06:28 <korvo> https://bpa.st/PNEA I wonder exactly what happened here. `makeList([])` should be fine; there is a top-level statement `emptyList = makeList([])` earlier in the module, so it can be executed, and the types should all line up.

06:28 <korvo> I know that what's actually happening is that rtyper doesn't think control flow will reach this position, but it's funny how I only know that because there aren't any other possible ways to get SomeImpossibleValue into an empty list.

08:05 <korvo> (I have conquered rtyper. I can't believe how proud I am of this. Like, this feels better than speedrunning in public.)

08:11 <korvo> (Wait, KeyError after translation. I take it back; my 2h50m on Zelda 3 last night was not great, but at least I got the Triforce.)

08:48 <mgorny> korvo: sorry, which one is "Zelda 3"?

08:48 <korvo> mgorny: Link to the Past.

08:49 <mgorny> ah

08:49 <mgorny> tried the first one but it was too hard for me xD

08:56 <korvo> For sure, the early games are quite difficult. Zelda 1 and 2 are beyond me. I run Zelda 3, I

08:56 <korvo> *I'm alright at 4 (Link's Awakening), but I can't do the tricks required for 5-6 (OoT, Majora) or later games.

09:07 <nimaje> I finally finished Link's Awakening DX some months ago after letting it lay around, because I was stuck in a temple, not finding a key I needed; in the end I looked up some map with the locations of the keys

09:12 <korvo> No shame in that. As a child I couldn't figure out the final few dungeons; Face Shrine is a serious complexity spike. It's also a very difficult game to speedrun because, while all Zeldas have terrible item menus, Zelda 4's items *move* around the menu as you play, so you have to plan to lay out your inventory.

19:55 <korvo> Huh, prebuilt lists of ints aren't mutable after translation? I have code that is basically `l = []` in global scope followed by a lambda with `rv = len(l); l.append(rv); return rv`. I've debugged everything else around it, but I guess I assumed that this was correct when it appears to not work.

19:58 <korvo> (The actual offender: https://bpa.st/WSTA )

23:00 <nimaje> (is that for a global dynamically sized union-find?)

23:03 <korvo> nimaje: Yep. I'm refactoring it to now be a class which is allocated in main(). Guess I shouldn't have been so lazy and hacky.