> I would indeed expect to use the same mechanisms for delivering the error > notifications as for all other notificaitons. > > And I suppose one could allow a client to unsubscribe from such > notifications. But it actually seems to encourage dangerous behavior to do > so. And adds extra complexity.
Can you explain how it adds extra complexity? I can argue it adds more complexity in terms of race conditions/etc. either way, so I don't see a net gain in complexity from the protocol side. What I do see is the option to reduce complexity on the agent side of things by disconnecting the two data stores (?). Again, my entire purpose here is to clarify and simplify, not to make things harder... Complexity always has tradeoffs, though. :-) Russ _______________________________________________ i2rs mailing list [email protected] https://www.ietf.org/mailman/listinfo/i2rs
