|
|
0b70948d08
|
feeds: unsubscribe from Byrne Hobart
|
2024-10-02 13:44:56 +00:00 |
|
|
|
dc4e86213d
|
firefox: bookmarks: remove the date/time fields
|
2024-10-02 13:31:49 +00:00 |
|
|
|
484a177e5c
|
firefox: add GitHub search bookmark
|
2024-10-02 13:10:20 +00:00 |
|
|
|
c8de203198
|
browserpass/sops-gpg-adapter: remove dead code "sane-secrets-unlock"
i don't give firefox access to ~/.ssh, so this wasn't doing anything
|
2024-10-02 12:32:16 +00:00 |
|
|
|
b532189b2f
|
overlays/preferences: remove overrides which are no longer needed/wanted
|
2024-10-02 12:28:52 +00:00 |
|
|
|
ff7d1639c9
|
pkgs: remove no-longer-needed clightning patch
|
2024-10-02 12:17:16 +00:00 |
|
|
|
5884ecde42
|
fix scripts/update to not fail because of linuxPackages
|
2024-10-02 12:08:44 +00:00 |
|
|
|
695e639fc9
|
git: add an alias for "git reset --hard HEAD"
|
2024-10-02 11:56:58 +00:00 |
|
|
|
cf33b0e244
|
rk818-charger: dont ship clk-rk808, rk808-regulator, rk8xx-spi since the mainline versions work fine
|
2024-10-02 11:51:07 +00:00 |
|
|
|
d0cd8ed691
|
moby: kernel: build the in-tree MFD_RK8XX, just plumb it so nixos doesnt complain about conflicting module suppliers
|
2024-10-02 11:51:07 +00:00 |
|
|
|
16d31669d1
|
yt-dlp: add a bandwidth limit when considering the preferred stream
|
2024-10-02 10:18:30 +00:00 |
|
|
|
69ae2df933
|
moby: add rk818_{battery,charger} to boot.kernelModules
with this they're loaded on boot and now i actually do get capacity info
|
2024-10-02 08:22:54 +00:00 |
|
|
|
80b3016363
|
moby: deploy my rk818-charger kernel module
it needs to be manually probed, but after that i *do* get current measurement, just no capacity
|
2024-10-02 01:29:04 +00:00 |
|
|
|
881aec85c4
|
rk818-charger: add clk-rk808 and rk808-regulator modules, unchanged from upstream
i'm not a fan of this approach: i should just build the mainline kernel all modules, then remove the ones i want to replace
|
2024-10-02 01:04:33 +00:00 |
|
|
|
936b197d03
|
moby: fix kernel/modules to be in a bootable state
probably i actually can enable myt rk818-charger now that i have the right modules in initrd, but this commit serves as a known-good checkpoint
|
2024-10-02 00:44:25 +00:00 |
|
|
|
4ff0bf7cde
|
pkgs: stop patching helix, pasta, firefox-unwrapped at the global scope
these patches could be re-applied if needed, via `sane.programs`.
if i remove my other patches, then i can migrate `pkgs/additional/` to
just `pkgs/`
|
2024-10-02 00:24:40 +00:00 |
|
|
|
99285c5911
|
picocom: fix cross compilation
|
2024-10-02 00:22:01 +00:00 |
|
|
|
34f0bc529b
|
programs: ship picocom
|
2024-10-02 00:01:34 +00:00 |
|
|
|
2b85563b65
|
rk818-charger: import rk8xx-spi from mainline
|
2024-10-01 23:30:46 +00:00 |
|
|
|
d00e26c044
|
plumb rk818-charger module into the moby config
|
2024-10-01 15:08:56 +00:00 |
|
|
|
6f38c7847f
|
SPECULATIVE: moby: build kernel with MFD_RK8XXX_* as a module instead of as builtin
|
2024-10-01 13:06:10 +00:00 |
|
|
|
a641dc6e39
|
rk818-charger: update rk8xx-{core,i2c}.c with patches from megi's kernel tree that add compat to rk818
|
2024-10-01 12:59:24 +00:00 |
|
|
|
2a4cbcf327
|
rk818-charger: import rk8xx-{core,i2c}.c from mainline (6.11) linux
|
2024-10-01 12:44:51 +00:00 |
|
|
|
dbd914b58f
|
rk818-charger: init at 0.1 (WIP)
this doesn't build yet.
i'll also need to import the patched mfd driver.
|
2024-10-01 12:29:00 +00:00 |
|
|
|
caabf1421b
|
bt/wifi: switch back to sane.fs; fix so WiFi creds are available before NetworkManager starts
|
2024-10-01 11:21:45 +00:00 |
|
|
|
b56aae444d
|
MASSIVELY reduce eval time for cross-compiled hosts (i.e. moby)
moby eval time is reduced from ~3m to 17s
|
2024-10-01 09:54:33 +00:00 |
|
|
|
31aeaf8f35
|
modules/programs: remove programs whose packages no longer exist
|
2024-10-01 09:49:16 +00:00 |
|
|
|
8795212c67
|
programs: discord: lift to own file
|
2024-10-01 05:45:28 +00:00 |
|
|
|
c7abda9393
|
impure.nix: add new pseudohosts: baseline-{aarch64,x86_64}
immediately apparent is that the config takes 4x as long to eval on cross than on native, regardless of anything else
|
2024-10-01 04:38:50 +00:00 |
|
|
|
044bf8b783
|
refactor: remove dead modules/lib/fs.nix file
|
2024-10-01 03:55:29 +00:00 |
|
|
|
61df81291b
|
refactor: optimize eval time
lifting `let` bindings up where possible helps reduce the number of thunks nix has to allocate. this patch only does that by 0.3%-ish, though
|
2024-10-01 03:54:44 +00:00 |
|
|
|
80c67caf19
|
modules/fs: remove unused symlink.targetName option
|
2024-09-30 15:18:32 +00:00 |
|
|
|
1eea81c4ff
|
refactor: sane.fs: lift acl up to the toplevel; drop generated options
|
2024-09-30 15:15:30 +00:00 |
|
|
|
97d38aecab
|
feeds: subscribe to timclicks.dev
|
2024-09-30 15:00:45 +00:00 |
|
|
|
3114b56f41
|
yt-dlp: fix video sort argument
|
2024-09-30 13:55:48 +00:00 |
|
|
|
894e2ad13a
|
ntfy-sh: dont ship the desktop component either
|
2024-09-30 11:44:29 +00:00 |
|
|
|
cf4f232b30
|
goaccess: disable
|
2024-09-30 11:34:43 +00:00 |
|
|
|
a568f1273c
|
ntfy-sh: disable
|
2024-09-30 11:34:34 +00:00 |
|
|
|
e88e693026
|
komga: disable
|
2024-09-30 11:14:35 +00:00 |
|
|
|
01363bf2e0
|
rsync: move from console packages to sysadmin package
|
2024-09-30 10:40:22 +00:00 |
|
|
|
e29842aa9d
|
sane.fs: simplify
|
2024-09-30 10:19:39 +00:00 |
|
|
|
50c52683ff
|
sane.fs: remove unused mount.mountConfig, mount.unitConfig options
|
2024-09-30 10:19:39 +00:00 |
|
|
|
ca85dac4ac
|
sane.fs: make bind a required sub-option of mount
|
2024-09-30 10:19:39 +00:00 |
|
|
|
e52f57f5a2
|
sane.fs: remove unused mount.depends
|
2024-09-30 10:19:39 +00:00 |
|
|
|
49b5da6385
|
sane.fs: dont generate systemd services for every file/dir/symlink
that's handled by systemd-tmpfiles now
|
2024-09-30 10:19:39 +00:00 |
|
|
|
572dd5854d
|
WIP: sane.fs: remove wantedBy/wantedBeforeBy options
|
2024-09-30 10:19:39 +00:00 |
|
|
|
0c270fe4a3
|
WIP: sane.fs consumers: avoid wantedBy/wantedBeforeBy
|
2024-09-30 10:19:39 +00:00 |
|
|
|
fb85b62d04
|
prosody: populate /etc/* using environment.etc
|
2024-09-30 10:19:39 +00:00 |
|
|
|
48c81610a5
|
sane.fs: remove public access to the "unit" fields
fs entries soon won't correspond to systemd units, and hence that option's a bit nonsensical
|
2024-09-30 09:10:40 +00:00 |
|
|
|
e7cf14cc4c
|
/var/lib/systemd/backlight: persist using ordinary symlink-based approach
it works, at least on lappy
|
2024-09-30 05:44:42 +00:00 |
|
