> But most Python project and programmers who are not making CLI tools or other startup-latency sensitive stuff may be better served by the old default import style.
That may well be the case, and Python's development processes would definitely tend to assume the same. That has a lot to do with why PEP 690 was rejected, and why this proposal is opt-in even though many people are concerned that a lot of projects will "opt in everywhere" and create a lot of noise.
Believe it or not, the process is actually very conservative. People complaining about the "churn" caused by deprecations and removals seem not to have any concept of how few suggestions actually get implemented, and how many are rejected (including ones that perennially occur to many new users). A browse through "ideas" forum where new pre-PEP ideas are commonly pitched (https://discuss.python.org/c/ideas/6) gives one the impression of a leisurely stroll through a graveyard.
And many people (including myself) can tell you that you'll often be put through a run-around: if your idea is good and can be implemented, then surely it falls on you to make and publicize (!) a third-party package, to prove the demand and the community support for your specific implementation; but if you somehow get there, now it's trivial for people to install support, so it doesn't need to be in the standard library (cf. Requests).
That may well be the case, and Python's development processes would definitely tend to assume the same. That has a lot to do with why PEP 690 was rejected, and why this proposal is opt-in even though many people are concerned that a lot of projects will "opt in everywhere" and create a lot of noise.
Believe it or not, the process is actually very conservative. People complaining about the "churn" caused by deprecations and removals seem not to have any concept of how few suggestions actually get implemented, and how many are rejected (including ones that perennially occur to many new users). A browse through "ideas" forum where new pre-PEP ideas are commonly pitched (https://discuss.python.org/c/ideas/6) gives one the impression of a leisurely stroll through a graveyard.
And many people (including myself) can tell you that you'll often be put through a run-around: if your idea is good and can be implemented, then surely it falls on you to make and publicize (!) a third-party package, to prove the demand and the community support for your specific implementation; but if you somehow get there, now it's trivial for people to install support, so it doesn't need to be in the standard library (cf. Requests).