You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Dynamic memory allocation is mandatory in protobuf-c, but it is a resource-consuming practice to use dynamic memory allocation in many embedded devices. Even some low-end chips do not support malloc operation, while nanopb can use static memory allocation to perform encoding and decoding operations, so we hope to add nanopb support.
@chenjingyuanku thanks for your feature request, I have attempted an initial naive implementation in #180
Are you familiar with building nanopb? I have a couple of questions:
The binary build still depends on 33MB of Python libraries, which would greatly increase the size of the final image. Is there any way to build a binary that doesn't depend on the presence of a Python runtime?
Is there any way to build this for aarch64? tools/make_linux_package.sh seems to assume the platform is linux-x86 but this is not always the case here. Can we do a PR to support different arch in this build script?
It uses Pyinstaller to package python source code into binary, I think aarch64 and x86 use the same package command. But since it uses Pyinstaller, it is unavoidable that the packaged binary contains the python runtime. I don't know if my understanding is correct.
Is there an existing issue for this?
Description
Dynamic memory allocation is mandatory in protobuf-c, but it is a resource-consuming practice to use dynamic memory allocation in many embedded devices. Even some low-end chips do not support malloc operation, while nanopb can use static memory allocation to perform encoding and decoding operations, so we hope to add nanopb support.
Acceptance Criteria
No response
Suggestions for a technical implementation
https://github.com/nanopb/nanopb
The text was updated successfully, but these errors were encountered: