| 20 Jun 2025 |
emily | so it is possible that it would require extensive patching that would be a maintenance burden for updates, in the case that upstream is not interested in changing that | 22:52:30 |
John Ericson | @emilazy:matrix.org so I view this push as kind of trying to get C libraries for FFI to work | 23:01:19 |
John Ericson | Like make us the go-to way to build the native deps for python packages or whatever | 23:02:13 |
John Ericson | Or static executables | 23:02:46 |
John Ericson | If whole systems a la NixOS fail to relocate, that's not ideal but it doesn't need to be fatal for that use-case | 23:17:33 |
| 22 Jun 2025 |
tobim | Redacted or Malformed Event | 05:05:32 |
tobim | Redacted or Malformed Event | 05:05:33 |
John Ericson | In reply to @emilazy:matrix.org IIRC #!/usr/bin/env -S is not portable wrt macOS so may not be a good solution for shebangs. https://ss64.com/mac/env.html actually the macOS env looks more powerful | 05:57:26 |
John Ericson | Any chance we can get these features in coreutils? | 05:57:44 |
John Ericson | Then we don't need a shim exe, because we can just expand an environment variable | 05:58:07 |
p14 | emily: Tristan Ross: Can we entertain merging [0] as-is since it fixes a regression? The question of providing a "fake libgcc" is something requiring further development/enablement. https://github.com/NixOS/nixpkgs/pull/417354 | 10:44:19 |
p14 | * emily: Tristan Ross: Can we entertain merging [0] as-is since it fixes a regression? The question of providing a "fake libgcc" is something requiring further development/enablement. [0] https://github.com/NixOS/nixpkgs/pull/417354 | 10:44:25 |
p14 | * emily: Tristan Ross: Can we entertain merging [0] as-is since it fixes a regression? The question of providing a "fake static libgcc" is something requiring further development/enablement. [0] https://github.com/NixOS/nixpkgs/pull/417354 | 10:44:32 |
emily | shebangs are parsed differently though | 11:22:11 |
emily | I forget the exact differences | 11:22:16 |