PyPI library entry

[DonaldTsang]

Would be good to have an entry in PyPI to make this more accessible

[dstebila]

We thought about that, and at this point don't want to add this to PyPI, because we don't want people to start using it for production code yet.

[DonaldTsang]

@dstebila estimated time of release?

[dstebila]

Closing issue. Won't consider a PyPI entry until NIST begins making recommendations on algorithms.

[Ceddicedced]

Maybe the time has come to reconsider this issue? As NIST now stated at least some finalists.

Also it would take care of the copycat who are publishing liboqs packages even-though not being affiliated.

[baentsch]

Also it would take care of the copycat who are publishing liboqs packages even-though not being affiliated.

What do you mean by "not being affiliated"? I personally wouldn't mind anyone taking open responsibility for maintaining a distribution as the OQS team arguably cannot do that, also due to its size. The "do not use this code productively" disclaimer has not been removed from the repo, in my eyes primarily as the OQS team did not author or vet the standardized core crypto routines beyond running some generic tests and also cannot give commitments to provide any specific distro support -- and surely not on the breadth of platforms supported by Python. Right now, this would IMO have to come from anyone willing to create distributions/package things up (or sell this level of support).

[dstebila]

Also it would take care of the copycat who are publishing liboqs packages even-though not being affiliated.

What do you mean by "not being affiliated"? I personally wouldn't mind anyone taking open responsibility for maintaining a distribution as the OQS team arguably cannot do that, also due to its size.

Getting more involvement in the project would be great. But I think it would still be desirable for anyone publishing a liboqs-python package to do so via this repository and integrated with our maintainers and processes.