| 1 Jun 2025 |
|  Ben Sparks joined the room. | 09:06:43 |
|  Sean Thawe joined the room. | 23:48:46 |
| 5 Jun 2025 |
 John Ericson | https://github.com/NixOS/nixpkgs/pull/414321 Tristan Ross wanna review? | 19:50:58 |
 Tristan Ross | Why is it still called libcCross? Shouldn't we move that to just libc and have it apply even when not cross? | 19:52:44 |
| John Ericson | done! | 20:46:52 |
| John Ericson | thanks for encouraging the scope creep :) | 20:47:01 |
| John Ericson | fixing it was easier than I thought! | 20:47:07 |
| John Ericson | glad this old tech debt is finally gone | 20:47:12 |
| John Ericson | once we get rid of the old gcc | 20:47:18 |
| John Ericson | I think we can finish the job of having nothing with "cross" in its name | 20:47:34 |
| Tristan Ross | Yeah, there's a few things I think that could be done in that PR lol. | 20:47:39 |
| John Ericson | crossLibcStdenv -> libcStdenv | 20:47:39 |
| Tristan Ross | Yeah, ideally you should be able to refer to toplevel attributes for the different components | 20:47:59 |
| Tristan Ross | The idea is then you can utilize whatever you need without going through the CC wrapper. | 20:48:15 |
| Tristan Ross | Like for example, if something wants to link to the libc without knowing the CC then it just does pkgs.libc | 20:48:45 |
| John Ericson | yeah | 20:49:34 |
| John Ericson | harder for tools as we discussed before | 20:49:56 |
| John Ericson | but for libs there is no problem | 20:50:05 |
| John Ericson | we should just do it! | 20:50:09 |
| 6 Jun 2025 |
|  @creepinson:matrix.org left the room. | 00:18:43 |
 Randy Eckenrode | pkgs.libc does nothing on Darwin. The libc is in the SDK. It needs to be to allow the libc to be overridden. The only way I can see breaking it out is to completely remove the link from cc-wrapper to libc and rely exclusively on libc being a package in the inputs (and I still expect some things to do the wrong thing). | 13:56:14 |
| Randy Eckenrode | * pkgs.libc does nothing on Darwin. The libc is in the SDK. It needs to be there to allow the libc to be overridden. The only way I can see breaking it out is to completely remove the link from cc-wrapper to libc and rely exclusively on libc being a package in the inputs (and I still expect some things to do the wrong thing). | 13:56:22 |
| Randy Eckenrode | * pkgs.libc does nothing on Darwin. The libc is in the SDK. It needs to be there to allow the SDK to be overridden. The only way I can see breaking it out is to completely remove the link from cc-wrapper to libc and rely exclusively on libc being a package in the inputs (and I still expect some things to do the wrong thing). | 13:56:26 |
| Randy Eckenrode | If you need to refer to libc, the correct paths are /usr/include and /usr/lib. Darwin tooling uses a sysroot for the SDK because that’s how native tooling works (and several languages expect tooling to work like native tooling). | 13:57:51 |
| Randy Eckenrode | (Considering that you can’t meaningfully change the libc on Darwin, it probably doesn’t matter what pkgs.libc is because anything other than the default one is wrong.) | 13:58:52 |
| Randy Eckenrode | * (Considering that you can’t meaningfully change the libc on Darwin, it probably doesn’t matter what pkgs.libc is because anything other than libSystem is wrong.) | 14:01:18 |
| 7 Jun 2025 |
|  matrixrooms.info mod bot (does NOT read/send messages and/or invites; used for checking reported rooms) left the room. | 22:12:41 |
| matrixrooms.info mod bot (does NOT read/send messages and/or invites; used for checking reported rooms) joined the room. | 22:38:33 |
| 15 Jun 2025 |
|  @t4ccer:matrix.org left the room. | 19:07:10 |
| 16 Jun 2025 |
 p14 | I'm guessing noBrokenSymlinks broke pkgsStatic.pkgsLLVM.llvmPackages.libunwind:
ERROR: noBrokenSymlinks: the symlink /nix/store/l44d59cxpyc7cw4kg99xr0ni375g1ikd-libunwind-static-x86_64-unknown-linux-musl-19.1.7/lib/libgcc_s.so points to a missing target: /nix/s…
:(
(cc Tristan Ross ) | 18:26:16 |
