Erik - absolutely right!

The addendum is, if your system is programmed as a background tasker, ALL your task, including the LONG ones, are SHORT (because they run as threaded strings of tasks). Hence the watchdog in the main loop is fine.

The nice thing about a tasker structure is the tasks are broken up into states and the threaded tasks can recover from timeouts are interrupted exchanges and reset them to the initial state.

Foreground simply needs to go on to other tasks or await/monitor a completion/failed task flag.

aka j