Im wondering so lets say I started to code a engine for my game (not any time soon) and I wanted to make a game wich will be played by gamers a fps with the grapics of allmost CS:Source how many people would it take for me to get it done and should I have more programers to work on it with(sound endine, physics engine,server,gui etc...) me or do all the coding myself.

You could do it yourself. If you write it in BlitzMax it'll take you about a year me recons. Do it in C++ and you're looking at maybe 2. But that means you'll have to work every available spare moment when not at school.

A typical 3d engine for commercial release takes at least a year to reach a relatively mature state, and requires the full time input of anywhere between 1 (very experienced) and 5 coders. Incorporating physics and ai into the engine will take equally as long. Development of the engine usually continues alongside and beyond the initial game project using that engine (for example the HDR capabilities that have been added to Source).

As a lone coder, it is incredibly difficult to produce a mature, usable engine and keep up with developments in the graphics field.

You are much better off using an available open source engine (Irrlicht, CrystalSpace, Ogre) and adapting those to your requirements, assuming you are experienced enough to be able to implement them. These libraries often have open source companion projects or bindings to other projects that may fulfil part of your requirements, such as the ODE bindings for Ogre.

The Source engine has been in development for a VERY long time. I'm uncertain of the team that was dedicated to it, but you can bet there were more than 5 coders having input to the project.

Similarly with audio, using an available library (OpenAL, FMOD) is the way to go.

In any case, if you're after a commercial quality product, you're either going to have to outsource part of the work, or get more people involved.

Several issues here.
1. To make an engine that competes with commercial engines you need to be a good programer BEFORE you start. If you are learning as you go along the quality of your work will be lower due to the errors you make and will take a lot longer.
2. Unless you are working full time on the project you will need a bigger team than the commercal games have because each memeber will be spending less time per week working on it.
3. The commercial engines will continue to develop so, unless your team is big enough and skilled enough to keep up your finished project will look like TODAY'S CS:Source engine but by that time CS:Source will have moved ahead.

My suggestion:

Create a design document that describes your game in detail.

Do an estimated technical analysis of your game and answer the following questions:

1. What major development tasks are there?
2. What resources do they need? (Software, hardware, skills)
3. How long will each task take to do?
4. What other tasks need to be completed first in order to finish this one? (Also known as a dependency)
5. Are there any risks are associated with this task?
6. What other alternatives methods can you use instead, if any?

Once you have answers to those questions, you'll know a lot more about the scope of your project - How many people you'll need, what major development tasks you'll have, etc. You may also be able to compare the pros/cons of creating your own engine vs. using a pre-built one.

The idea is just to look at what your goal is and the best (and most realistic) way to achieve it. Other than that, if you really have the passion and drive to succeed, my only other advice is to just never give up! :)

Goodluck,

~RJ

"The best way to predict the future, is to invent it."

Quote:Original post by Obscure
2. Unless you are working full time on the project you will need a bigger team than the commercal games have because each memeber will be spending less time per week working on it.

Um,. I've seen a very nice flashy racing game demo by a guy who worked by himself
for 2 years and he says he spent few hours(not sure if he said several hours) a week.
If he's experiences then it *might be possible *if he works on the top layer of the game development.