stated by URTOS author:
the need to ensure each task completes in less than one tick time

Thus stated by me:
What a nightmare maintenancewise. Just visulaize: some task "cross the edge" when 5 random facts concur (i.e. once a month or once a year) and the poor sod that is tasked with fixing it has the chance of a tailor in hades of reproducing it.

False statement in prev post:
A time triggered sheduler only runs one task per tick so this situation never occurs unlike interupt driven systems where you have no control over when things happen.
because
If a task that has to "complete in less than one tick time" has 47 if statements testing flags etc set by external evants, how are you to guarantee that the "task completes in less than one tick time". For example my thingy here that run with a powerful processor and a URTOS that has no such restrictions, test several random events (a simple example "key received") in many tasks. These tasks has a very large "completion time" variance.

Erik

PS have you ever seen an ICE (truly Maintenance Heaven) getting fit for the insane asylum? If not, try running it with a URTOS.