[gameprogrammer] Re: "Mission Efficiency" - Blue Collar coding

On Thursday 13 May 2004 21.27, Bob Pendleton wrote:
[...]
> > What I think would be nice is to invoke the Python runtime
> > environment and have it read my C program and run, so my script
> > and compiled code source all look and feel the same. Now that
> > would be useful!!! Okay, well maybe not so much for strings, the
> > entire point of using the scripting language is to get away from
> > walking through character arrays.  I know you had started a
> > thread a while back on what would make the perfect language to
> > write games in, but you had taken that offline.  I would be
> > interested in any feedback you'd care to share though.
>
> I didn't take it off line, the discussion just died. No feed back
> at all.

Being that I'm hacking a real time scripting engine as we speak, I'd 
be very interested as well. Right now, the primary goals are real 
time control, data aquisition, configuration and communication - ie, 
what I get paid for - but my spare time plans include audio synthesis 
and games.

Problem is that I haven't done any serious game scripting before, and 
I only have limited experience in hacking game logic in BASIC, asm, 
Pascal, C and C++; no scripting langugages, and certainly no 
languages designed for game scripting. (Well, apart from Q3 shaders, 
if that counts. ;-)

So, I figured I'd just lurk and take note of any interesting ideas, 
but that didn't work out too well... :-)


> > Honestly i'm beginning to think games require a whole new
> > paradigm. The graphics portions (and internal datastructures)
> > lend themselves well to being organized in an objective manner,
> > the system integration (signals, threading) and much of the
> > mathmatic areas seem to scream for a functional layout, while the
> > AI is a wonderful candidate for an associative set of
> > relationships (Go Prolog!).
>
> Yep. I can see that. I think that there is a lot of thinking that
> needs to be done about a data structure that subsumes the concept
> of structures/classes/matrices/databases. Some kind of a persistent
> associative store that can store code as well as data.

Tables (Lua style; <name, value>, where both 'name' and 'value' can be 
of any type) can cover some of that... What's missing? I'm 
particularly interested in the idea of applying associative database 
properties to something like these objects.


> And, I been thinking about the thread and storage management
> requirements for the run time of a language. I think you have to
> start at both ends of the problem to do this design. You need to
> work back from the desired semantics and forward from what is
> implementable.

That's pretty much what I'm doing right now; looking for ways to get a 
nice language without too insane requirements on the implementation. 
Problem right now is that I'm short on time, so I'll have to stick 
with mostly tried and tested designs for now.

Oh well... Need to get back to hacking.

If this turns out right, maybe EEL could serve as a testbed or 
starting point for our game scripting ideas? It's LGPL (considering 
X11), some 11k lines of (hopefully) portable C, designed for hard 
real time applications, and has a simple and lightweight interface to 
C. Raw VM speed was on par with Lua last time I checked, but the 
stupid compiler in that version generated about twice as much code as 
needed, resulting in roughly half the performance.


//David Olofson - Programmer, Composer, Open Source Advocate

.- Audiality -----------------------------------------------.
|  Free/Open Source audio engine for games and multimedia.  |
| MIDI, modular synthesis, real time effects, scripting,... |
`-----------------------------------> http://audiality.org -'
   --- http://olofson.net --- http://www.reologica.se ---



---------------------
To unsubscribe go to http://gameprogrammer.com/mailinglist.html

Other related posts: