| 2 Aug 2025 |
 emily | In reply to @elvishjerricco:matrix.org
Well my point was that nested automounts don't work. The need for the child dir to exist forces the parent to be actually mounted it could automount both partitions when you want /boot/efi is my point. (I know Lennart doesn't like this because what if you only need one of them but like, c'mon. like hooray I can mount one FAT32 instead of two if and only if I'm updating systemd-boot but not kernel/initrd or vice versa. big win) | 11:56:06 |
| emily | I do agree /efi is sensible I just think it is a bit mountain out of a molehill. | 11:56:25 |
 ElvishJerricco |
it could automount both partitions when you want /boot/efi is my point.
I don't understand what you mean by this. If you make two .automount units where one is a child of the other, the parent will just literally always be mounted and it won't behave like an automount. So no, you can't automount both
| 11:57:42 |
 aloisw | In reply to @emilazy:matrix.org
it's actually as old as the initial recs around XBOOTLDR on UEFI That still means that the recommendation has changed since XBOOTLDR was a later addition. | 11:58:11 |
| emily | In reply to @aloisw:julia0815.de
That still means that the recommendation has changed since XBOOTLDR was a later addition. I'm not sure people were regularly mounting ESP on /boot before at least gummiboot? /boot/efi was pretty ubiquitous for GRUB and so on | 11:59:29 |
| emily | but I can believe there were some months between the two :) | 12:00:01 |
| emily | oh what | 12:00:24 |
| emily | https://0pointer.net/blog/linux-boot-partitions.html is from 2022 | 12:00:27 |
| emily | ok | 12:00:29 |
| emily | I swear I read a very similar post way back | 12:00:37 |
| emily | maybe Lennart just repeats himself a lot | 12:00:45 |
| emily | but I may be totally wrong about the timeline here instead too | 12:01:00 |
 K900 | @emily OK so I have a branch called calamares6 | 16:41:58 |
| K900 | That at least runs | 16:42:01 |
| K900 | And needs more fixes in calamares-nixos-extensions | 16:42:09 |
| K900 | But I dropped SO MUCH SHIT | 16:42:30 |
| K900 | Already | 16:42:32 |
| emily | glad we finally have a graphical installer maintainer : ) | 17:00:49 |
| K900 | Fuck off | 17:01:29 |
| K900 | I will literally die | 17:01:40 |
| 3 Aug 2025 |
| K900 | https://github.com/NixOS/nixpkgs/pull/430636 | 13:21:50 |
| K900 | Someone please test this | 13:21:52 |
| K900 | I have calamared | 13:21:57 |
| emily | can you pls do the vendoring with a merge commit | 13:29:30 |
| emily | to preserve history and attribution | 13:29:37 |
| emily | (I am not sure I will have time to test properly today sadly) | 13:30:22 |
| emily | we also definitely cannot drop LICENSES/ | 13:30:41 |
| emily | and we should preserve the pytest stuff unless it's completely irrelevant now | 13:31:38 |
| emily | can't really review the extensions without them being on top of a merge commit so I can see the diff but seeing the patches go away makes me happy | 13:36:52 |
| K900 | I can do it later | 13:37:18 |
