| 6 Jun 2021 |
 maralorn | Maybe there is stupid mistake in there. | 15:02:33 |
 pwmosquito | Hey, so I wanted to try out https://github.com/kowainik/stan on a work codebase and was going to add it to my nix shell and noticed that it's not in unstable but it is in 21.05:
https://search.nixos.org/packages?channel=21.05query=haskellPackages.stan
https://search.nixos.org/packages?channel=unstable&query=haskellPackages.stan
any idea why? | 15:45:15 |
| pwmosquito | * Hey, so I wanted to try out https://github.com/kowainik/stan on a work codebase and was going to add it to my nix shell and noticed that it's not in unstable but it is in 21.05:
https://search.nixos.org/packages?channel=21.05&query=haskellPackages.stan
https://search.nixos.org/packages?channel=unstable&query=haskellPackages.stan
any idea why? | 15:46:57 |
| maralorn | pwmosquito: I have no clue, what the search is doing. | 15:47:08 |
| maralorn | stan works for me. | 15:47:13 |
| maralorn | on unstable. | 15:47:17 |
| maralorn | Can you just try to install it? | 15:47:29 |
| pwmosquito | ha! let me try again | 15:47:31 |
| pwmosquito | i mean i actually went berzerk and nixified it 😅
https://github.com/pwm/stan/commit/afc048fae34c899530ddc57a1651d101b7060c27
but i'm not sure if it'd be useful to do an upstream PR if it's in nixpkgs already and kowainik iirc don't use nix | 15:48:34 |
| maralorn | Hm, yeah I think nix files in upstream repos only make sense if the maintainers care about it. | 15:49:29 |
| pwmosquito | and yep, it works ¯_(ツ)_/¯ | 15:51:22 |
| pwmosquito | tnx maralorn | 15:51:28 |
| maralorn | sterni (he/him): While we are talking about contribution guidelines: Do you have a system when you write haskellPackages and when haskell.packages? | 16:06:03 |
 sterni (he/him) | I usually do haskellPackages.<package> since it is shorter than haskell.packages.*.<package> | 16:06:50 |
| sterni (he/him) | so when I do something that affects the set(s) in general I do haskell.packages | 16:07:03 |
| maralorn | Interesting. | 16:07:18 |
| sterni (he/him) | but probably always using haskellPackages is better since it is more searchable | 16:07:22 |
| maralorn | Yeah. | 16:07:34 |
| sterni (he/him) | think would make more sense if we were to make it a policy right | 16:08:14 |
| maralorn | I am wondering a bit if the autocommits by our scripts are actually conforming to the guidelines. | 16:08:49 |
| maralorn | Stackage Nightly: date -> date, kinda does not. | 16:09:08 |
| sterni (he/him) | lets clean that up in the next iteration maybe | 16:09:23 |
| sterni (he/him) | and update HACKING.md | 16:09:28 |
| maralorn | But that’s mainly because there is no package this belongs to. | 16:09:45 |
| sterni (he/him) | also i'd like to prescribe a PR title so it is easier to search for the current PR | 16:09:45 |
| sterni (he/him) | since /pulls/peti doesn't work anymore… | 16:09:52 |
| maralorn | good idea. | 16:09:58 |
| sterni (he/him) | In reply to @maralorn:maralorn.de
But that’s mainly because there is no package this belongs to. haskellPackages: update stackage date -> date | 16:10:15 |
| sterni (he/him) | something like that? | 16:10:20 |
| maralorn | sterni (he/him): It could quickly become to long. the word nightly should be in there. | 16:10:43 |
