RE: BSC - Distributed intelligence DRAFT

[Patrick Lidstone \(Personal e-mail\)]

A couple of thoughts:

1. There have a been a number of occasions where we have proposed
null-bodies as part of "probes" or "requests". Perhaps
the spec
should be revised to make messages with no body at all legal?
If anyone is concerned about backwards compatibility, we could
do this by using a new header (xap-request perhaps?), although
I don't like this very much.

2. The current schema doesn't allow for any translation between
the source and target mapping. Perhaps this is overcomplicating
a straight forward concept, but the most obvious extension would
be to allow for (a) inverted operation (switch=on, light=off /
switch=off, light=on), and (b) timed/pulse operation (switch=on,
output fires for x secs. A simple way of dealing with this would
be to allow an association to include a condition name, value
pair. We could then extend the condition capabilities as
real-world needs arise.

Patrick