On 15May2019 1610, Victor Stinner wrote:
Thanks to the constructive discussions, I enhanced my PEP 587. I don't plan any further change, the PEP is now ready to review (and maybe even for pronouncement, hi Thomas! :-)).
My view is that while this is a fantastic PEP and the groundwork that already exists as private API is excellent, it is too early to commit to a new public API and we need to do more analysis work. We should not accept this PEP at this time.
So far, the API being exposed here has not been tested with embedders. We have very little feedback on whether it meets their needs or would help them simplify or make their projects more robust. I have concerns about the number of new functions being added, the new patterns being proposed, and both forward and backwards compatibility as we inevitably make changes. (I have discussed all of these in person with Victor, Nick, and Thomas at PyCon last week, which is why I'm not doing a point-by-point here.)
As soon as we publish a PEP declaring a new embedding API, users will assume that it's here to stay. I don't believe that is true, as there is much more we can and should do to improve embedding. But we don't get to totally revise the public API on each release without alienating users, which is why I would rather hold the public API changes until 3.9, investigate and design them properly. It does our users a disservice to make major changes like this without due process.
Cheers, Steve _______________________________________________ Python-Dev mailing list Python-Dev@python.org https://mail.python.org/mailman/listinfo/python-dev Unsubscribe: https://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com
