Newest at the top
| 2024-04-19 03:44:46 +0200 | otto_s | (~user@p5de2fb1c.dip0.t-ipconnect.de) |
| 2024-04-19 03:42:59 +0200 | otto_s | (~user@p4ff271d1.dip0.t-ipconnect.de) (Ping timeout: 256 seconds) |
| 2024-04-19 03:27:34 +0200 | <tcard> | Thanks! |
| 2024-04-19 03:26:07 +0200 | <sm> | tcard: aha. Here's such a build, for example/testing: https://github.com/simonmichael/hledger/actions/runs/8744554461 , it used https://downloads.haskell.org/ghcup/unofficial-bindists/ghc/9.8.2/ghc-9.8.2-x86_64-alpine3_12-linu… |
| 2024-04-19 03:25:19 +0200 | <masaeedu> | the key is in what you require of `f` |
| 2024-04-19 03:25:00 +0200 | <masaeedu> | probie: there's no formal content here to "obvious", but if you'd like me to be pedantic i can distinguish your function from what is intended quite easily |
| 2024-04-19 03:22:27 +0200 | <monochrom> | In fact, I bet not an obvious semantics either. |
| 2024-04-19 03:22:02 +0200 | madeleine-sydney | (~madeleine@c-76-155-235-153.hsd1.co.comcast.net) |
| 2024-04-19 03:22:00 +0200 | <probie> | However, I wouldn't say that there is "an obvious implementation of `(a -> Maybe b) -> [Tree a] -> [Tree b]`". You're probably unhappy with something along the lines of `\f -> map (:< []) . catMaybes . map f . flatten` |
| 2024-04-19 03:20:57 +0200 | xff0x | (~xff0x@2405:6580:b080:900:746b:6d2b:7905:31b8) (Ping timeout: 268 seconds) |
| 2024-04-19 03:19:40 +0200 | <masaeedu> | monochrom: `head` is not a total function |
| 2024-04-19 03:19:18 +0200 | waldo | (~waldo@user/waldo) (Quit: waldo) |
| 2024-04-19 03:19:13 +0200 | <c_wraith> | the real fun is when you turn the whole mess into the crazy forest thing used by priority queues optimized for Dijkstra's algorithm. |
| 2024-04-19 03:18:15 +0200 | <probie> | I think the suggested semantics are that when a node is deleted, then all its children become its parent's children |
| 2024-04-19 03:17:51 +0200 | <monochrom> | OK I see. Then it is not weird at all. |
| 2024-04-19 03:17:43 +0200 | <c_wraith> | the forest version usually results in a forest with more trees than the input |
| 2024-04-19 03:17:23 +0200 | <c_wraith> | that's only true if you enjoy throwing out nodes that the predicate accepted |
| 2024-04-19 03:16:56 +0200 | <monochrom> | OK more accurately singleTreeVersion pred = head . forestVersion pred . (: []) |
| 2024-04-19 03:16:23 +0200 | <monochrom> | Because singleTreeVersion = head . forestVersion . (: []) |
| 2024-04-19 03:15:52 +0200 | <monochrom> | No, I would think the forest version is even more problematic than the single tree version. |
| 2024-04-19 03:14:45 +0200 | <masaeedu> | but assuming `Tree` is e.g. `Cofree []`, (this works for anything Filterable), there is an obvious implementation of `(a -> Maybe b) -> [Tree a] -> [Tree b]` |
| 2024-04-19 03:13:25 +0200 | <c_wraith> | jackdk: if you delete a node with two children, it's not obvious how to reconstruct a tree out of it. But it is obvious how to turn the remainder into a forest. |
| 2024-04-19 03:13:23 +0200 | <masaeedu> | well, assuming `Tree` represents a tree with internal nodes, there isn't an obvious unique way to implement `(a -> Maybe b) -> Tree a -> Tree b` |
| 2024-04-19 03:12:21 +0200 | <jackdk> | WDYM? |
| 2024-04-19 03:12:16 +0200 | <masaeedu> | s/don't// |
| 2024-04-19 03:12:01 +0200 | <masaeedu> | it's kind of weird how trees with internal nodes don't can't be sensibly filtered, but their forests can |
| 2024-04-19 03:07:16 +0200 | <tcard> | That is great news! I see that ghcup installs unofficial builds (from https://downloads.haskell.org/ghcup/unofficial-bindists/ghc/) for recent versions of GHC, but it still installs the broken official builds for older versions (such as 9.2.4). I will investigate further when I get a chance. |
| 2024-04-19 03:03:16 +0200 | tv | (~tv@user/tv) |
| 2024-04-19 03:01:00 +0200 | itscaleb | (~itscaleb@user/itscaleb) (away) |
| 2024-04-19 02:58:35 +0200 | tdammers | (~tdammers@41-138-178-143.ftth.glasoperator.nl) |
| 2024-04-19 02:56:49 +0200 | dehsou^ | (~cd@c-98-242-74-66.hsd1.ga.comcast.net) |
| 2024-04-19 02:50:26 +0200 | tv | (~tv@user/tv) (Ping timeout: 268 seconds) |
| 2024-04-19 02:48:51 +0200 | tzh | (~tzh@c-73-164-206-160.hsd1.or.comcast.net) |
| 2024-04-19 02:46:59 +0200 | <sm> | tcard: I am guessing maerwald's ghcup is providing a fixed ghc on alpine. It seems to work for me at least |
| 2024-04-19 02:46:47 +0200 | tdammers | (~tdammers@219-131-178-143.ftth.glasoperator.nl) (Ping timeout: 264 seconds) |
| 2024-04-19 02:38:19 +0200 | tok | (~user@user/tok) (Remote host closed the connection) |
| 2024-04-19 02:38:05 +0200 | kilolympus | (~kilolympu@31.205.200.246) (Read error: Connection reset by peer) |
| 2024-04-19 02:28:57 +0200 | masaeedu | (~masaeedu@user/masaeedu) |
| 2024-04-19 02:23:05 +0200 | masaeedu | (~masaeedu@user/masaeedu) (Read error: Connection reset by peer) |
| 2024-04-19 02:21:01 +0200 | n8n | (n8n@user/n8n) |
| 2024-04-19 02:20:48 +0200 | demon-cat | (~demon-cat@dund-15-b2-v4wan-169642-cust1347.vm6.cable.virginm.net) (Ping timeout: 252 seconds) |
| 2024-04-19 02:17:04 +0200 | n8n | (n8n@user/n8n) (Quit: WeeChat 4.2.2) |
| 2024-04-19 02:16:22 +0200 | demon-cat | (~demon-cat@dund-15-b2-v4wan-169642-cust1347.vm6.cable.virginm.net) |
| 2024-04-19 02:16:14 +0200 | yin | (~yin@user/zero) |
| 2024-04-19 02:16:08 +0200 | peterbecich | (~Thunderbi@47.229.123.186) |
| 2024-04-19 02:14:01 +0200 | vgtw | (~vgtw@user/vgtw) |
| 2024-04-19 02:13:37 +0200 | <tcard> | Official Alpine builds of GHC have unfortunately not worked for a while, though. I have not had time to work on this recently, but the issue I have bookmarked (https://gitlab.haskell.org/ghc/ghc/-/issues/22237) is still open. Olivier's project works great because it builds GHC from source, and I think he might be using the gold linker if I remember correctly. |
| 2024-04-19 02:08:14 +0200 | bitdex | (~bitdex@gateway/tor-sasl/bitdex) |
| 2024-04-19 02:07:28 +0200 | <tcard> | If building on Alpine, you can indeed create static executables directly. Stack's docker support just makes it easy to do so using an Apline container from distributions that are not musl-based. |
| 2024-04-19 02:07:02 +0200 | <sm> | jackdk: yes, that's why alpine is used |