Removed halt and replaced it with a delay and reboot

[wel97459]

Replaced the halt function with a delay and reboot, instead of an infinite loop. This removes an edge case were there's a bad brown out and the radio fails to initialize properly.

As @liamcottle suggested I have added an error message that is customizable to other failure cases. Available on node with a display enabled.

[LitBomb]

halt means stop. if you delay 10 seconds and then reboot, then it is a reboot and not a halt any more, is it?

[fschrempf]

This could keep some boards from booting when there's dirty power, and the radio chip fails to init correctly.

And I guess that's exactly what is intended. If there is a critical failure, there is no reason to continue. Rebooting could cause nasty boot loops that won't help either.

[liamcottle]

For radio init failure, if the device has a screen, I'd probably prefer to show a message that says radio init failed.

[wel97459]

The halt function is only called in one place in the whole code base, just for that radio init in the being of boot. It's definitely a contributing factor for solar nodes not booting up properly.

And I have already seen improvements in getting my repeaters to start from a BMS cut off.

I would also hazard to say that a rover on Mars would not have a function like halt or a call to a function that doesn't try to get the rover back to an operational status. Most repeaters are in places where reset buttons are a pain to get too.

And I thinking adding a message for nodes with a screens is a good idea with a count down to reset or something.

[wel97459]

I'll change the function name. A halt function like this one is bad way to handle a radio init problem.

halt means stop. if you delay 10 seconds and then reboot, then it is a reboot and not a halt any more, is it?

[wel97459]

I personally would rather have a boot loop then a non-responsive repeater, with a chance of a fully functional boot.

This could keep some boards from booting when there's dirty power, and the radio chip fails to init correctly.

And I guess that's exactly what is intended. If there is a critical failure, there is no reason to continue. Rebooting could cause nasty boot loops that won't help either.

[wel97459]

I removed the halt function and added delayedReboot. The function takes in a message string and delay amount, and displays the message like in the image.

Without DISPLAY_CLASS defined it just delays and reboots.