From the outside in, it's amazing to me that such a decision would actually pass muster and ship:

- The amount of human productivity wasted by people trying to find why major packages broke. There are still thousands of unfixed references on Github a year and a half later

- The effect it will have on adoption of future versions as people are stuck on packages that aren't updated: the most common fix mentioned is to simply downgrade your Python version

- Python doesn't acknowledge that there's such a known cause for this specific import error

- The fact fixes requires changing code, afaik there's no sort of backwards compat flag or something.

It just seems like incredibly short-sighted engineering. There was a deprecation warning, but isn't this an ideological change at most? Was there some greater benefit I'm missing?

I'd love for people more involved in the ecosystem to share their thoughts on this.