| 20 Nov 2025 |
 John Ericson | I wanted you all to take my libexpr -> libflakes separation long ago :) | 20:02:57 |
| John Ericson | * OK I guess I did not announce this very clearly :(, whops! | 20:05:39 |
 just1602 | I'm really not great at C++, but I'd be really interested to see this libexpr -> libflakes project | 20:52:03 |
 raitobezarius | Thank you for the suggestion but we probably won't take it, it doesn't achieve what we want | 21:11:27 |
| raitobezarius | It's not really the important piece of the Flakes extraction project, just moving code around and exposing APIs to register certain operations that are related to Flakes in the evaluation machinery (mostly around NIX_PATH) | 21:12:24 |
| raitobezarius | I have a tree that already finished that part, more difficult parts involves libcmd and generalizing installables | 21:12:49 |
 edef | anything NIX_PATH-y should probably be explicitly passed in, rather than being grabbed from env by default | 23:55:46 |
| edef | like, objcap-y pure-by-default rather than "here's a flag to turn off the default impure behaviour" | 23:56:17 |
| 21 Nov 2025 |
| raitobezarius | Doing that properly would require rewrapping a bunch of impure calls and rewiring an objcap system, I want to get there | 01:01:43 |
| raitobezarius | But that's mostly orthogonal to how Flakes processes the search path entries | 01:02:12 |
