@nnnnnnnn brought up that Sequence.min/max implementations in the stdlib do not specify which element they return if there are duplicates. I think they should!
This would also be a chance to make this consistent. Currently, the top-level max functions return the last dupe, while Sequence.max returns the first. We should choose one behavior and consistently stick to it. (Given that the behavior of the top-level max is documented, Sequence.max should probably adopt it, too.)
So the Swift 5.6 release process was just published. Do you think there is any way this would make it into that release?
What would be the next step? A pitch thread to try to gather more changes? At first I thought this would just be changes in documentation, but if the idea is to also change the behavior at the same time, some actual implementation will be required
I don't see why a narrow and relatively uncontroversial proposal that ties down behavior through doc updates couldn't fit into 5.6! I don't think changing the behavior of Sequence.max would complicate things, either.
S-E proposal scheduling bottlenecks could happen, of course -- it wouldn't be the end of the world if the updates get delayed by one release, as long as they eventually happen.