[cap-talk] Throwing down the gauntlet
iang at systemics.com
Fri Jan 13 10:23:52 EST 2006
John C. McCabe-Dansted wrote:
> David Wagner suggested that it is impossible to write reliable software
> without bidirectional communication. I am not yet convinced this is the case.
As a generalism I think it is quite true, and I'd
be fascinated if you could find a case where you
could challenge it. As a comp sci problem it is
called "the coordination problem" I gather although
there isn't an easy / good reference for it.
One of the common things that happens is that designs
ignore this and use uni-directional comms. Then,
the software becomes "unreliable" and as a consequence
users implement a reliability protocol over the top,
often in meatspace not digitalspace.
The other common failing is to take a "reliable"
protocol like TCP/IP and assume that it reliably
does what you want it to. Same result, the
unreliability eventually results in a reliability
protocol being added over the top.
More information about the cap-talk