The informative half of 'informative + rock-stable': both rebuild
wrappers capture /run/current-system before the switch and print a
package-level nvd diff after it — store-path-pinned so no PATH
dependence, || true so an ornamental diff can never fail a succeeded
rebuild, and an identical-path fast-path says 'no changes'. README §3
notes it. V1: template-system toplevel rebuilt, generated scripts
bash -n clean, nvd smoke-run against the real closure.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Recovers the crashed iteration #7 (items 21+23, Latitude QA findings):
- nomarchy-waybar supervisor (waybar.nix, exec-once'd from
hyprland.nix): any waybar exit respawns the bar; crash-loop guard
(5 fast exits -> critical notify + stop); TERM trapped for a real
stop. Fixes the bar-less session after a theme-switch crash.
- nomarchy-theme-sync prefers a clean `pkill -x waybar` (supervisor
restart with fresh config+style) over the crash-prone in-place
SIGUSR2 reload when the supervisor is running; SIGUSR2 stays as the
unsupervised fallback.
- sys-rebuild: snapshot-first system rebuild against the CURRENT lock
(no `nix flake update`) — the no-update twin of sys-update;
README §3/§5 + motd list it.
Verified: V0 re-run green after crash recovery (flake check --no-build
+ py_compile); V1/V2 per the crashed session's journal entry (HM
renders exec-once=nomarchy-waybar; headless software-GL VM:
SIGKILL -> new pid, clean kill -> respawn, SIGUSR2 -> alive).
V3 pending on the Latitude (queued in HARDWARE-QUEUE.md).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Re-diagnosis of the "btrfs-assistant 2.2 segfaults on launch" bug, and
the fix is not the planned nixpkgs override:
1. gdb: the crash is btrfs_util_subvolume_iterator_next() —
libbtrfsutil's UNPRIVILEGED subvolume-iteration path in btrfs-progs
6.17.1 (upstream-fixed after 6.17.1, kdave/btrfs-progs 886571653).
Symbol versions were checked and match: not ABI/link drift.
2. VM A/B on real btrfs: as root it works (exit 0); as a user it
segfaults (139). The pkexec launcher runs it as root — so the GUI
path was fine all along, except…
3. …the distro shipped NO polkit authentication agent: every pkexec
prompt in the session failed silently. That is the actual root
cause of "the GUI doesn't open".
Shipped: hyprpolkitagent (Hyprland's Qt agent — the prompt is
Stylix-themed) via exec-once in hyprland.nix; `nomarchy-menu snapshot`
prefers btrfs-assistant-launcher again, with the nomarchy-snapshots
fzf flow kept as fallback (and for SSH); checks.snapshot-gui guards
the root path + offscreen GUI event loop on a real btrfs volume so a
lock bump can't silently regress it. No btrfs-progs patch: all our
flows are root-side; the unprivileged fix arrives with a lock bump.
Verified: V0 (flake check) + V2 (snapshot-gui check, executed green) +
V1 (HM generation builds; hyprland.conf carries the agent exec-once;
generated menu passes bash -n with the launcher wiring). Remaining V3
queued: first on-hardware themed polkit prompt.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
modules/nixos/oom.nix — running out of memory now kills the offending
process (with a desktop notification via systembus-notify) instead of
freezing the desktop for minutes until the kernel OOM killer fires.
earlyoom over systemd-oomd, deliberately: oomd kills whole cgroups,
and a Hyprland session runs as ONE scope (nothing here spawns per-app
systemd scopes, unlike GNOME) — under pressure oomd would take out the
entire desktop to save it. earlyoom kills the single largest process
before the thrash point. nixpkgs default-enables oomd in an inert
state (no slices monitored); it is disabled outright so there is one
owner. Session plumbing is --avoid-listed (unanchored — NixOS wrappers
rename comm to .foo-wrapped); no --prefer tuning, largest-RSS
selection already finds the hog. All mkDefault; opt out with
services.earlyoom.enable = false (README note added).
Verified: V0 (flake check) + V2 — new checks.oom-protection
runNixOSTest, executed: a chunked allocator (686 MB peak, 1 GB VM) is
SIGTERM'd by earlyoom in 0.1 s, a bystander unit survives (the
process-level granularity the module exists for), and systemd-oomd is
asserted inactive. 25 s runtime.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Completes the remaining "Distro branding, round 2" work.
GRUB theme: hosts/live.nix builds nomarchyGrubTheme whose background is
the same composed splash as the isolinux/BIOS splash (accent logo on
base), with a palette-coloured boot menu and accent timeout bar. Derived
from nixos-grub2-theme only to reuse its bundled DejaVu .pf2; the stock
NixOS logo is dropped (ours is in the background). Wired as
isoImage.grubTheme so UEFI boot matches BIOS.
distroId: set system.nixos.distroId = "nomarchy" so os-release is honest
(ID=nomarchy, ID_LIKE=nixos). Safe — switch-to-configuration builds its
"is this NixOS?" guard from the configured distroId (and /etc/NIXOS is
the fallback), so rebuilds keep working. isNixos→false blanks the
nixos.org URLs, so extraOSReleaseArgs restores them to the project.
Adds checks.distro-id (runNixOSTest): boots distroId=nomarchy and runs
switch-to-configuration dry-activate green; os-release eval-verified from
the real downstream config; grub theme built + structure-verified.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The threshold was written boot-only, so a firmware that clears
charge_control_end_threshold on unplug would silently lose the cap
until the next reboot. Add a services.udev rule on the mains adapter
(ATTR{type}=="Mains", vendor-neutral) that restarts the oneshot via
systemctl --no-block on every AC state change. Uses restart (not
try-restart) so it re-applies even if the boot run was inactive.
Eval-verified: rule present when the charge limit is on, absent when
off; downstream-template-system still evaluates.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A built-in RGB+IR webcam (common on recent ThinkPads, e.g. the T14s AMD
Gen 4) exposes its IR face-unlock sensor as a second, identically-named
"Integrated Camera"; apps that pick it show a dark, 8-bit-greyscale
image. nomarchy.hardware.camera.hideIrSensor disables that node on the
V4L2 PipeWire path (WirePlumber monitor.v4l2.rules) so only the colour
camera is offered, matched by card name (overridable irMatch).
- libcamera is left untouched: an external camera you plug in is never
affected, and surgical internal-only libcamera scoping isn't possible
(its distinguishing props bind after the monitor rule runs).
- Only the PipeWire node is hidden; the kernel /dev/video* stays open,
so Howdy face-unlock still reads the IR sensor directly.
- The installer's hardware-db.sh auto-detects a paired RGB+IR webcam
from /sys/class/video4linux/*/name and bakes the toggle into
system.nix; commented example added to the downstream template.
Validated live on a T14s: collapses the four enumerated "Integrated
Camera" entries to one colour source; the exact shipped regex disables
the IR node (confirmed in the WirePlumber log) and leaves libcamera
intact.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
nomarchy.services.tailscale now makes the login user the Tailscale
operator (services.tailscale.extraSetFlags = ["--operator=<username>"],
mkDefault), so `tailscale up/down/set` — and the System → VPN menu's
Tailscale controls — run without sudo. The user is already in wheel, so
the operator grant is no real new privilege, just skips the prompt.
nomarchy-vpn's Tailscale actions are now operator-aware: a ts_priv
helper runs the subcommand inline and only falls back to a sudo terminal
if the operator grant is absent (downstream dropped it). Connect brings
an authed-but-down node up inline; a node that still needs interactive
login goes to a terminal so the auth URL is visible.
Option description + template comment updated; ROADMAP follow-up closed.
System + home builds green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A dedicated System → VPN flow (nomarchy-vpn) past the wifi picker:
- NetworkManager VPN/WireGuard connections shown ● active / ○ inactive,
toggled up/down on select (networkmanager-group users need no sudo).
- Import a *.conf / *.ovpn via the fd picker → nmcli connection import,
type chosen by extension. networkmanager-openvpn ships system-side
(networking.networkmanager.plugins, mkDefault) so the openvpn type is
available; WireGuard is native to NetworkManager.
- Tailscale block, self-gated on the CLI (= nomarchy.services.tailscale):
read-only status + up/down + exit-node; privileged actions go through a
sudo terminal (the snapshot pattern), no operator wiring.
Waybar custom/vpn shield (nomarchy-vpn-status): self-hides unless a NM
tunnel or Tailscale is up; @good tone; click opens the submenu. Wired
into the generated bar and both summer whole-swaps. Decided import-first;
creation deferred to nm-connection-editor. ROADMAP marked shipped.
Eval + home/system builds green; both scripts pass bash -n. Pending an
on-machine check (live nmcli import/connect + Tailscale paths).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
nomarchy.system.autoTimezone (off by default): geoclue + automatic-
timezoned drive /etc/localtime from your location, so travelling to
another zone updates the Waybar clock on its own.
Menu-driven, in-flake state (the night-light / keyboard philosophy):
the flag is settings.autoTimezone in theme-state.json (git-tracked); a
System-menu entry toggles it. Because it's a SYSTEM service — not a user
unit it can start/stop instantly like night-light — the toggle
(nomarchy-autotimezone, run in a terminal) writes the flag --no-switch
then drives `sudo nixos-rebuild` (bakes the service + the time.timeZone
override) plus `home-manager switch` (the Waybar-refresh watcher); both
sides read the one flag.
- modules/nixos/timezone.nix: parses settings.autoTimezone from
nomarchy.system.stateFile (like Plymouth), gates geoclue2 +
automatic-timezoned, and mkForce-nulls time.timeZone — automatic-
timezoned sets it null at normal priority, which collides with the
installer's static value (a hard eval error), so force wins and
reverts cleanly when disabled. Ships the nomarchy-autotimezone toggle.
- modules/home/timezone.nix: a tiny user service watching timedate1's
change signal that reloads Waybar (SIGUSR2) on a real zone change, so
the clock follows live (Waybar captures the zone at construction).
Gated on the same flag via nomarchy.settings.autoTimezone.
- rofi menu: "Auto timezone (on/off)" in the System submenu.
- option + template example + ROADMAP entry; theme.nix seeds the flag.
Eval-verified on (geoclue+daemon on, tz forced null over a static
installer value, watcher present) and off (static tz intact, nothing
extra); home + system both build green. Needs an on-hardware check
(geoclue detection + the SIGUSR2 clock refresh aren't CI-testable).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A `nomarchy-menu printers` entry opens system-config-printer (the CUPS
admin GUI), mirroring the bluetooth/blueman pattern: the printing service
ships the package via environment.systemPackages and the menu execs it,
self-gated on the system-config-printer binary so the entry appears only
when nomarchy.services.printing is on. No direct keybind (rare one-off).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Two menu improvements.
Back everywhere: every list menu now ends with a "↩ Back" entry that
returns one level up — power, theme, power-profile, clipboard, files,
capture, not just Tools/System — so you no longer have to Esc out and
reopen to back up. A `back` helper + a shared `BACK` label keep it uniform;
matched exactly so it can't collide with clipboard/filename content. Esc
still quits instantly.
Snapshot fallback: btrfs-assistant 2.2 segfaults on 26.05 (libbtrfsutil
ABI), so `nomarchy-menu snapshot` now launches nomarchy-snapshots instead —
a keyboard-driven snapper browser/restore. snapper is root-only here and
rofi can't run as root under Wayland, so it runs in a terminal via sudo
(one password prompt) and uses fzf to pick: browse/diff (read-only),
restore files (undochange), or roll back (root config only), each behind a
typed-`yes` confirmation — safer than a one-click rofi rollback.
btrfs-assistant stays installed for when nixpkgs fixes it.
Verified: flake check clean; home generation builds and the menu passes
bash -n with all Back + snapshot wiring; nomarchy-snapshots passes
shellcheck (writeShellApplication) in a snapper-enabled system build.
Pending an on-machine check of the restore/rollback paths.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Nomarchy roots are BTRFS, never ZFS, but the zfs module's `true` default for
forceImportRoot emits a warning on every eval/build ("recommended to set it
to false, the new default from 26.11"). Adopt that default early via
mkDefault — silences the noise, future-proofs for 26.11, and a real ZFS-root
downstream can still force it back on.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Harden the hardware-enablement broad seed for newer hardware:
- xe vs i915: intel.guc emits the i915 param (i915.enable_guc=3), which the
newer `xe` driver (Lunar Lake / Battlemage / Panther Lake, recent Xe GPUs)
ignores — it enables GuC by default. hardware-db.sh now reads the in-use
GPU driver (lspci -k) and writes `intel.guc = false` on xe hardware, so the
toggle isn't a misleading no-op there.
- NPU detection by PCI accelerator class [1200] (+ keywords), attributing the
vendor, instead of a fixed device-ID list — so new NPUs (e.g. Panther Lake)
are caught without a code change. Known IDs kept as a fallback/reference.
- Kernel awareness: a nomarchy.hardware.latestKernel escape hatch
(linuxPackages_latest) for very-new hardware whose drivers only just landed,
and a build-time warning when npu.enable predates the shipped kernel
(amdxdna needs 6.14+, intel_vpu wants recent). Downstream default is 6.18,
which already carries amd-pstate + amdxdna; the reference host pins latest.
- VM test (checks.hardware-toggles, pkgs.testers.runNixOSTest): boots a
minimal VM with the toggles on and asserts the config landed — amd_pstate
+ i915.enable_guc in /proc/cmdline, fprintd.service present, pam_fprintd in
/etc/pam.d/sudo. Passing. (Hardware behaviour still needs bare metal.)
Docs: README + template note latestKernel; ROADMAP records the hardening.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The nixos-hardware "common-*" profiles the installer selects cover the
basics (microcode, the Intel/AMD VA-API media stack, weekly fstrim), and
power.nix adds thermald + power-profiles-daemon. This adds the gap above
them, generically and detected at install time.
New modules/nixos/hardware.nix exposes a vendor-keyed nomarchy.hardware.*
surface — broadly-beneficial bits default ON when the vendor is detected
(opt-out), heavy/experimental bits behind opt-in toggles:
- intel.enable -> GuC/HuC (i915.enable_guc=3); intel.computeRuntime
(opt-in: intel-compute-runtime + vpl-gpu-rt)
- amd.enable -> amd_pstate=active + radeonsi VA-API env; amd.rocm.enable
+ amd.rocm.gfxOverride (opt-in: ROCm HIP/OpenCL)
- fingerprint.enable -> fprintd; fingerprint.pam (opt-in: login + sudo)
- npu.enable (opt-in/experimental) -> the in-kernel driver
(amdxdna/intel_vpu) keyed by vendor; userspace runtime is BYO
hardware-db.sh now detects Intel/AMD, a fingerprint reader (libfprint USB
vendor IDs) and an NPU (Intel VPU / AMD XDNA PCI IDs), emitting NOMARCHY
lines the installer bakes into system.nix — safe defaults active, opt-ins
commented. Audited against the commons to avoid double-setting.
Verified: flake check green; a toggles-on build has amd_pstate=active +
i915.enable_guc=3 in kernel-params and ships fprintd.service; detection
runs correctly on the (AMD Ryzen-AI) dev machine. On-hardware verification
of the AMD/NPU/Intel-compute runtime bits is still pending.
Docs: template commented examples, README option table, ROADMAP status.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
VM-verified the full batch of opt-in service toggles (libvirt, docker, obs
+ v4l2loopback, ollama, printing, openrgb, wireshark, adb, gamemode,
kdeconnect, restic) in a headless nixosTest — all green, no module bugs.
The one thing worth surfacing: services.restic ships a pre-configured
`restic-nomarchy` wrapper (repository + passwordFile baked in) for listing
and restoring snapshots. Mention it in the option description and README so
the restore path is discoverable — a backup you can't restore is half a
feature.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Finishing the four remaining "opt-in services" candidates — but two of
them aren't service-shaped, so the package-vs-toggle principle applies:
- Nextcloud client → a bare tray app; already in the app-suite menu,
not a service.
- WireGuard → NetworkManager (on distro-wide) imports .conf
tunnels natively, so there's no daemon to "just
enable". Only the wg/wg-quick CLI is missing →
wireguard-tools added to the app-suite menu.
The two that ARE config-backed become nomarchy.services.* toggles:
- openrgb → services.hardware.openrgb (daemon + device udev rules).
- restic → a small scaffolded surface (repository / passwordFile /
paths) over services.restic.backups: daily timer, 7/4/6
retention, --exclude-caches, repo auto-init. Asserts that
repository + passwordFile are set when enabled. passwordFile
is typed `str`, not `path`, so a flake-relative secret can't
be copied into the world-readable store.
README option table and ROADMAP (services now "Seventeen shipped";
cloud-sync / networking / backup / devices areas updated) + the downstream
system.nix examples. Both branches eval-verified; the restic assertions
confirmed firing when repo/password are unset.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Make the whole workstation menu available, all off/commented by default.
home.nix: the bare GUI apps become a categorized commented menu (~50
suggestions across browsers, comms, office, notes, code, graphics, A/V,
media, gaming, files/sync, security, system, local AI) — uncomment to add.
services.nix: the members that need system config (not just a package)
become opt-in nomarchy.services.* toggles, each with a commented example
in the downstream system.nix:
- docker → rootful Docker + user in docker group; asserts against the
podman docker-compat (both provide the `docker` command)
- kdeconnect → programs.kdeconnect (opens phone-pairing ports)
- gamemode → Feral GameMode daemon
- adb → android-tools (programs.adb was removed; systemd ≥258 does
the device uaccess udev rules, so no group needed)
- wireshark → Qt GUI + user in wireshark group (capture without root)
- ollama → local LLM runtime on 127.0.0.1:11434
- printing → CUPS + Avahi/mDNS network printer discovery
README option table and ROADMAP (services item now "Fifteen shipped";
local-AI / containers / gaming / devices areas marked) updated. All seven
branches eval-verified with the toggles forced on; the docker/podman
assertion confirmed firing.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The "default application suite" roadmap item, but deliberately NOT a
nomarchy.apps.* option surface: a bare package install has no config to
put behind a toggle, so a thin `mkIf x { systemPackages = [p]; }` only
reinvents what a package list already gives. Instead the suite is a
curated home.packages list in templates/downstream/home.nix (extending
the existing `# firefox` convention) — active staples (libreoffice,
vscode, gimp, inkscape, mpv, amberol) plus commented suggestions. The
opt-out is deleting a line; it also keeps the apps out of the live ISO
automatically (the ISO never scaffolds from the template).
The members that DO need system config become nomarchy.services.* opt-in
toggles instead (config behind the switch, like the existing services):
- steam → programs.steam (32-bit stack, controller udev, RP ports)
- libvirt → libvirtd + virt-manager, login user added to libvirtd group
- obs → obs-studio + a v4l2loopback virtual camera (exclusive_caps=1)
selectable as a webcam in Zoom/Teams
README option table, ROADMAP (services item now "Eight shipped"; suite
item marked done with the no-surface rationale) and the downstream
system.nix examples updated to match.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Extend nomarchy.services.* with three more opt-in toggles:
- podman: rootless containers, `docker` aliased to it, container DNS, and a
subuid/subgid range for the login user.
- flatpak: services.flatpak + a oneshot that adds the Flathub remote
system-wide (idempotent, retries when online).
- pika: ships pika-backup, the GUI for scheduled Borg backups.
Commented examples in the downstream system.nix template, per convention.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Start the opt-in services surface: nomarchy.services.{tailscale,syncthing},
off by default, in modules/nixos/services.nix. Tailscale ships the daemon
(authenticate with `sudo tailscale up`); Syncthing runs as the login user
with its GUI on 127.0.0.1:8384 (overrideDevices/Folders off so GUI-managed
folders survive rebuilds). username is read lazily, only when Syncthing is
on. Commented examples in the downstream system.nix template per the opt-in
convention.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Surface snapper snapshots from the desktop: `nomarchy-menu snapshot` (in the
SUPER+M picker, self-gated) launches btrfs-assistant, the polkit-aware
browse/diff/restore/rollback UI -- safer than a fat-fingerable rofi rollback
for a destructive root-only op. Shipped system-side gated on
nomarchy.system.snapper.
Also snapshot /home by default (hourly 5 / daily 7 / weekly 4) when it's its
own BTRFS subvolume -- the installer only set up snapper for /. A boot
oneshot creates the required /home/.snapshots subvolume if missing, so
already-installed machines get it too (no disko change).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The previous nomarchy-hibernate-unlock ran `pkill hyprlock` after a
hibernate resume, but killing a Wayland session-lock client without
releasing the lock trips hyprlock's "go to a tty" crash failsafe -- the
compositor keeps the screen locked for safety. So resume showed a hyprlock
error screen instead of unlocking.
Fix it the right way: never engage the lock before an encrypted hibernate
(the LUKS passphrase at resume is the gate). Replace the post-resume unlock
with a nomarchy-lock-before-sleep unit that takes over hypridle's
before_sleep_cmd: it locks on the RAM-resume sleeps (suspend / hybrid-sleep
/ suspend-then-hibernate) always, and on hibernate.target only when the disk
is unencrypted. idle.nix drops before_sleep_cmd accordingly.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The downstream flake at ~/.nomarchy is a live working tree by design —
nomarchy-theme-sync rewrites theme-state.json on every switch — so Nix's
dirty-tree warning fired on every sys-update/home-update and was just
noise. Default nix.settings.warn-dirty to false (mkDefault, overridable).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Ship `services.fwupd.enable` distro-wide (mkDefault) for firmware updates
via LVFS — the daemon + metadata-refresh timer only, no automatic
flashing, so applying an update stays an explicit `fwupdmgr update`. A
no-op where LVFS sees no devices; overridable natively. Sits with the
other mkDefault desktop services rather than a nomarchy toggle (it's a
single service, like upower).
Also documents the `nomarchy.system.power.*` options in the README table
(missed when the power module landed) and adds a short note on the
fwupd default.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Adds a `nomarchy.system.power.*` surface (modules/nixos/power.nix),
replacing the bare `services.upower` (Waybar reporting only) baseline:
- power-profiles-daemon ships by default; switched through polkit (no
root prompt). TLP is the opt-in `backend = "tlp"` for deeper battery
tuning — the two are mutually exclusive (asserted).
- thermald behind `power.thermal.enable` (Intel-only).
- `power.batteryChargeLimit` (e.g. 80): a backend-independent sysfs
oneshot writes charge_control_end_threshold; gated on `power.laptop`.
Switcher + indicator live home-side and self-gate on a battery being
present + powerprofilesctl running, so they hide on desktops and under
TLP — no system->home wiring (like the Waybar battery widget):
- rofi.nix: `nomarchy-menu power-profile` module + a picker entry shown
only on laptops with PPD.
- waybar.nix: first `custom/*` module — click cycles the profile.
Installer writes `power.laptop = true` (battery probe) and
`power.thermal.enable` (GenuineIntel) into the generated system.nix, and
scaffolds the charge limit commented-out.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
hypridle locks hyprlock before every sleep -- right for suspend, but
redundant for hibernate: the LUKS passphrase entered at boot already gates
the resume, so the user typed a password twice. Add a nomarchy-hibernate-
unlock systemd unit (WantedBy hibernate.target, After systemd-hibernate.service
=> runs post-resume) that dismisses hyprlock, gated on the disk being LUKS-
encrypted. Suspend (and the RAM-resume phase of suspend-then-hibernate) keep
locking -- they have no passphrase gate; an unencrypted hibernate keeps its
lock too. idle.nix gains a pointer comment; roadmap item marked done.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Setting only services.xserver.xkb.layout moves the graphical session but
leaves console.useXkbConfig unset, so the virtual-console keymap -- and the
systemd initrd the LUKS passphrase prompt runs in -- fall back to US (seen
on a Latitude 5410: X11 Layout gb, VC Keymap unset, LUKS prompt US).
Default console.useXkbConfig + console.earlySetup + boot.initrd.systemd.enable
distro-wide, so xkb.layout alone reaches the console and the encrypted-disk
prompt. Moved the latter two out of plymouth.nix -- they no longer hinge on
the splash being enabled. All mkDefault: an explicit console.keyMap or a
scripted initrd still wins.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Two branding follow-ups, using the vendored legacy assets:
- fastfetch (modules/home/fastfetch.nix, nomarchy.fastfetch.enable): the
vector logo recolored to the palette accent and rendered to compact
block-art via chafa at build time (so it tracks the theme), fronting a
curated module list. Replaces the oversized legacy ASCII.
- Waybar logo glyph: vendor Nomarchy.ttf (modules/nixos/branding/),
install it via fonts.packages, and switch the summer-day/night menu
buttons to its U+F000 mark with `font-family: Nomarchy` pinned in CSS
(Nerd Fonts also occupy U+F000, so the pin is required). Glyph stored as
the JSON escape for robustness.
Also log a roadmap item for a quality-of-life alias collection.
Verified: font family resolves as "Nomarchy" and the ttf maps U+F000;
both summer jsonc parse; chafa logo builds headless; nix flake check passes.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Vendor the vector logo from the legacy branch
(modules/nixos/branding/logo.svg) and use it for the live ISO's isolinux
splash: rsvg-convert renders it, recolored to the palette accent and
centred on the theme base (from theme-state.json) at build time, wired via
isoImage.splashImage in hosts/live.nix. Verified the generated PNG by
building it in the sandbox.
UEFI/GRUB still uses the stock theme (grubTheme needs a full theme dir —
left on the roadmap), as does the distroId switch.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Round 2 of distro branding, text half: tuigreet now shows
'Welcome to <distroName>', and users.motd carries a branded banner that
doubles as a cheat sheet for the distro helpers (sys-update, home-update,
nomarchy-theme-sync, SUPER+?). Both key off system.nixos.distroName so the
brand name stays in one place.
Deferred (need an asset / a boot test): ISO splashImage/grubTheme art and
the distroId switch.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The shipped boot art was a fixed navy that vanished against dark theme
bases. Recolor each element from theme-state.json at build time (flat
fill, alpha preserved): foreground glyphs (logo, padlock, password dots)
→ text, the field/track boxes (entry, progress_box) → surface, and the
progress fill → accent. text/surface both contrast with base in either
polarity, so the splash now reads on light and dark themes alike.
Adds imagemagick as a build-time input to the theme derivation. Like the
background tint, this follows the theme as of the last system rebuild.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Three quick known-issue fixes:
- yazi: 26.x made the fetcher `group` field required, so the git
prepend_fetchers failed to parse and yazi fell back to presets. Add
`group = "git"` to both entries (the git plugin's setup() only renders
the linemode — the fetcher still needs registering here).
- file-manager: xfce.exo and the three Thunar plugins moved to top-level;
use pkgs.xfce4-exo / pkgs.thunar-{archive-plugin,volman,media-tags-plugin}.
Clears the eval-time deprecation warnings.
- nix-ld: enable programs.nix-ld distro-wide so foreign dynamically-linked
binaries (downloaded tools, npx-fetched claude-code, language servers)
run without patchelf.
Also drops the now-stale claude-code mention from the allowUnfree comment
(the ask module uses npx). nix flake check passes; xfce warnings gone.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
zsh is now the default login shell (programs.zsh.enable +
users.defaultUserShell, mkOverride 500 to beat bash.nix's mkDefault;
per-user shell= still overrides). modules/home/shell.nix configures the
interactive experience (nomarchy.shell.enable):
- zsh: completion, autosuggestions, syntax highlighting, big shared
history, autocd.
- starship prompt themed from the palette (a `nomarchy` palette built
from theme-state.json colors).
- bat as cat (theme "ansi" → tracks the terminal/theme palette), eza as
ls (icons + git + dirs-first), zoxide as cd (--cmd cd).
- rg/fd shipped as themselves — deliberately NOT aliased over grep/find
(flag differences surprise).
home.shell.enableZshIntegration = true is the single lever; starship,
eza, zoxide, ghostty and yazi all emit their zsh hooks from it (no
per-module edits). Verified: flake check green, default shell resolves
to zsh, .zshrc carries the starship/zoxide/eza/yazi hooks + aliases,
starship palette renders, all tools in the live ISO closure.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The media keys called wpctl/brightnessctl directly with no visual
feedback — changing volume showed nothing on screen. Now they drive
swayosd-client, which performs the action AND shows a themed OSD.
- modules/home/osd.nix: services.swayosd, style.css generated from the
palette (base bg, accent progress/border); nomarchy.osd.enable.
- hyprland media binds → swayosd-client (--output-volume raise/lower/
mute-toggle, --input-volume mute-toggle, --brightness raise/lower).
- modules/nixos: services.udev.packages = [ swayosd ] installs the
backlight rule (chgrp video + g+w) so brightness works for video-group
users without root (the installer/live user is in video).
swaync is unaffected (it was already running — this was purely the
missing OSD). Verified: flake check green, swayosd.service bound to
graphical-session.target, hyprland config verifies, swayosd in the live
ISO closure. On-screen rendering needs a real session to confirm.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Legacy shipped only Thunar (and a stray nautilus extension, mid-migration).
This pairs a keyboard-driven TUI flagship that fits the distro's identity
with the GUI half for point-and-click + the "open folder" handler.
yazi (modules/home/yazi.nix, nomarchy.yazi.enable, SUPER+E, `y` shell
wrapper cd's on exit): themed from theme-state.json (mgr/mode/status
mapped from the palette; merges over yazi's defaults so it tracks the
theme) with a curated, correctly-wired plugin set —
full-border, git (setup + prepend_fetchers), smart-enter (l/Enter),
chmod (cm), mount (M), compress (caa), mediainfo (audio/subtitle only —
images/videos keep native thumbnails), glow (markdown previews).
Preview/tooling deps: ffmpegthumbnailer, poppler-utils, p7zip, jq, fd,
ripgrep, fzf, chafa, mediainfo, glow.
Thunar (modules/nixos/file-manager.nix, nomarchy.system.fileManager.enable):
GTK file manager (Stylix-themed) + archive/volman/media-tags plugins +
gvfs/tumbler/udisks2 backends + thumbnailers + exo. udisks2 also backs
yazi's mount plugin.
Verified: flake check green, HM generation builds (yazi TOML/init.lua
render, all 8 plugins resolve), every package present in the live ISO
closure (offline install draws from it — no source-build leak).
Not verified: yazi's runtime preview/plugin behaviour needs a real
session (TUI; can't drive in CI here).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
- Plymouth boot splash ported from the legacy branch (modules/nixos/
plymouth/): logo + eased progress + LUKS entry, background tinted from
theme-state.json via the new nomarchy.system.stateFile (wired by
mkFlake/lib.nix; null → Tokyo Night fallback). Default on; OFF on the
live ISO (boot-message visibility on the install medium). Pulls
boot.initrd.systemd, which also drives the keyboard-at-LUKS feature.
- system.nixos.distroName = "Nomarchy" (os-release PRETTY_NAME,
systemd-boot entries, ISO menu label). distroId left "nixos" (feeds
DEFAULT_HOSTNAME + upstream isNixos checks — roadmapped).
- nixpkgs.config.allowUnfree distro-wide (here + both import-nixpkgs
sites) — unblocks claude-code for the menu's ask-Claude module.
- systemd-boot.configurationLimit = 10 so entries don't fill the ESP.
- Live ISO: nomarchy.idle.enable = false — hypridle was suspending the
VM mid-install (the install-hung regression); installed systems keep it.
- flake.nix offline pins (verified 0-leak via a foreign-identity
gap-analysis probe): the repo's own standalone HM gen + inputDerivations,
mustache-go + stdenv (stylix re-renders base16 per switch), microcode-amd/
intel (enableRedistributableFirmware activated updateMicrocode →
source-build cascade), buildEnv's builder.pl, findXMLCatalogs, and the
representativeInstall mirror (xkb/initrd-systemd/microcode).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
nixpkgs.config.allowUnfree in the system module (covers the live ISO
via useGlobalPkgs and every mkFlake machine) plus the two explicit
`import nixpkgs` sites (repo + lib.nix) so the standalone HM desktop
sees the same package set. Unblocks claude-code for the menu system's
ask-Claude module, vendor drivers, etc. Opt out with
`nixpkgs.config = lib.mkForce { allowUnfree = false; }`.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Roadmap items 1+2 plus the issues found testing on a Latitude 5410:
- Ship the 10 most popular Nerd Fonts by default (iosevka swapped for
mononoki: 1.1 GB vs 27 MB) and warn from nomarchy-theme-sync when a
configured fonts.mono/fonts.ui family isn't installed (fc-match) —
fontconfig substitutes silently otherwise.
- hardware.enableRedistributableFirmware: installed systems shipped NO
firmware blobs (wifi/SOF audio/BT) — nixos-generate-config only emits
microcode lines referencing this flag, it never sets it.
- Live ISO wifi: networking.wireless.enable = mkForce false killed
NetworkManager's supplicant — since 26.05 the NM module drives its
wifi backend THROUGH networking.wireless (dbusControlled). Devices
vanished from nmtui with iwlwifi loaded and no rfkill block.
- Hyprland 0.55: launch sessions via start-hyprland (watchdog; bare
binary warns), add the new gesture keyword (workspace_swipe is gone —
touchpads had no gestures at all), media keys via wpctl/bindel/bindl
plus the missing mic/play/next/prev binds.
- Offline theme switching: pin mustache-go + stdenv + the repo's own
standalone HM generation (incl. home-files inputDerivation) into the
ISO — stylix re-renders base16 templates per switch and an offline
`apply` cascaded into building stdenv from source (1107 drvs).
Verified with tools/vm/gap-analysis.py: 17 config drvs, zero fetches.
- Stylix: track release-26.05 (kills the HM version-skew warning and
the stale home-manager follows), lock updated.
- Roadmap: swaync (no notification daemon ships today).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
- LUKS2 is the installer default; in exchange the generated config sets
the new nomarchy.system.greeter.autoLogin (greetd initial_session) —
the disk passphrase already gates the machine.
- @snapshots subvolume + nomarchy.system.snapper.enable: hourly/daily
timeline snapshots of / and the nixos-rebuild-snap helper, ported from
the previous iteration (3bdfc35), guarded to no-op on non-BTRFS roots.
- @swap subvolume with a swapfile sized to RAM by default (disko
mkswapfile handles NOCOW); the installer computes the resume offset
and wires boot.resumeDevice + resume_offset for hibernation.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Full replacement of the previous iteration, rebuilt around three ideas:
- Pure evaluation: theme-state.json lives inside the flake and is read
via the nomarchy.stateFile option — no --impure, ever.
- All-Home-Manager theming: `nomarchy-theme-sync apply` writes the JSON
and runs `home-manager switch`; every theme change is one atomic,
rollbackable generation. Wallpaper (swww) is the sole runtime piece.
- Flat, downstream-first layout: modules/{nixos,home} with one
options.nix each, exported as nixosModules/homeModules + overlay +
flake template; system (nixos-rebuild) and desktop (home-manager
switch) rebuild paths are fully split.
Ships 21 themes imported from the previous iteration (palettes,
wallpapers, btop themes, six whole-swap Waybar identities), Stylix for
the GTK/Qt/cursor long tail, a live ISO target with offline theme
switching, and docs/TESTING.md with the QEMU verification workflow.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
