Sorry for yet another mail in this long thread but it looks like it is
preemptive after all. Maybe the scheduling is a bit primitive for tasks
having same priorities or there is another isuue (locking, PACE not
being multithreaded ...) but the preemptive nature works at least for
bluetooth and network stack. Maybe same could be verified also with
debugging server running in background (..2).
I tried this example again without SysTaskDelay but enabled network over
bluetooth and pinging the device continuously and even if the program
does not return to launcher pinging still works so the device is not
completely dead. I don't think whole TCP/IP and bluetooth layer is done
in interrupt context or something like that so the preemptive nature
seems to be there. All tried on Tungsten T2 (OS 5.2.1).
Frantisek
--
For information on using the PalmSource Developer Forums, or to unsubscribe,
please see http://www.palmos.com/dev/support/forums/
