| 5 May 2026 |
 Rutile | Yep, we do have sandbox enabled by default, but if sandbox isnt available they are run without iirc
Though it is possible to mark a test as requiring sandbox which will skip it if sandbox isnt available (be it darwin or ubuntu apparmor stuff) | 19:26:24 |
| 6 May 2026 |
 zoë (@blokyk) | in the lix codebase, is there anyway to get the Hash object that corresponds to a StorePath? | 09:31:50 |
 thubrecht | You can use path.hashPart() that gives you a base32 encoded string and reconstruct the Hash | 09:35:55 |
| zoë (@blokyk) | oh i'm so dumb i didn't notice the string constructor for Hash | 09:36:54 |
| zoë (@blokyk) | wait no it's private | 09:37:40 |
| thubrecht | parseAny | 09:39:52 |
| zoë (@blokyk) | alright Hash::parseNonSRIUnprefixed seems to do what i want? | 09:39:57 |
| thubrecht | Also yes | 09:40:06 |
| zoë (@blokyk) |
hash 'qwshrp27a7s6cp7l1ambcrmanlmdzg9v' has wrong length for hash type 'sha256'
:(
| 09:48:39 |
| zoë (@blokyk) | *
hash 'qwshrp27a7s6cp7l1ambcrmanlmdzg9v' has wrong length for hash type 'sha256'
i specifically want a sha256 :(
| 10:00:54 |
 flokli | The hash you see in front of a store path is not a sha256 digest | 10:21:58 |
| flokli | If you prefer looking at some rust code with maybe more comments, nix_compat has all that factored out. So you see how it is constructed, both for CA paths (FODs) or input-addressed paths: https://snix.dev/rustdoc/src/nix_compat/store_path/utils.rs.html#124-137 | 10:23:40 |
| flokli | There's only 20 bytes in a store path: https://snix.dev/rustdoc/nix_compat/store_path/struct.StorePath.html | 10:28:07 |
| flokli | (the previously linked file has the compress_hash impl too. | 10:28:26 |
 Sergei Zimmerman (xokdvium) | In reply to @blokyk:matrix.org
hash 'qwshrp27a7s6cp7l1ambcrmanlmdzg9v' has wrong length for hash type 'sha256'
i specifically want a sha256 :(
Do you want a narHash maybe? | 10:32:35 |
| Sergei Zimmerman (xokdvium) | Ie input address vs output hash | 10:33:14 |
| zoë (@blokyk) | i'm trying to improve the error message for builtins.path's hash mismatch error: https://git.lix.systems/lix-project/lix/src/commit/010d0629261248ad846f9f12b356053514e464eb/lix/libexpr/primops.cc#L1819 | 10:37:52 |
| zoë (@blokyk) | so this should be the value of outputHash if this was a normal dérivation | 10:38:25 |
| zoë (@blokyk) | ideally without having to re-hash the path a second time ^^; | 10:38:40 |
| Sergei Zimmerman (xokdvium) | You can do a queryPathInfo from the store and get the narHash that way. But there’s no way to get the flat hash without recomputing. Or maybe looking at the ca field in the path info? Not sure what the state of it is lix though | 10:53:38 |
|  mrspits4ever joined the room. | 20:06:17 |
 just1602 | Does git.lix.systems load on your side? It doesn't load here. | 21:56:44 |
 hexa | 
Download | 21:58:30 |
| hexa | cpu goes brrrrrrr | 21:58:33 |
| just1602 | :( | 21:58:46 |
| just1602 | LLM bots hammering again? | 21:58:52 |
 raitobezarius (DECT: 7248) | Yes | 22:00:23 |
| raitobezarius (DECT: 7248) | https://zulip.afnix.fr/#narrow/channel/4-infra/topic/Lix.20Forgejo.20is.20being.20hammered.20of.20Git.20operations/near/40315 | 22:00:29 |
| hexa | raitobezarius my friend, is this reaction? | 23:41:03 |
| hexa | how costly is reaction on the lix infra? | 23:41:18 |
