| 13 Oct 2025 |
 K900 | This is a special case for paths relative and fully contained in flake root | 10:02:48 |
| K900 | I still have no idea why | 10:02:55 |
 MangoIV | oh, i do not wish to partake in that discussion, I just want to explain to myself why I cannot use this nix code with lix that I can use with upstream nix. | 10:02:56 |
 raitobezarius | it's helpful for us to decide what to do | 10:03:16 |
| raitobezarius | it's not a debate :) | 10:03:18 |
| MangoIV | probably usability issues? you don't have to update the lock file every time something in the repo changes | 10:03:23 |
| MangoIV | that is a really good usability feature, fwiw | 10:03:40 |
| K900 | The short answer is we didn't backport https://github.com/NixOS/nix/pull/10089 | 10:03:43 |
| raitobezarius | it makes sense to me to say that relative paths to a locked node are locked | 10:03:43 |
| K900 | I don't see a use case where a subpath is actually meaningfully useful as its own flake input unless you're doing weird subflake stuff or submodules | 10:04:36 |
| K900 | And you shouldn't do those things | 10:04:46 |
| K900 | And I don't even know if this works with submodules | 10:04:56 |
| K900 | I'm pretty sure this is just some lazy trees nonsense that's three steps removed from the actual rationale | 10:05:14 |
 aloisw | To avoid https://git.lix.systems/lix-project/lix/issues/586 I guess. | 10:05:33 |
| MangoIV | our specific use case is a tutorial. there are two sub directories that each are a separate tutorial and the top level flake import the other flakes to allow checking that they still work | 10:05:59 |
| raitobezarius | the change above mentioned do change evaluation semantics of the outputs of flakes though | 10:07:00 |
| raitobezarius | that massively sucks | 10:07:15 |
| aloisw | It was extremely broken before though. | 10:07:36 |
| raitobezarius | you mean subdirs as flake inputs or dependencies of the outputs fields in other values of the flake? | 10:08:12 |
| aloisw | Subdirectories as flake inputs. | 10:08:57 |
| raitobezarius | yeah my remark is not a major blocker | 10:09:55 |
| raitobezarius | but that change is not a trivial thing to port seems like | 10:10:01 |
| raitobezarius | https://git.lix.systems/lix-project/lix/issues/1012 | 10:13:19 |
| raitobezarius | thanks aloisw for helping me to clarify the core feature | 10:13:34 |
| MangoIV | So the answer is
„This is a good change, semantics wise, but poses some technical difficulties and is probably hard to port“? | 10:13:38 |
| raitobezarius | yeah, ideally we should avoid the shortcomings of cppnix's change | 10:13:56 |
| MangoIV | I think this is a good answer. For now a workaround is just using the cppnix evaluator in a shell. Maybe you can add that to the issue.
Thank you! | 10:15:20 |
| raitobezarius | added the workaround | 10:16:01 |
|  MoonstoneHearts changed their display name from tarynnya to MoonstoneHearts. | 10:17:36 |
 Sofie 🏳️⚧️ (she/her) | 
Download photo.jpg | 10:37:14 |
