Game Programming in Squeak

Andrew C. Greenberg werdna at mucow.com
Tue Oct 30 17:56:46 UTC 2001


On Tuesday, October 30, 2001, at 07:54  AM, Andreas Raab wrote:

> Jon,
>
>> As anyone who has done any serious 3D game development realizes,
>> Squeak will probably never be fast enough to do a "modern"
>> first-person shooter.
>
> I'm sure that if you start from that point you'll have plenty of ways to
> prove it ;-)  Ah ... do you remember the days when people "knew" that 
> you
> can't possibly write an operating system in a high-level language like 
> C?!
> ;-)

Indeed.  It has been YEARS since "serious 3D game development" required 
game logic to be integrally related to graphics engines.  Indeed, most 
games have the graphics engine under the hood, and many rely on open 
source engines for rendering, with an intermediate wrapper that in turn 
faces the high end game logic.

No doubt, the days where Mike Abrash's "Zen of" bibles were de riguer 
are long gone for most game coders.  They are GREAT books, full of fun 
code and insights.   But that kind of optimization is as obscure and 
unnecessary as the five-layers-of-interpreter we used to fit Wizardry 
onto a 48K Apple with a 140K floppy.  The art of small and slow machine 
optimization is deep and wonderful.  I love it.  It is also irrelevant 
to modern games and game design.

I have hacked games in Smalltalk without running into any serious 
resource issues, from time to time resorting to plugins to provide the 
(mostly input) capabilities lacking in Squeak and interfaces to graphics 
cards.  Now that we have a 3D engine framework, even that may prove 
wholly unnecessary.  The only distance between a 3D game programmer and 
a game today is the game, using Squeak or C.

And as a graduate game designer laureate, let me tell you, the game is 
the hard part.  Next, graphic design is probably the biggest challenge 
for an amateur.  Code isn't an issue.





More information about the Squeak-dev mailing list