somewhere along the way the hack to disable of gettys for kiosk appliances completely broke.
if gettys can not be disabled on core systems mir-kiosk and the getty program fight over who owns the tty. this typically ends in mir dieing on first boot (it works on subsequent boots but in case there is a UI supposed to come up right after installation of a kiosk appliance this looks indeed very broken)
also, having login prompts at all on kiosks that you can reach by simply plugging in a kbd and switching ttys is often considered a security risk by customers.
this PR fixes the regressed behaviour of allowing to set a "nogetty" option in cmdline.txt. if the option is set it will mask the getty@tty1.service and also turn off auto-spawning of tty's in systemd. in case the option is removed from the cmdline the original behaviour is restored.
This PR also fixes the broken travis test alongside
somewhere along the way the hack to disable of gettys for kiosk appliances completely broke. if gettys can not be disabled on core systems mir-kiosk and the getty program fight over who owns the tty. this typically ends in mir dieing on first boot (it works on subsequent boots but in case there is a UI supposed to come up right after installation of a kiosk appliance this looks indeed very broken)
also, having login prompts at all on kiosks that you can reach by simply plugging in a kbd and switching ttys is often considered a security risk by customers.
this PR fixes the regressed behaviour of allowing to set a "nogetty" option in cmdline.txt. if the option is set it will mask the getty@tty1.service and also turn off auto-spawning of tty's in systemd. in case the option is removed from the cmdline the original behaviour is restored.
This PR also fixes the broken travis test alongside