At 12:05 PM 11/26/2007, Tim Ellison wrote:
>What happens when all of the devices start changing the state of 
>PowerSDR and one device expects it to be in one state, but another 
>CAT device changes it to something else.  If it is PTT assertion, 
>this could get very ugly.


This is the fundamental problem with one process maintaining a model 
of the internal state of another process.

You can resolve it (in a theoretical sense) by using different 
control strategies (most of which, of course, didn't happen to be 
implemented by the mfr/creator in at least one logical component in 
question)...

a) Your controlling process can interrogate the controlled process 
for its state and take appropriate action accordingly. This presumes 
that there are alternative courses of action that make sense AND that 
the controlled process doesn't spontaneously change state during that 
time.  You can get around this if there are semaphores and/or "test 
and set" sorts of indivisible operations available.

b) you can create an intermediate arbiter that manages the 
interactions.  Rather than Programs 1 and 2 both talking to device A, 
they talk to program 3, which then talks to device A. Program 3 
implements some algorithms to manage the interaction (exclusive use 
with timeouts is a very popular strategy...)

What's nice about languages like Erlang (and many similar things) is 
that they give you a fairly structured way to describe what you want to do.


Jim, W6RMK



_______________________________________________
FlexRadio mailing list
[email protected]
http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz
Archive Link: http://www.mail-archive.com/flexradio%40flex-radio.biz/
FlexRadio Knowledge Base: http://kb.flex-radio.com/
FlexRadio Homepage: http://www.flex-radio.com/

Reply via email to