That's basically an extremely simplified version of how game development already works.

In actuality, all three phases need to be interleaved because of things that you will encounter during implementation: Problems in implementation (conflicts between different design goals), problems with performance, or just deciding that something you made isn't as fun as you thought it would be. The more complex your project, you more opportunities there are for things to deviate from the plan.

Well imagine you are so advanced that you don't make any mistakes... and you have all the knowledge to take the game from step 1, to step 2, to step 3 without retracing your steps and having to learn/figure out a workaround.

A game is a computer program.
Every non trivial computer program consists of at least one variable, a branch, a loop at least one bug.

That's basically an extremely simplified version of how game development already works.

In actuality, all three phases need to be interleaved because of things that you will encounter during implementation: Problems in implementation (conflicts between different design goals), problems with performance, or just deciding that something you made isn't as fun as you thought it would be. The more complex your project, you more opportunities there are for things to deviate from the plan.

Well imagine you are so advanced that you don't make any mistakes... and you have all the knowledge to take the game from step 1, to step 2, to step 3 without retracing your steps and having to learn/figure out a workaround.

A game is a computer program.
Every non trivial computer program consists of at least one variable, a branch, a loop at least one bug.

Speaking of Bugs. Selective Quote no longer works on the forums :/

I agree with Tom Sloper on this one: game DESIGN has so many different approaches. Depending on what you approached first: the aesthetic, the mechanics (or a core mechanic) etc. you'd end up with two very different games because each angle you approach the design from can come with different mind sets.

It's an artistic process. Enjoy it! Let it challenge you and take you on an adventure equal to the game play you're trying to develop. By all means, structure your day and actively develop habits that work your idea muscle, study games and build an arsenal of useful mechanics, identify useful design patterns, apply constraints when you're in a creative rut etc. But just go with the flow or you could be missing out on the opportunity for a really good game.

Game development, to my mind, is an umbrella term. I think you need to go through the game design process first to truly appreciate how to tackle prototyping and the technical delivery of a video game.

Also, I really feel a game in general is more than just a computer program! Ha ha.

Also, I really feel a game in general is more than just a computer program! Ha ha.

Well it could be a board game or a card game but from a computer science standpoint a video game is nothing but a fancy database.

O_o badly disguised troll or for real?

The point I'm trying to make is that when you create the "framework"... it shouldn't take you a long time... it shouldn't even take you a few days. And I think if it takes you any longer than that it means that you are actually LEARNING how to make your game... meaning you are trying to FIGURE SOMETHING OUT... meaning you don't actually know what you are doing.

So say if you wanted to make a clone of WoW or Halo or CoD or StarCraft or GTA... even in a different language/IDE then what they were originally made in. It shouldn't take you more than a few days to craft the framework because the scientific theory on how to build all of those features already exists.

a) what is the point of recreating an existing game? You would basically set yourself for legal trouble if it wouldn't be just for personal learning never to leave your own PC.

b) do you have ANY idea how real game development works? Doesn't sound like, else you wouldn't put "clone WoW" and "in a few days" into the same sentence. As soon as you have to recreate even a fraction of the needed assets and code, even if you have access to the original and just have to type/draw/model it to copy it, it will take WAY more than just a few days.

c) Creating games is a constant process of learning. If you stop learning, improving old and inventing new things with every new game, you basically started to push out clone after clone... how long that will work out without gamers getting bored off your games is anyone guess, it might look like it works for the likes of EA and similar "sweat shops", but even their clonetastic games like CoD and FIFA have to come up with something new with every iteration, or risk having to make a big reinvention of the series like what was direly needed with CoD modern warfare.

The theory presented is much like a physisist trying to determine the smallest field suitable for a given number of cows, who begins his solution with "First, we assume a spherical cow..."

Here, OP begins with "First, assume perfect knowlege, perfect execution, and an infinite library of predefined components that are all perfectly composable; Let us also assume that components of the library are simultaneously highly-detailed and universally well-understood, and are sufficiently high-level to be productive...."

If we assume that does exist, or could exist, sure, I guess one could bang out step 2 in a relatively short period, but it doesn't exist, likely can't, and certainly won't for a very long time. Creating it would be a herculean if not impossible task. Furthermore, for it to be a win, we also have to assume that achieving perfect understanding of our creative design space is firstly achievable (but no bother that...) and that expressing it in terms of this framework (which is by definition too large for any human or group of humans to understand) is quicker than a more iterative, collaborative, organic process. Then, you also add in step 3--content generation--which is post-hoc and therefore does not benefit from any assumed productivity gains (creating and importing content in this perfect framework is no easier or quicker than in a custom framework) -- since content generation is already the largest part of man-hours invested in a large game by far (easily one, perhaps two orders of magnitude), then any gains in step 2 are swallowed up by step 3 anyways -- for any proposed method to decrease development times, it must decrease the time spent on content creation.

All of that said, the theory presented is basically a Polaris or 'North Star' -- where we're at with game technology today lies somewhere along the line between where we started and this idealized framework, and where we want to go in years to come is somewhere along that same line but closer still to the ideal. Frameworks like Unreal and Unity already strive more and more that direction with each version but are, reasonably-so, still a long, long ways away. Technology will continue moving this way, broadly speaking, but we're as likely to reach this destination as we are to count to infinity.

The theory presented is much like a physisist trying to determine the smallest field suitable for a given number of cows, who begins his solution with "First, we assume a spherical cow..."

Make it a cubical cow, that would be the ideal to get as many cows in a field as possible... they might not be able to turn anymore, but to hell with that! Why don't we put multiple layers of tightly packed cows into that field....

"What do you say? They are there to eat grass to actually produce milk? There is not enough grass anyway for even a single layer of tightly packed cows? .... well that is not MY concern, I am only here to improve the packing density of cows in the field."

I would go one step above what Ravyne said and would actually say that even IF we could reach that level of optimization one day, the output of such an optimized process would be questionable.

Would it still be innovative and fresh enough to have any market value? Would it just be copy-pasta elevated to a new level of genericness?

With all creative processes there is a fine line between productivity and churning out generic crap. Of course not every work you do can be the greatest work of art of the century, else no game would ever be finished. At the same time, if you just take a generic human model and put generic clothes on it, you just created a generic character.... without any memorizability, nothing to set it apart from all the other generic characters out there... a character without, erh... character.

That is why I am quite ambivalent to all the cool productivity improving tools out there, worst of all character creator tools. Are they a great help to boost an artists productivity? Hell yes. Do they eliminate the need for manual tuning of the result and creativity? Hell no, else all assets churned out from these tools look the same.

You can use a texturing tool like DDO or a character creator like Fuse to create high quality content quicker by giving you a head start where it doesn't count (character body, dirt and grime for the texture), so you can spend more time where it does count (Characters face and clothing, bigger details on the texture).

These things are still tools that needs to be mastered, and are not replacements for missing skill.

In the end, the goal cannot be to churn out the most generic content in the least amount of time. The goal must be to reduce the time needed to create the 80% of the content that is generic anyway while still spending the full time on the 20% that count.

Hello.

The point I'm trying to make is that when you create the "framework"... it shouldn't take you a long time... it shouldn't even take you a few days. And I think if it takes you any longer than that it means that you are actually LEARNING how to make your game... meaning you are trying to FIGURE SOMETHING OUT... meaning you don't actually know what you are doing.

Lately Ive been thinking about where all the time went on my game engine thing. Your right learning.