Potential Panda User, please forgive the silly questions

Hello, I’m doing my research on 3d engines. I’ve partially committed to Ogre3d with bad consequences as it doesn’t work reliably with my hardware. this led to lots of wasted time until I found this out. Unfortunately, it was very compatible with my workflow: Blender for models, c/c++ for coding, linux, etc.

My question is this: will Panda play nice with Blender and C++ on Linux? I’ve seen references to blender and c++ on this site, as well as linux downloads. The reason I ask is because downloading 38mb on dialup is not trivial and I’d like to put my toe in the water first…

My hardware:

Dell Inspiron 4000
ATI Rage 128 w/ 8mb VRAM
256mb RAM
Debian testing (Etch is it now?)

I’d really appreciate your input.

As for blender — we don’t have a blender exporter for panda3d. Blender does have an X exporter, and we do have an X importer. However, I’m not sure how reliable the two are: we’ve had mixed reports.

As for C++: panda is written in C++, theres no problem calling it from C++. There is one minor issue, though: the documentation and examples are all written with python. Now there’s a fairly straightforward translation from one to the other, but that’s still a bit of an inconvenience anyhow.

Note that python is a very easy language to learn — it took me about 8 hours to really get the hang of it. You might consider learning python and panda at the same time, that way, you’ll be able to make sense of the documentation.

Thanks!

Hm. Are there other linux-compatible modelers that are more reliable at exporting this format (or more direct, without the intermediate step)? My reason for asking is that I could export to my work to that modeler and make final adjustments. What other solutions are there?

Python isn’t the problem, I know python. It’s just that I want to learn C++ too. Plus, performance is an issue on my machine.

Rob

One last question: how does Panda stack up performance-wise? And I don’t mean how much it can do triangles-per-second style. How scalable is it for lower-end hardware?

Panda performance on low-end machines is quite good, assuming that your artwork doesn’t overdo it. (It’s always possible to create lots of huge textures, and that sort of thing will make low-end cards thrash no matter how good the engine). Disney uses the engine for kids’ games, and little kids very often have low-end machines. So they put a lot of effort into it.

There’s no intermediate step involved in using X file format. Panda can load X files directly into a running game. The only downside is that the load times are longer than when you load panda’s native BAM format (it’s doing translation at load time). But that’s the sort of thing you can optimize when it’s time to ship the game.

I think the only issue you’ll be up against is that blender X exporter. Again, I don’t know how it works, or how reliable it is.

Thanks Josh!

That was really helpful. I’ll definitely keep Panda in mind, depending what the project will require and how I can manage with Crystal Space.

Rob