LGTM, but would be nice to add a test.

We can check the size of the set in the loop. This would give the nefit not only for empty set, but in the case of substracting the larger set from the smaller set. But this is new feature and different issue.

FYI, I don't want to put a size check inside the loop.

Please add also a test for an iterable of unhasables, e.g. [[]]. It will trigger different error.

I don't know what you're talking about. That gives a TypeError just like the existing test.

Thanks @rhettinger for the PR 🌮🎉.. I'm working now to backport this PR to: 3.7, 3.8.
🐍🍒⛏🤖

I'm having trouble backporting to 3.8. Reason: 'Error 110 while writing to socket. Connection timed out.'. Please retry by removing and re-adding the needs backport to 3.8 label.

Sorry, @rhettinger, I could not cleanly backport this to 3.7 due to a conflict.
Please backport using cherry_picker on command line.
cherry_picker 1f11cf9521114447b3e32e2ac88f075ffaa37555 3.7

Thanks @rhettinger for the PR 🌮🎉.. I'm working now to backport this PR to: 3.8.
🐍🍒⛏🤖

Sorry @rhettinger, I had trouble checking out the 3.8 backport branch.
Please backport using cherry_picker on command line.
cherry_picker 1f11cf9521114447b3e32e2ac88f075ffaa37555 3.8

Thanks @rhettinger for the PR 🌮🎉.. I'm working now to backport this PR to: 3.8.
🐍🍒⛏🤖

GH-13967 is a backport of this pull request to the 3.8 branch.

GH-13968 is a backport of this pull request to the 3.7 branch.

These errors are different.

>>> set().difference(123)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: 'int' object is not iterable
>>> set().difference([[]])
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: unhashable type: 'list'

If add the check for the set size in the loop, the former will left, but the latter will disappear. Someone can add such optimization and miss the behavior change. See https://bugs.python.org/issue28071#msg306973.