Re: Closing on CONNECTION_CLOSE

[Patrick McManus <[e-mail]>]

I agree that the rst-like semantic in the pull request is sensible here.
Some more language around expecting applications to do graceful shutdown
before this happens would be helpful - especially with the definition of
GOAWAY migrating out of the transport document. This works well with the
silent close concept so that makes me happy.

I would like the text to say that the receiver should not discard data that
it has acknowledged before receipt of the CLOSE (discarding it is the TCP
receiver RST behavior). Are there objections to adding that?

I disagree a little bit with (I think) Subodh and Wily regarding what to
send when you recv data after you have sent a CLOSE. I want to strongly
encourage implementations to have a TIME_WAIT like period of being able to
a fully protected CLOSE (it can be a memcpy retransmit). public reset ought
to be left as a last resort.