Skip to content

Conversation

@KennyChenBasis
Copy link

Fixes #50

I have kept the current defaults, though, again, we might want to change the defaults (if we were to change the defaults to match numpy's, a possible nice interface for patch would be to have it have allow_pickle=None which would mean to not patch Packer, Unpacker, etc., so they'll all use their defaults for allow_pickle).

@arogozhnikov
Copy link

arogozhnikov commented Oct 31, 2022

Hi @KennyChenBasis and @lebedov

My big +1 here.

Just found this package, very compactly written. My thought was to recommend this as a replacement for distributing data by pickles.
However I've hit the issue #50 (pickles are 'on' by default and there is no way to prevent that).

Regarding defaults: I for sure expect that loading has pickles off, and prefer that saving has pickling turned off (but that's an opinion, just saving does not introduce vulnerability per se).

@akapocsi
Copy link

akapocsi commented Jun 8, 2023

This would be great to have in a pypi or conda-forge release!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

pickle is always enabled

3 participants