# Vigil TODO list 

## Reminders

- Never panic from non-critical failures (e.g. non-root filesystem mount failure).
- All errors should be logged.
- Configuration via TOML.

## Roadmap

### System init

- [x] Mount filesystems (`/proc`, `/sys`, `/dev`).
- [x] Spawn `udev` (or compatible device manager) as child process.
- [x] Mount user-defined filesystems from `/etc/fstab`
- [x] Activate `swap`
- [x] Set hostname, timezone, and locale from config.
- [x] Load kernel modules (via `modprobe` or direct `init_module` syscall).

### Core runtime features

- [ ] **Service management**
  - [ ] Parse TOML unit files (`/etc/vigil/units/*.toml`)
  - [ ] Start/stop/restart/status via `vigilctl`
  - [ ] Handle `Wants=`, `After=`, `Before=` dependencies
  - [ ] Auto-restart failed services (configurable via option `restart = always|on-failure|never`)
- [x] **Child process reaping**
  - [x] Install `SIGCHLD` handler
  - [x] Call `waitpid(-1, ...)` in loop to reap zombies
  - [ ] Log exit status, signal, and runtime duration per service
- [ ] **TTY & login**
  - [ ] Launch `getty` on configured TTYs (e.g. `tty1`–`tty6`)
  - [ ] Support custom `getty` paths/args per TTY in config

### System lifecycle control

- [ ] Handle `reboot`, `halt`, `poweroff` via `vigilctl`
- [ ] Properly terminate all services in reverse dependency order
- [ ] Sync filesystems and unmount (best-effort)
- [ ] Invoke `reboot(2)` / `halt(2)` syscalls directly

### Signal handling

- [ ] `SIGINT` / `SIGTERM` = graceful shutdown
- [ ] `SIGUSR1` / `SIGUSR2` = reload config or trigger debug dump
- [ ] Block all non-fatal signals during critical sections (e.g. mount)

### Logging

Logging should not stop on printing colored messages to stdout
in addition, we __should__ send logs to /var/log/vigil.log or smth simmilar.
<!-- Also time to think how and where to store service logs, create generic /usr/local/log/vigil/*service_name*; or store them with /var/log/vigil/*service_name* etc.)-->