Peter, > sounds like things are going well - excellent.
Thanks. >>*Session* >> >>It now also enshrines a session id. The session id is a long. It is an >>int serial number shifted left 16 bits with a random int added (to >>prevent spoofing). >> > >I wonder if this is needed. Some transports (ie SSL and TCP) already indicate >a session because you have constant comms and the session is broken when the >transport is teared down. > Two of the transports Piped & Direct are not marshalled over links that have session, though it would be interested to share for those that do. Perhaps those that are in VM no not need session at all as we could assume the machine is safe from internal hacking. >It would be nice not to require the API to have a >session id associated with it but instead have the session id a >characteristeric of the transport. > It is invisible to the user/developer of the client and server sides. I.e. it is not expsed in the API. Regards, - Paul H -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
