is because you use a tool (a 3D modeler) to assist you in that endeavor (I assume).

It follows the simple idea of the path of least resistance:
if (time to create something > time to create a tool to automatically build an equivalent resource) then CreateTool()

We could argue all day whether or not you think static buildings in the background are worth an artists time, but I'd rightly assume you'd rather be modeling something a little more creative.

 

Hey, I think automating what can be automated is a pretty good idea!