Not sure if this would actually bring any performance improvements, but it is possible it could since the implementation would be entirely lock-free and im has some interesting optimizations. A benchmark would be required.
I doubt I'm going to work on this much, but would be open to PRs.
Not sure if this would actually bring any performance improvements, but it is possible it could since the implementation would be entirely lock-free and
im
has some interesting optimizations. A benchmark would be required.I doubt I'm going to work on this much, but would be open to PRs.