Matthew, > I know this was discussed a while back and am just curious where your > thinking is at in regards to a timeout parameter supplied to recv for > [REQ / REP] RPC use cases where the reply is never going to happen... Yes, it should definitely be included. Want to give it a try? Martin