On Tue, Oct 13, 2020 at 8:16 AM Ma Lin firstname.lastname@example.org wrote:
Zstd has some advantages: fast speed, multi-threaded compression, dictionary for small data, etc. IMO it's suitable as a replacement for zlib, but at this time: 1, If it is included into stdlib, it will take advantage of the huge influence of Python and become popular.
Let's stipulate that zstd is "better" than bzip2 or lzma is relevant ways (although the reality is less unambiguous).
Zstd was created in 2016, Brotli in 2013. Those are pretty new. Better is nice, but adoption of either of those is only middling. And the two are different and incompatible, although offer largely similar benefits. However, I am not at all certain that someone won't introduce something new in 2021 that is better in every technical way than either Zstd or Brotli. That doesn't make those bad, but Python isn't trying to have the "optimal cutting-edge" thing in its standard library. More like "the well-established, widely-used" thing.
I feel like you making Zstd available on PyPI is wonderful, and a huge service to the community. But I'm -0.5 on adding either it or Brotli to the standard library at this time. Until they are older and more established, it feels premature.