What is the disposition on this PR (asking in relation to #15049)? Looks like the conflict is in generated code, and can be resolved rather simply.

There's no final decision between this and #14130 yet. In the end, I don't have a very strong opinion anymore on which one to choose. The important thing is that we get one of them in, so that we can properly rename parameters and propagate that to pyplot. Since this is internal functionality, we could even switch out the solution if we realize the other one was better.

In practice, we need a second positive review so that this can get merged. The confict is indeed trivial to solve by regenerating the code.

rebased

I want to merge, but I'm wondering if there's an example that would trigger this yet?

@QuLogic I think there are others, but one such example is the s argument to plt.annotate being changed to text (See #15049, as that change is also required for this example)

The argument (and docstring) were changed in the actual implementation, but pyplot still has the old kwarg (with the new docstring)