| 9 Aug 2023 |
 John Ericson | Warnings from forcing the old attribute? | 02:07:52 |
| John Ericson | Fair enough, if so | 02:08:09 |
 infinisil | Yeah nix repl at least uses drvPath to print, I imagine drvPath is used in other parts of Nix too | 02:08:54 |
| John Ericson | I guess it would be nice to say we can make use of this builtin irrespective of any deprecation policy: we can make a nicer way to keep this weird expressive power | 02:09:24 |
| John Ericson | The discard way is quasi unsafe because it "weakens" the string context throwing stuff away | 02:09:46 |
| infinisil | It's very late here, can you write a concrete proposal for Nixpkgs into some issue? We can take another look then | 02:09:57 |
| John Ericson | The one I made "strengthens" the string context so I think we can just set aside `drvPath` an hope this is good primop which we can e.g. make a Nixpkgs lib function from which would hide some of the details | 02:10:42 |
| John Ericson | Yeah | 02:10:47 |
| John Ericson | Indeed it is late | 02:11:00 |
| infinisil | I didn't bother with a PR, but I updated the NAT GitHub profile with information regarding the new member-exclusive Matrix channel used to review team applications | 23:01:00 |
| infinisil | * I didn't bother with a PR, but I updated the NAT GitHub profile with information regarding the new member-exclusive Matrix channel used to review team applications: https://github.com/nixpkgs-architecture/.github/commit/c8eb7b0c34bf4faa9a6cb1b06408b55bb9c32c90 | 23:01:14 |
| infinisil | That was mentioned in the last meeting notes as well. I will make sure that we really only use that private channel for things that can't be shared publicly. | 23:05:54 |
