Probably not much I can add to this discussion, but it is interesting. I would say that there are two ways game development can be hard. I also have a third point.

The first is not being able to implement your vision due to crappiness of stuff, as a programmer, I find lots of times I want to do something, but because someone had to make it overcomplicated, I can''t do it. API''s like OpenGL are moving in the right direction. Many people reinvent the wheel... and do it poorly. I have messed some with COM objects, an example of something which is so frustrating it prevents good programming(personal example, some people like COM).

The other way it is hard really hacks me off. I had a guy who wanted me to "teach him how to make a 3d engine". So I started with some basic math(BTW, the math skills are only requisite when you''re rolling your own 3d engine, it helps elsewhere, but not as much as clear thinking). Turns out he had been programming three weeks. I suggested he invest some skills in learning programming, this horribly insulted him, as he felt that he should just be able to do it, or I should give him code(which I did, but he couldn''t figure out how to use it), or that in general, the world should somehow provide him with what he wanted. The process should be lubricated as much as it possibly can, but to some its like saying, "I want to write great literature, but I don''t want to bother with actually having to write the book". Suck it down.

My last point(after the flamebait) is that eventually some kinds of programming are refined until the creation of application specific tools are the penultimate expression of the requirements. Perl is an example for text processing. For the old-school text adventure games, I understand it has reached that stage of refinement. For the rest of gaming? Nowhere close, but I would ask you to consider the mod community and Unreal Script, these are the beginnings. Sure there is always a need for better graphics and sound, but at some point, an avenue has reached a stage of refinement where the specialization involved makes the creation of certain tools a neccessity. In fact, I would suggest games will arrive at such a solution in a hybrid fashion. The parts of games which are not fully explored(graphics) will be written in general programming languages with API''s, and the rest of the game will be written in a refined language that will allow the rest of the game features to be expressed concisely.</p>

Greg, you''re wondering why these powerful game-creation tools don''t exist... I think it''s because nobody wants to make them. A powerful, flexible game-creation tool (API, whatever) would be (1) more complex to program than your typical professional game, (2) much less fun, because you''re not really creating anything tangible or visible, and (3) less profitable than your typical professional game, particularly considering how quickly such a tool would become technologically obselete.

Well, aside from hobbyists (which is basically everyone on this board), people are interested in reducing the difficulty of creating interactive media (including games). Because that means more money made with less people in less time. This is the basis for licensing engines. To use an example, BioWare''s Inifinity engine was created and used to make several RPGs, including Baldur''s gate and recently Icewind Dale. But, the engine itself was so complete that the Icewind Dale project only had a single engine programmer (lots of scripters though) and everyone else was mostly artists and writers/scripters.
But, this simplicity comes at a cost: the technology in Icewind dale is really out of date. That''s the basic paradox here: in order for an engine to become popular it has to be used to create a really popular game. But two years after that popular game is released, when all the games using that engine are released, the engine is already obsolete.
In the not-too-terribly distant future I''m sure we''ll reach a threshold in both hardware (speed and power) and software (realism) where a single engine can be used for many games across a large span of time. When that time comes, it''s easy to foresee game development teams consisting of a single programmer and dozens of content creators.. but until then, you''re basically out of luck.
Still, it''s not like you need to have games that use the absolute latest technology . Making mods is (relatively) easy, depending on the engine gives you a ton of freedom, and requires only basic programming skill. Plus you get to see the world you create emerge right before your eyes, while programming a game from scratch can take months before you can even see anything.

Good luck =)
-RWarden (roberte@maui.net)

Greg,

I am probably one of those "math geniuses" now, but we were all where you are now - i used to visit the university library to read books on 8086 assembly when i was at school. and i hated it. i would have been happy to use QuickBasic to do a game that was fun.

And this is where the importance is: Fun.

There are many people out there who are hell bent on writing Yet Another 3D Engine (except theirs will be even better). They make good programmers to join a team, but they are not true Game Developers.

Game Development is much more then rotating points and plotting them Ever-So-Fast{tm}. It is making something that is FUN TO PLAY.

so start small and make sure what you are doing is fun.