@gvanrossum this one should also be ready.

I thought there was less agreement on this? Or is it a done deal now that reveal_type() is in? (Personally I use the latter all the time but I don't think I've ever used reveal_locals(), so I'm rather lukewarm.)

I'm also not as enthusiastic about this one as reveal_type. The main reason I added it was that you wrote (on typing-sig):

I think we should also add reveal_locals() and assert_type(), to promote, document and standardize them.

@erictraut and Martin DeMello also supported adding reveal_locals(); others in the typing-sig thread just discussed the implementation or reiterated concerns about reveal_type().

I'll send another message to the thread asking people to speak up if they really want reveal_locals(). If nobody likes it we can close this.

FWIW, I don't feel strongly one way or another about reveal_locals. I suggested it for completeness.

I agree that assert_type is more useful.

Nobody has spoken out in favor of reveal_locals(), so let's close this. We can always bring it back later if there's more demand.