ANONYMOUS wrote:
> So, there can be multiple implementations of this? The way I thought it would be is that if something is in transition between states (and that transition takes some amount of time) it is neither in its current state nor in the place it’s going but rather in some other ‘transition queue’ or something like that. Depending on the implementation will that not change the output for the total time.
There's 'always' multiple ways to implementing something. Different (correct) implementation should produce the same results - as long as a process requests something requiring a state transition at a certain time, and ends up at its eventual destination at the correct time, then 'where' it sits (in implementation) during that period doesn't matter.
> Moreover I checked the sample solutions for a simple commands file, one which after some time sleeps for 6 usecs. In this case the program first went to blocked, did nothing for 10 usec, then left blocked. I thought the 10 usec transition time was independent of the sleep time.
Was the requested sleep time shorter or longer than the timequantum?
If the sleep-time was for 6usecs and the timequantum 10usecs, all seems correct.