[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

Re: [pygame] Where Do I Go From Here?

Kris Schnee wrote: 
Many questions here; I could use help from someone more experienced.

But what now? Although I could and probably should focus on the
"worldsim" itself -- the physics and gameplay -- I'm not sure about how
to develop the graphics system further. Assuming that I can successfully
combine this landscape with my Pygame-only widget system and with
Pyglyph so that I can do text windows, buttons etc., it's still going to
be ugly! I thought about switching from these awkward-looking textures
to plain colors, but (perhaps because there's full lighting) that turns
the landscape into green soup. And there's still a bad case of the
jaggies (and terrible FPS) from the fact that I'm using 100x100 data
points. _And_ the sky's not animated. And so on.

So, as a lone developer, is it even worth trying to do this 3D system?

Truthfully, I wouldn't even attempt to write a 3D system from scratch. There are lots of open source and commercial 3D engines out there - I'd pick one and go with it, rather than reinvent the wheel. 3D programming is much harder than 2D programming, and bigger teams than yours have already plowed that same ground.

My reasoning is:
-That bad 3D is worse than good 2D;

I personally don't think so - take a look at http://www.kamilche.com for some screenshots of a 2D game I'm currently developing. In truth, I've been working on it for several years now. I've programmed it in a variety of different programming languages, but Python is the one that 'stuck.'


-That Pygame scrolls slowly (I'm getting just 15 FPS on my non-scrolling 2D engine with Psyco, vs. 32-34 on a minimal Pygame program with/without Psyco)

Pygame is slow for scrolling games - they have a hard time keeping their frame rate up. There are probably faster engines out there, but the ones I've seen force you to stick to the Windows world, or don't have Python bindings, or don't have per pixel alpha, etc.


-That OpenGL is potentially faster if I can make lists work properly (it's around 4-7 FPS right now!); 

Ouch.

-But that I'm tired of being bogged down switching between graphics engines and still having the program look bad. I want to build gameplay and AI, not write yet another display system!

It seems to me that you really want a 3D game, and won't be satisfied with 2D. If that's the case, you really have one decision to make - to continue to develop your own engine, which it sounds like you don't want to do), or to look at pre-built 3D engines that have Python bindings.

--Kamilche