Francesc Alted

Ok, so `global_dict` is not honored since that commit (May 23, 2023)? That probably means that not many people use it. Maybe should we deprecate it; or do you think...

Please do not mark as stale.

Hello. As you probably know, you should not use private variables in (like `numexpr.necompiler._names_cache`) in your wrappers. Having said that, maybe @andfoy has some suggestion to your issue.

Unfortunately, there is no explicit support for 32-bit OS/CPUs in Blosc2 anymore. We are actively supporting AMD64 and ARM64 as the main deployment platforms, so support for others will need...

Ok. So your issue should stem from here: https://github.com/Blosc/python-blosc2/blob/main/CMakeLists.txt#L51-L54. The reason is that normally there is quite of dependency of C-Blosc2 sources, and this is a nice way to provide...

Can you tell us what advantages that brings for users?

Interesting. Looks like a weird interaction with old numexpr packages when used with new NumPy 2.3. At least we have been fortunate to have recently released a new version of...

Yes, that [commit](https://github.com/pydata/numexpr/commit/33ee71b0c6f13224f3031cd8b42921c748ce9ede) was added during the support for free-threaded python. So you can proceed with this new assumption.

In general, Blosc2 only allows for 1 byte per [user-defined codec](https://www.blosc.org/posts/dynamic-plugins/) to be stored in the header of the compressed chunk; see [udcodec and compcode_meta](https://github.com/Blosc/c-blosc2/blob/main/README_CFRAME_FORMAT.rst). So, if metadata for the...

Uh, I am afraid that we did not try to make vlmeta API threadsafe, so assume that it is not. Also, Zstandard is a very nice codec, and with the...