Robin Sommer

On further consideration, I think this is ok to leave as is, I don't think this actually a bug, C++ has the same issue when returning an `optional` vs `std::nullopt`...

That said, #1143 needs coercion for ternary, too, so that fix will address this too.

My thinking here so far is that we can't actually reasonably support a default case for look-ahead switches. I need to mull over that some more, but if that's right...

Looks like we can test this, too: https://github.com/cirruslabs/cirrus-ci-docs/pull/905

@bbannier please see if this is still relevant (not immediately clear to me how to test)

Update: It's not *that* low in priority anymore, but remains something to do until other things stabilize.

Intuitively I'd say it *should* work, but there may be reasons to disallow it that I don't recall right now. Either way, agree, that's not an error message we want....

I take that back, don't think this should work, must have misread this originally. I'm adding type checking.

Thanks for the feedback. I'll take a stab at it, but it might take me a bit.

Updated this as requested, take a look. I added support to Apple SSL as well, but can't test that one because my tests spawns up an SSL server, which seems...