For my current project, I need to modify a tiny bit (add ~20 lines) to the panda3d source code, particulary the BulletHeightFieldShape.
I have succesfully located the file I require to modify, and also managed to recompile the engine with the changed module.
However, it is really uncomfortable to recompile and reinstall panda3d for every single change I make.
So, how do I recompile only panda3d/bullet.pyd, without any changes to the rest?
I currently use windows (visual studio 2010 and winsdk)
Thank you for any help.
makepanda only recompiles the parts of Panda you changed. Changing one of the bullet headers and re-running makepanda shouldn’t cause anything else to be recompiled.
It is sometimes the case that modifying a compilation setting or a header file used by almost all of Panda will end up causing almost all of Panda to be recompiled if they use this header, but this just isn’t the case for the Bullet headers.
I also strongly recommend compiling with --no-eigen and setting a number of threads for the build using --threads (to as many cores as you have). This will greatly speed up the build.
Ok, I managed to do the compilation part, which works fine now.
However, the changes seemingly apply only if I reinstall panda3d overall, copying bullet.pyd changes nothing.
So, what files do I need to copy in order to apply the changes made to bullet?
Probably built_x64/panda3d/bullet.pyd and built_x64/bin/libpandabullet.dll. Note however that you can run Panda straight from the built directory if you don’t want to reinstall Panda every time. Just use the copy of Python in built_x64/python/python.exe to run your programs and it will use the copy of Panda straight from the built directory, no installation needed.