Newest at the top
| 2025-02-11 20:40:04 +0100 | <exarkun> | each on a line by itself is likely quite parsable, even if it is a type error |
| 2025-02-11 20:39:43 +0100 | <euouae> | what is the big difference? |
| 2025-02-11 20:39:23 +0100 | <exarkun> | not valid code, just parseable |
| 2025-02-11 20:38:56 +0100 | <euouae> | Yeah exactly, I need to have valid code before I can even look up the symbol :P exarkun |
| 2025-02-11 20:38:45 +0100 | <exarkun> | (+/- some random hls quirks) |
| 2025-02-11 20:38:31 +0100 | <exarkun> | if lsp has managed to parse the buffer including it, you can |
| 2025-02-11 20:38:25 +0100 | <euouae> | ah, `s` is nice |
| 2025-02-11 20:38:12 +0100 | <euouae> | I can't e.g. do it on `each` if it sits by itself in a random line |
| 2025-02-11 20:38:10 +0100 | <mauke> | euouae: on many hackage pages you can hit 's' for an interactive search |
| 2025-02-11 20:38:10 +0100 | merijn | (~merijn@host-vr.cgnat-g.v4.dfn.nl) merijn |
| 2025-02-11 20:37:59 +0100 | <euouae> | hmm yeah but first I have to write a valid expression using the identifier |
| 2025-02-11 20:37:40 +0100 | <exarkun> | C-c l h h |
| 2025-02-11 20:37:19 +0100 | <exarkun> | if you have lsp you can get some |
| 2025-02-11 20:37:10 +0100 | <euouae> | hm -- trying :doc gives me `try recompiling with --haddock`, maybe I need to enable it in the cabal file |
| 2025-02-11 20:36:42 +0100 | xdej | (~xdej@quatramaran.salle-s.org) (Ping timeout: 244 seconds) |
| 2025-02-11 20:36:34 +0100 | <euouae> | e.g. I want to look up `each` from the lens library |
| 2025-02-11 20:36:20 +0100 | <euouae> | is there a good way to look up documentation somehow? maybe from the repl or emacs? I end up searching with control+f on hackage pages |
| 2025-02-11 20:30:01 +0100 | JeremyB99 | (~JeremyB99@2607:ac80:407:7:faad:86de:6310:fa61) (Remote host closed the connection) |
| 2025-02-11 20:27:19 +0100 | merijn | (~merijn@host-vr.cgnat-g.v4.dfn.nl) (Ping timeout: 260 seconds) |
| 2025-02-11 20:26:35 +0100 | JeremyB99 | (~JeremyB99@2607:ac80:407:7:faad:86de:6310:fa61) |
| 2025-02-11 20:26:32 +0100 | <euouae> | right, makes sense |
| 2025-02-11 20:26:07 +0100 | <int-e> | No, it was just the unqualified version of the same thing |
| 2025-02-11 20:25:58 +0100 | <euouae> | I thought Endo might've been public but that base:...Internal thing might've been an optimization |
| 2025-02-11 20:25:46 +0100 | <euouae> | Yeah but you said Endo a, and I thought maybe my type was different |
| 2025-02-11 20:25:14 +0100 | <int-e> | euouae: note how my first answer was two modules that you can import it from, maybe you missed it |
| 2025-02-11 20:25:05 +0100 | <euouae> | I just went with what :t told me |
| 2025-02-11 20:24:55 +0100 | <euouae> | mauke, I didn't realize it was a public type! |
| 2025-02-11 20:24:46 +0100 | <euouae> | Oh, I see... |
| 2025-02-11 20:24:31 +0100 | <int-e> | (It's slightly unfortunate that ghci doesn't tell you where to import Endo from but that information is hard to track.) |
| 2025-02-11 20:24:09 +0100 | <mauke> | but Data.Monoid re-exports it |
| 2025-02-11 20:23:55 +0100 | <mauke> | it's just originally defined in GHC.Internal.Data.Semigroup.Internal |
| 2025-02-11 20:23:49 +0100 | <int-e> | But yeah I think those two questions are not connected. |
| 2025-02-11 20:23:41 +0100 | notzmv | (~umar@user/notzmv) (Ping timeout: 268 seconds) |
| 2025-02-11 20:23:26 +0100 | <mauke> | because https://hackage.haskell.org/package/base-4.21.0.0/docs/Data-Monoid.html#t:Endo is public |
| 2025-02-11 20:23:07 +0100 | <int-e> | I'd be surprised if it didn't |
| 2025-02-11 20:22:53 +0100 | <mauke> | wait, are you saying lens is using ghc internal modules? |
| 2025-02-11 20:22:48 +0100 | merijn | (~merijn@host-vr.cgnat-g.v4.dfn.nl) merijn |
| 2025-02-11 20:22:45 +0100 | <int-e> | GHC is the dominant implementation anyway. |
| 2025-02-11 20:22:16 +0100 | <euouae> | I understand that there are be optimizations in lens that are GHC specific, but I hadn't seen anything before, maybe this is my first exposure |
| 2025-02-11 20:21:40 +0100 | <euouae> | It's just that... I guess what I'm asking is, if you're a library writer (like lens here) and you dig into GHC internal modules, are you committing a sin? Perhaps GHC documentation says this is okay? |
| 2025-02-11 20:21:06 +0100 | <euouae> | I guess that's the optimization |
| 2025-02-11 20:20:59 +0100 | machinedgod | (~machinedg@d108-173-18-100.abhsia.telus.net) machinedgod |
| 2025-02-11 20:20:55 +0100 | <euouae> | Yeah this is a fold |
| 2025-02-11 20:20:43 +0100 | <int-e> | Endo [a] is usually used as a difference list, to avoid a quadratic slowdown when concatenating many lists |
| 2025-02-11 20:20:20 +0100 | <mauke> | you're allowed to depend on yourself |
| 2025-02-11 20:20:11 +0100 | <EvanR> | wherever it came from |
| 2025-02-11 20:20:09 +0100 | <EvanR> | well whatever is exported is by definition the public API |
| 2025-02-11 20:20:07 +0100 | <mauke> | euouae: ?? it's your own implementation details |
| 2025-02-11 20:20:00 +0100 | <euouae> | lol |
| 2025-02-11 20:19:42 +0100 | <EvanR> | and you never would have known about it if it wasn't for that pesky verbose output and their dog! |