[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]
[pygame] Niss
Luke Paireepinart wrote:
I hope this is an addon to NISS and you haven't straight-up abandoned
her ;)
Heh! No way. Now that exams are done I can actually do some work...
Could use any input you have.
<rambling>
My problem with Niss right now is that I don't know what sort of
application I'm building. I toyed with several 3D engines:
-Panda3D: native to Python, but it's generally ugly. Creepy, disturbing
"Hello World" video.
-CrystalSpace: extensive hazing process just to install for Python;
finally gave up
-Ogre3D: PyOgre finally has the CEGUI interface working. Would look nice.
I look at Ogre screenshots and think that anything I'd make with it
would automatically look nicer than anything I've done in Pygame. Plus
it's 3D. But unless I use billboarding, I'd need fancy custom 3D
characters, and they'd look bad unless they were also animated a bit.
So, should I even attempt 3D? In contrast, won't people likely dismiss
anything I do in Pygame because it's not nearly as pretty as an Ogre
program?
Where I left off, I had the "Aquablue" system: an ugly but functional
endless scrolling map with crude editor and mostly-there sprite code.
Should I just use that and focus on refining it? It's innately 2D unless
I build in warps to other maps that represents building interiors etc,
and that means no hills -- just flat islands. But that's OK for an AI
demo and maybe even for a game.
Should I be focused on making a playable game out of it, or on a simple
demo of the AI doing smart things? Or on some kind of potentially
marketable AI chat thing?
I'm _not_ going to enter the Loebner Prize Contest again this year. I
don't have anything substantially better for that than I did last year,
and Loebner's doing odd things with the contest requirements. I've got
various partly-implemented new ideas for the knowledge/decision system.
There's more on my brainstorming at
http://niss-the-ai.livejournal.com/2006/05/13/ .
</rambling>
Kris