gh-74690: Document changes made to runtime-checkable protocols in 3.12

[AlexWaygood]

Adds docs, NEWS and WhatsNew entries for changes made in these PRs:

• gh-74690: typing: Don't unnecessarily call _get_protocol_attrs twice in _ProtocolMeta.__instancecheck__ #103141
• gh-74690: Micro-optimise typing._get_protocol_attrs #103152
• gh-74690: typing: Simplify and optimise _ProtocolMeta.__instancecheck__ #103159
• gh-74690: typing: Call _get_protocol_attrs and _callable_members_only at protocol class creation time, not during isinstance() checks #103160
• gh-74690: Further optimise typing._ProtocolMeta.__instancecheck__ #103280
• gh-103193: Micro-optimise helper functions for inspect.getattr_static #103195
• gh-103193: Use LBYL idioms rather than EAFP in inspect.getattr_static #103318
• gh-103193: Speedup and inline inspect._is_type #103321

• Issue: Performance of typing._ProtocolMeta._get_protocol_attrs and isinstance #74690

[hauntsaninja]

The changes look good. Only thing I'm thinking of is that most users reading What's New still don't have the context that runtime_checkable isinstance is like 400x slower than regular isinstance or whatever. Feels hard to communicate "this thing is 20x faster than it was, but is still 20x slower than you might think it is", but maybe there's a good way :-)

[AlexWaygood]

Only thing I'm thinking of is that most users reading What's New still don't have the context that runtime_checkable isinstance is like 400x slower than regular isinstance or whatever.

Well, only ~15x slower in 3.12 for simple protocols ;)

Feels hard to communicate "this thing is 20x faster than it was, but is still 20x slower than you might think it is"

Yeah... I know what you mean. I feel pretty torn. On the one hand, I agree that this is important context. On the other hand, I feel like this is pretty clearly documented in typing.rst now; that the performance improvement is worth celebrating on its own; and that "What's new" should be primarily focussed on the difference between the Python versions.

I think I'll leave this as it is for now, but happy to accept improvements in followup PRs if we can think of a better way of phrasing it :-)