rdb

Hey @atulsaswat! That's great to hear. Would you mind getting in touch with us on Discord or IRC? More information is on [this page](https://www.panda3d.org/gsoc-2020-ideas/).

Great news! You may want to coordinate the filter API design with @lachbr who has been working on the FMOD back-end.

I'm fine with auto-detecting `.zstd` based on the header in ZStream, just like what we do with `.gz` files. We don't actually care that much about the file extension, Panda...

I'm going to unapprove this PR. Many of these, like guiScaleTask, setFrameSize, frameInitialiseFunc, etc. are actually internal methods and not meant for public use, so we shouldn't be making aliases...

Indeed, I had intended to comment what @el-dee just pointed out. This isn't the heightfield being transferred, but the data texture. The default PRC settings massively oversize the data texture...

Nim looks like an interesting language whose semantics do seem to be a good fit for creating Panda3D bindings. The fact that it compiles to C++, can directly call C++...

I hacked up a proof of concept here: https://github.com/rdb/nim-panda3d If anyone wants to work on this they can use that as a starting point. It probably makes sense to do...

Agreed.

Note that optimize level 1 doesn't strip the docstrings, so we don't have to disable it altogether. Can we get away with a module blacklist in this case, to not...

Is this fixed by #1359, or does fixing this require a new optimization level setting?