AUUUUGH

Perhaps I should just label everything I say as "Subject to change. Do not consider any evidence as permanent, for any period of time longer than never."

Okay, okay, shall I explain.

Jetmen Survival's out, I'm back to Crystal Blue (3d engine) again. Although this time, I've hacked out pieces of JMS and CB and plugged them into an improved engine concept, renamed GE (Generic Engine, or Game Environment).

JMS let me accomplish some pretty good stuff; essentially I think I'm going to be using the glee behind a game concept each time to go further into coding. "Pretty good stuff" being:
1. Texture upload and an understanding of using them. (Have not dealt with texture environments yet.)
2. The data transfer bandwidth limitation of most graphics hardware and ways of dealing with it.
3. Fast manipulation of images in software, from a blur algorithm to safely copying areas between images.
4. The standard C family of block memory functions.
5. Better file access and pointer safety methods.


The new engine has a radically (hehe, I said radically) different structure, as it is based around abstracting reality, instead of around what computers use to simulate it.

It also has a different resource control method which takes a leaf from Python's book and uses reference counting on object pointers.

Meshes support multitexturing and materials, in storage. (The construction of Crystal Blue was too complex for me to understand how to support it!)

At this point I am still creating OpenGL setup functions and reference counting/autodeletion.

..

Does none of that make sense?

I'd just like to say something here.
Understanding the things others build is hard. This problem is something I have run into every other day. The complexity of a program can far exceed that of a physical object and it has no basis in reality, no rules to adhere to.

Therefore, I have personally tasked myself with creating machines understandable on instinct. Perhaps not when you go deeper, below the layer marked "This is the boundary between your right brain hemisphere and the computer", but on the surface. GE somewhat represents my first strong attempt at this goal, so I am hoping it will take me somewhere.

Some guy somewhere once said something meaningful about computers and people. It had a thing in it, I thought it was good because it pointed out how computers were.. somethingish, and humans weren't meant- WAIT NO now I remember also paraphrasing obviously:
"A program should be more easily read by the human, than by the computer running it."

The interface "spec" of GE, in terms of what your right brain can think about:

Create an environment.
Create an object.
Place the object in the environment.
Look at the environment.

I see an object!


As I've said to a couple of people while describing it.. it's somewhat depressingly simple. But it works..! And the C++ isn't that bad at matching it, either.

ALTHOUGH YOU HAVE TO PUT THE BIT WHERE YOU LOOK AT IT IN A LOOP CONTROLLED BY A CLASS WHICH CAUSES MULTIPLE LOGICAL TIMESTEPS BEHIND THE RENDERING OPERATION SO THAT THE FRAMERATE KEEPS UP WITH ALL YOUR OBJECTS MADE OF MESHES AND DEFORMER PIPELINES AND I HAVEN'T FIGURED OUT HOW TO KEEP THAT BEHIND A FUNCTION BECAUSE YOU CAN SET A 2D RENDERING MODE AND BECAUSE USING 3D CAMERAS WILL UPSET THE PROJECTION MATRIX OF THAT (and i'm too scared to put in a safeguard) IT MEANS YOU SHOULD BE ABLE TO RENDER MULTIPLE ENVIRONMENTS FOR OVERLAYS AND THAT'D REQUIRE THE USER TO DEAL WITH A BLASTED, BLASTED CLASS FOR STORING MULTIPLE ENVIRONMENTS AND THAT DOESN'T DAMN WELL WORK IN REALITY UNLESS YOU HAVE A PORTAL GUN.