Francesc Alted

@jstriebel Yeah, as said, it is up to you to decide whether Blosc2/Caterva can bring advantages to you or not. It is just that I was thinking that this could...

Well, if it helps, HDF5 includes support for checksumming, but it is disable by default, so +1 for disabling it by default here as well. Regarding omitting the offsets by...

Well, I think what you describe is completely compatible with the docstrings for blosc_compress(): > The `dest` buffer must have at least the size of `destsize`. Blosc guarantees that if...

Reviewing this, I think I did not understand well the question. In fact, Blosc ensures that `destsize

Yes, using scikit-build is probably the reason why this is happening. @thewtex do you have some suggestion here?

Yeah, this kind of wild variations in compression ratio for small buffers is expected. Blosc is actually meant towards compressing large datasets, so priority in optimizing such small buffers is...

Absolutely. When Win32 CI crashes, we re-run the check and it normally passes there. Being so elusive does not help to find a fix. All help is welcome.

Thanks for the feedback. Unfortunately I don't keep a recent sample of failure, but it is a good idea to activate `setvbuf` and add a failed log here.

@rlizzo Since 1.8.3, python-blosc requires scikit-build (as specified in `requirements-dev.txt`). Can you try to install it and see if that works? @dkleissa Can you send some hints on what part...

Well, what it happens is that the default is not to compile snappy, but it can be activated via `cmake -D DEACTIVATE_SNAPPY=OFF` again. It should be a matter of recompilation....