Re: BEEP adjustments [was: Re: I-D ACTION:draft-ietf-sacred-protocol-bss-01.txt]

> > Drop 2.4 since it does not make things more secure or efficient but adds
> > confusion (e.g. is there some operational requirement that requires the
> > server to drop the connection after a successful exchange?)
>
> Don't we need to say somewhere whether the session is expected to be
> kept open? I don't see any requirement that a sacred server keep the
> connection open, and I think its simpler if it just handles a single
> transaction per connection/session. Hence the current text. I wouldn't
> agree that saying nothing is right though.

well, actually, saying nothing is exactly what you want to do.

it's pretty much up to the initiator to decide what they want to do with the
session.

while a listener should certainly have an idle timeout, it's rather
presumptuous on the part of the listener to handle one request and then
exit.

all section 2.4 does is add a fair amount of implementational burden for no
added gain.

/mtr