>> You should not have to deal with the delayed closing: MINA is already >> allowing you to do that, if you call closeOnFlush() instead of closeNow() -or close(), which maps to closeNow()-. It will then flushall the pending messages before closing the session. No message written in the session *after* you have called close(false) will be sent.
For all I know you may be right (I have not reviewed this option in depth) - please note though that the SSHD code does not deal directly with MINA but rather uses an *abstraction layer* that enables it to invoke various different network transports - MINA being one of them (NIO2 and Netty being others - and in the future... who knows). That being said, we try to keep the transport implementation requirements rather simple so others may find it easier to implement a new transport. Requiring a specialized feature such as this one (or the other you mention) is likely to complicate future implementations. Furthermore, even if we made such features optional (e.g., via *isFeatureXXXSupportedI) *both the SSHD code and the compatibility matrix (which feature is required for which SSHD functionality) would make it (IMO) complex (more than already is) and cumbersome to use (users would have to be aware what are they *giving up* when switching transports. E,g, what if users need 2 features - one of which is supported by one transport but not the other and the other feature only by its counterpart. >> I would suggest filling a JIRA ticket for this one, it requires some further investigation, but I suspect it would ease your implementation A LOT Again, you may be 100% right, but in view of the above I do not want to request lots of work (perhaps complex) on what may prove to be a "niche" piece of code especially tailored for SSHD. Furthermore, I am not sure this is exactly what is needed in this case - and until I am reasonably sure I can define the exact requirements I don't think it would be a good idea. Thanks for the feedback, Lyor
