json_benchmark VS ucall

If you're primarily targeting Python as an application layer, you may also want to check out my msgspec library[1]. All the perf benefits of e.g. yyjson, but with schema validation like pydantic. It regularly benchmarks[2] as the fastest JSON library for Python. Much of the overhead of decoding JSON -> Python comes from the python layer, and msgspec employs every trick I know to minimize that overhead.

[1]: https://github.com/jcrist/msgspec

[2]: https://github.com/TkTech/json_benchmark

- Adding nvme drive support to SMARTie, https://github.com/tktech/smartie, which is a pure-python cross-platform library for getting disk information like serial number, SMART attributes (like disk temperature) - json_benchmark, https://github.com/tktech/json_benchmark, which is a new benchmark and correctness test for the more modern Python JSON libraries - py_yyjson, https://github.com/tktech/py_yyjson, which is still a WIP and provides Python bindings to the yyjson library, which offers comparable speed to simdjson but more flexibility when parsing (comments, arbitrary sized numbers, Inf/Nan, etc) - And some fixes to https://github.com/TkTech/humanmark, which is a markdown library used to edit the README.md in json_benchmark above.

Type checking was included, and union support is trivial to add. We have just added a feature request and will release it in a few days.

You are right! For the convenience of Python users, we have to introspect the messages and parse JSON into Python objects. Every member of every dictionary being allocated on heap.

To make it as fast as possible we don't use PyBind, NanoBind, SWIG, or any high-level tooling. Our Python bindings are a pure CPython integration. There is just no way to beat that combo, not that I know.

https://github.com/unum-cloud/ujrpc/blob/main/src/python.c

TLDR: UJRPC reaches 230K TCP/IP round-trips per second on 1 socket. Faster than gRPC and much faster than FastAPI.