Even if you had managed to get everything to compile, it wouldn't have been built with OpenGL ES 2.0 GPU support. I'm working on a script that will build the whole shebang, including python OGLES GPU support, checking for max CPU RAM and min GPU RAM in /boot/config.txt (during build), swap enabled (preferably on a USB hard drive - at least 20 ~ 50 times faster than the slowest SD cards, and USB flash drives are 5 ~ 20 times faster than the slowest SD cards), etc.
There are some other things that also need to be optimized, but, it's undetermined if the Panda3D engine will even fit in the Pi RAM, much less support rendering anything beyond a mono-color triangle. The problem is that the GPU probably needs max RAM, and the Debian OS alone eats about 80 MB in the stock configuration (that's part of the optimization I'm working on - carving out the OS parts not needed to run Panda3D). It may be necessary to trim Panda 3D down in functionality in order to run reasonably within the tight Pi RAM constraints.
It will be much easier to just post an SD card image with everything configured and built for the proper CPU/GPU RAM split for GPU-accelerated Panda3D, minimal OS resources loading into CPU RAM, optimal swap space, etc. That's going to take a few weeks working on it part-time along with other Pi-finity! educational game system experiments, Pi configuration optimization, etc., (e.g., I'm still wrestling with flaky WiFi issues, USB hub power, building a reasonably robust case with built-in USB hub and 5v power, etc.).
BTW, swap space bigger than twice the CPU RAM in Linux versions from the past few years generally results in increasingly poor performance if lots of swapping is actually occurring, especially to a non-hard disk, which is typically at least twice as fast as the fastest USB flash drives, and upwards of ten times as fast as slower USB flash drives. The upper limit to any drive on the USB bus is going to be less than about 50 MByte/sec sustained, even with a faster hard drive, due to the 480 Mbit/sec USB 2.0 instantaneous max burst speed, and depending on the number of USB devices on the bus and Ethernet and/or WiFi activity that will likely be going on along with everything else.
Oh, and we're not going to be running Panda3D with any kind of windowing system, as there just isn't any spare CPU RAM for that, not to mention no GPU-accelerated X window server (yet), anyway, even if there were enough RAM. That's OK, though, as we want to fill large screen HDTVs with gorgeous 40,000,000 polygon/second animated 3-D images in 1080p video, anyway