| 20 Apr 2025 |
 aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | Yeah, but how do we handle this as a general case, and also for config.allowVariants | 15:35:24 |
 emily | also the date-based system is silly, since stuff gets merged later than it gets written | 15:35:24 |
| emily | it should be based on releases instead | 15:35:27 |
| emily | splitting out deprecations and preserved aliases makes sense. I think we don't need to separate out removed things | 15:36:15 |
| emily | it could just look like deprecatedPkg = { release = "25.05"; alias = newName; }; or deprecatedPkg = { release = "25.05"; }; for a removed thing | 15:36:56 |
| emily | which would also make automating dropping stuff easier, there's currently a sorta gross script that doesn't handle all the cases | 15:37:08 |
 Tristan Ross | Yeah, something which automagically handles deprecations is better | 15:37:33 |
| emily | other scopes are where it gets awkward | 15:37:36 |
| emily | probably we'd want a thing to process these in lib so that it could be reused across the tree | 15:37:45 |
| Tristan Ross | Also, should have a suggestion in case something was removed. | 15:37:46 |
| emily | right, you could have a field for an additional message | 15:37:59 |
| emily | like reason = "known vulnerabilities and unmaintained for years"; suggestion = …; | 15:38:21 |
| emily | rather than manually writing throws and putting dates in comments | 15:38:36 |
| Tristan Ross | I feel like some of the reasons could be defined so it didn't have to be copied and pasted. | 15:39:02 |
| emily | aliases.nix could contain stuff that's meant to stay (the "preserve" stuff) and deprecated.nix could contain stuff on the chopping block | 15:39:04 |
| Tristan Ross | Yeah, aliases could hold things like aliases to latest versions and a sort of legacy-compatibility | 15:39:36 |
| emily | yeah there's mostly like 4 common reasons that get reused a lot :) | 15:39:42 |
| Tristan Ross | I can think unmaintained and vulnerable are the two most common | 15:40:08 |
| Tristan Ross | If we make it a list then we could have: reasons = [ unmaintained vulnerable ] | 15:40:35 |
| Tristan Ross | It could throw an error that looks something like this:
Package "xyz" was removed since xx.xx due to the following reasons:
- Unmaintained: package has not been updated or maintained for an extended period of time
- Vulnerable: package contains security vulnerability and should not be trusted
| 15:42:25 |
| Tristan Ross | It's a pretty clear error | 15:42:46 |
| aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | Well the real question I'd like to ask is how do we tell users the pkgs variant cannot be used is because they didn't enable config.allowVariants | 15:44:53 |
| aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | Other wise it feels like "I sometimes can use them, sometimes it just isn't there but I don't know what is exactly going wrong" | 15:45:36 |
| Tristan Ross | For now, I made it to where you have to explicitly disable it. | 15:46:32 |
| aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | In reply to @rosscomputerguy:matrix.org
For now, I made it to where you have to explicitly disable it. Yeah, but if people trigger ofborg errors, they should get an explanation in place | 15:47:14 |
| aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | The aliases are already confusing but we can make them correct | 15:47:41 |
| Tristan Ross | So what we can do is "fake-import" it | 15:48:03 |
| Tristan Ross | I tested in repl and got this:
nix-repl> builtins.attrNames (import ./pkgs/top-level/variants.nix { nixpkgsFun = import ./.; lib = pkgs.lib; stdenv = pkgs.stdenv; overlays = pkgs.overlays; makeMuslParsedPlatform = null; } pkgs pkgs)
[
"pkgsArocc"
"pkgsExtraHardening"
"pkgsLLVM"
"pkgsLLVMLibc"
"pkgsMusl"
"pkgsRocm"
"pkgsZig"
]
| 15:48:46 |
| Tristan Ross | With this, we can map the names and throw an error saying that config.allowVariants need to be enabled. | 15:49:18 |
| aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | Can we use some = allow: builtins.mapAttrs (k: v: if allow then v else throw "not allowed") { a = 1; } | 15:52:04 |
