Hi Everyone,

I'm trying to get a better grip on process scheduling in Pharo but
don't understand the behaviour I'm seeing.

My understanding is that the VM (which does the real work of
scheduling the processes) is a pre-emptive co-operative model,
meaning:

- Higher priority processes are always scheduled ahead of lower
priority processes.
- Processes at the same priority get the CPU until something is done
to explicitly stop processing, e.g.:
-- terminate the process, wait on semaphore (I/O, timer, mutex, etc.),
#yield the processor

But given the following:

| p x a index |

p := 35.
a := Array new: 20.
index := 1.
[ 1 to: 10 do: [ :i |
�� �� [ a at: index put: i.
�� �� index := index + 1.
�� �� x := 0.
�� �� 300000000 timesRepeat: [ x := x + 1 ].
�� �� a at: index put: i.
�� �� index := index + 1. ] forkAt: p named: 'Proc', i asString ].
] forkAt: p+1.
a.


The code above was written to be as simple as possible (it isn't going
to be creating streams, have semaphores etc. called somewhere deep
within).�� The 300000000 number was manually determined as something
that will run for a few seconds on my laptop.

Since the forking process is higher priority than the 10 worker
processes, I expect that all 10 processes will be created before the
first gets any CPU, and they will then be executed in order (since
there is nothing obvious to cause a context switch), and the results
to be:

#(1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10)