Carl Meyer

PyPerformance results show neutral impact on geometric mean of benchmarks: https://gist.github.com/ericsnowcurrently/01b79f54360efa52d4c5bb536cc885e5

For the record, discussed this in person with @markshannon and agreed to expand it so that you watch individual types instead of subscribing only to all changes to all types....

@markshannon I think this is ready for your review now; I've implemented the changes we discussed. You may want to take a look at how I chose to document the...

Thanks @erlend-aasland, updated the test coverage according to your comments. @markshannon I think this is ready, pretty straightforward along the same lines as dict watchers, would be great to just...

@markshannon good suggestion, updated!

I guess it still makes sense to "fix" the specialized instructions in 3.11 so that this always-broken SWIG code isn't _more_ broken than it was pre-3.11. I think for 3.12...

We have code in cinder's compiler to inline comprehensions instead of creating a function. It is a perf win but there is a semantic compromise in scoping / name visibility,...

> I'm thinking that any object that is depended on, whether dict, function or class would be allocated an ID. Optimized code would depend on a set of IDs. If...

It has to be added manually. It would be up to @ambv how he wants to address this issue in retype, if at all.

In principle I think the tools already exist for this with `typing.overload` and `typing.NoReturn`. One could imagine this annotation for `assertIsNotNone` (defined as if it were a standalone function for...