effectfully
effectfully
We use Haskell-like syntax for constants and I don't see any reason to be wrapping those subjects into parens. Would you expect a list of integers to require parens like...
Eek, now that I looked at that diff again, I realize it's not sufficient in the general case, although it's probably sufficient for what you're asking for. I'll try something...
I mean, I don't mind adding support for that to the parser, here's a PR: #7391
This issue was for discussion, no discussion happened, the issue is getting closed. Feel free to reopen.
Yeah, me too haha. Ok, reopening. @zliu41 do you know what our debugger has been up to?
@michaelpj any idea?
> What happens if you get rid of the bit where it tries to prune dead constructors? The test becomes green, see 229b6903412086ed05b468a65e751fb59dd2fb08.
> Since you have a small test case I would dump the output from just after the bad dead code pass and see what's wrong. It looks like a simple...
> I don't see why from just looking at it, some debugging is going to be required I think! Wanna do some debugging on a weekend? :slightly_smiling_face:
I've added an entry to #5967, so that we don't forget about this bug, and let's close this PR for now.