Re: Poison

[A E Lawrence <adrian.lawrence@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx>]

P.H.Welch@xxxxxxxxx wrote:

> 
> So, despite the semantic complexities introduced by exception handling,
> its application (for termination at least) on processes trying to use
> a poisoned channel seems quite understandable - so there ought to be
> a (formal) semantics out there that simply captures it ...


I've been too busy and distracted to follow this thread in detail, but
as someone else pointed out, there is a formal semantics for an
interrupt in standard CSP. It just doesn't appear in occam. 

Maybe that is because PRI ALT subsumes interrupts? And occam has to
worry about the state of memory.  I'm not sure: maybe we need a bit of
plumbing to ensure that a process can always responds immediately.

What is the semantics if a processes has "committed" to a communication
and then receives an interrupt? And the communication is between two
remote chunks of hardware? Presumably we must wait, perhaps for a very
long time, for the partner to complete the event. Otherwise we need some
sort of end-to-end negotiation allowing for an offer to communicate to
be withdrawn?

Sorry if this has already been discussed. I will try and catch up
sometime :-(

Adrian
-- 
Dr A E Lawrence