Make "device is inactive" not spam when TTY is inactive

[YaLTeR]

Discussed in #2169

^{Originally posted by stefonarch August 2, 2025}
For testing purposes I often use a different LXQt Wayland session started on another tty, when it's niri after exciting I find lots of messages

ago 02 20:50:18 casa niri[547577]: 2025-08-02T18:50:18.551529Z  WARN niri::backend::tty: device is inactive

printed on the screen. I've added systemd-cat --identifier=niri to log things in the start script and now all those end up in the journal. I went to count them...

ago 02 20:50:17 casa niri[547577]: 2025-08-02T18:50:17.354128Z  WARN niri::backend::tty: device is inactive
ago 02 20:50:18 casa niri[547577]: 2025-08-02T18:50:18.348144Z  WARN niri::backend::tty: device is inactive
ago 02 20:50:18 casa niri[547577]: 2025-08-02T18:50:18.348432Z  WARN niri::backend::tty: device is inactive
ago 02 20:50:18 casa niri[547577]: 2025-08-02T18:50:18.350041Z  WARN niri::backend::tty: device is inactive
ago 02 20:50:18 casa niri[547577]: 2025-08-02T18:50:18.355751Z  WARN niri::backend::tty: device is inactive
ago 02 20:50:18 casa niri[547577]: 2025-08-02T18:50:18.551483Z  WARN niri::backend::tty: device is inactive
ago 02 20:50:18 casa niri[547577]: 2025-08-02T18:50:18.551529Z  WARN niri::backend::tty: device is inactive
stef@casa:lxqt-panel$ journalctl -b |grep "device is"|wc -l
34487

I guess there's no easy way to disable this warning.

---

IIRC the underlying cause is that queue_redraw() will still queue even inactive outputs, then their redraw() gets called and prints this message. Need to think what is the best place to stop this.