[Bf-committers] restore Game Engine project

[Gregor Mückl]

Am Sonntag, 10. August 2003 03:31 schrieb Patrick:
>   I think I see the advantage of seperating it out now.  I'm looking
> into open scene graph as an alternative rasterizer, as it already has
> culling, is fairly up to date with special effects and opengl
> extensions, and seems less bloated than the current favorite Crystal
> Space.  If we seperate out the engine it will have more flexibility in
> plugging these kinds of things in.  We need to get together and make a
> major combing through of the source documenting/figuring out how it all
> fits together.
>

You got me all wrong here. It never was my intention to use Crystal Space as a 
replacement for blender's current game engine. My goal is making Blender 
useable for modelling real-time contents for external game engines.

Therefore my emphasis was on the following possibilities:

1. easy integration and testing of game models/levels into an external game 
without the need of a programmer to do the integration work
2. ability to add game- or engine-related metadata
3. best possible preview, which would ideally rely on the game engine's 
renderer.

This affects different parts of blender. Some of the work is already under 
way. But the third and (to me) least important point would need a lot of 
extra interfaces in blender.

Most of the features above are interesting for all external game engines as 
well as blender's own game engine. So seperating it and defining 
well-abstracted interface between blender and game engine plugins would help 
the community a lot.

Actually I believe that dropping the current game engine and focussing on game 
engine *support* - meaning making blender a tool for game developers - would 
be better than developing yet another open source game engine.

I recently noticed that game engines that are deeply integrated into game 
authoring tools are generally frowned upon and actually do not wind up 
generating good games. 

Furthermore, good state of the art game engines are a lot more work than you 
possibly expect. Needless to say that many game engines are very specialized 
to the game's environment for a reason. True general purpose engines have to 
be slower and are harder to develop.

I'm not saying this in favor of CS alone. Actually I am currently not really 
supporting any specific game engine.

>     And to the fellow who says kill the game engine, I can see you're
> point of view, as I hardly ever use the rendering functions, or nurbs,
> or metaballs etc.  I pretty much exclusively use the game engine.  It
> might be a good idea in the future to restructure blender in such a way
> that the user can decide what things they want in their blender.  Like,
> I would have the game engine, but I would take out the render module,
> and metaballs, and paths, etc.  And you could not put in the game
> engine.  But there are too many people using the game engine right now
> to just can it.  And there are a lot of really good games so far, you
> must have missed them:)  There are less good games than there should be
> however because there are 4 or 5 major bugs that slow development down
> trying to find workarounds.
>

When you were given the chance to choose from a variety of specialized, highly 
developed and sometimes even modular stand-alone game engine for which you 
can create models using blender, would you still choose blender's own engine?

>     The modular approach would be desirable I think, and we can start by
> seperating the game engine out.  The rest of blender would be much
> harder to do that with :/
>

Seperating the game engine is a good idea. Making a new type of binary plugins 
for external engines (including blender's engine) is an even better idea in 
my eyes. Note that many good game engines do not have python bindings and 
therefore it will be hard to impossible to tie these engines to blender using 
python.

> Anyway, I am optomistic about the future of the engine.  I think with
> some work it could become the premier game engine for hobbyist game
> designers and proffesional RGD (rapid game development), although I
> really do hope some more programmers can help us out:)
>

No. Blender would also need to satisfy the needs of the game programmers who 
need a good script editor at least. And this is yet another hell of a task.

Regards,
Gregor