Re: [stella] stella charter

Subject: Re: [stella] stella charter
From: Doug Dingus <opengeek@xxxxxxxxx>
Date: Thu, 25 Aug 2005 03:31:24 -0400
On 8/24/05, Lee Fastenau <stella@xxxxxxxxxxxxxxx> wrote:
> > about a program that writes kernels?
> Can I ask a silly question? Why not learn assembly?

Of course you can. It's not a silly question. To be honest, I know how to 
write in assembly. (6502, 6809, x86) At some point, I'll be working on bb 
kernels for specific game visuals.

What I don't have is the time it takes to complete projects in assembly 
right now. I also lack the skill required to put larger programs together 
because my skills are rusty and underdeveloped from lack of use. BB is nice 
because I can focus the assembly just where it might be needed and get 
something done that others might enjoy, while re-learning the art of things 
as I have time to do so.

It seems to me the dynamics of how the TIA & CPU interact are well 
understood. Not completely understood, but almost. Given a particular visual 
goal, we might know enough now to write a program that can translate that 
visual into a 2600 display. The kernel gods here will be able to do it 
better of course, but that's not the point. (And my point is coming)

Bb works the same way right now with game logic. Given a particular gameplay 
scenario, bb can write assembly language code that works. Of course the 
assembly gods here can do it better, but not faster. Well, maybe not faster 
--some folks here are really good, but I hope you get my point. --And by 
faster I mean in terms of development time, not cycle usage.

Of course any one particular display kernel and gameplay combination 
developed with the higher level tools is not going to compare to a 
concentrated, low-level effort by a skilled person. However, being able to 
realize these things at a higher speed will make new game scenarios 
possible, by virtue of the shorter time frame the approach would bring.

Given large rom / ram addressing schemes and these two things together, 
games with a greater variety of visuals all appearing in the same game might 
be developed in a timeframe similar to the games we see today. Multi-screen 
games that present the information to the player in a greater variety of 
ways are still somewhat rare on the 2600.

I simply was musing about the possibility of advancing that particular 
front, that's all.

In the very near future, we will have progressed one step toward that goal 
via multiple Bb kernels being used in the same Bb program. The current 4k 
limit will sharply limit that, but not for too long. In any case, I think a 
kernel program is as possible as bb was and that it will be interesting in 
the same way bb is right now.

> -Lee
> Archives (includes files) at
> Unsub & more at
Archives (includes files) at
Unsub & more at

Current Thread