W3C home > Mailing lists > Public > ietf-http-wg@w3.org > January to March 2019

Re: Shouldn't we add a new CLOSING state in H2 ?

From: Martin Thomson <mt@lowentropy.net>
Date: Sun, 27 Jan 2019 09:31:38 +1100
Message-Id: <1548541898.1607094.1644271784.51AAFB02@webmail.messagingengine.com>
To: Patrick McManus <mcmanus@ducksong.com>
Cc: HTTP Working Group <ietf-http-wg@w3.org>
On Sun, Jan 27, 2019, at 09:29, Patrick McManus wrote:
> Perhaps you are looking at a broader issue than I am (and thought was what
> wily was describing).. I was specifically referring to the pattern that
> forces the RST when tcp data hits the closed connection which in return
> forces the truncation and dropping of buffered valid recvd (even ack'd)
> data on the client.. That should go away (ha) in h3, no?

So if the server closes the connection, it will re-send the packet containing the CONNECTION_CLOSE, but that's all.  Whatever data the client might have received should still be available.
Received on Saturday, 26 January 2019 22:32:00 UTC

This archive was generated by hypermail 2.3.1 : Saturday, 26 January 2019 22:32:01 UTC