Re: [quicwg/base-drafts] Proposed change to #1965 (#2212)

[Nick Banks <[e-mail]>]

nibanks commented on this pull request.



> @@ -496,11 +496,14 @@ MAY use alternate strategies for determining the content of probe packets,
 including sending new or retransmitted data based on the application's
 priorities.
 
-If no new data or unacknowledged data is available to send, an ack-eliciting
-packet SHOULD be sent.  Sending a retransmittable frame ensures that any in
-flight packets are acknowledged or declared lost in a timely manner.  If no
-ack-eliciting packet is sent, any packets currently in flight should be
-declared lost to avoid repeatedly arming and firing the PTO timer.
+When a PTO timer expires, new or previously-sent data may not available to send,
+and data may still be in flight.  Under these conditions, a sender MUST mark any
+packets still in flight as lost.  This can happen for example when data is sent
+on a stream which is then reset by a sender, and a PTO timer expires after this
+stream is reset.  A sender can be blocked from sending new data in the future if

If the stream is reset, there is still ack-eliciting data pending, the RESET_STREAM frame. When that frame is sent, a new PTO would be sent.

Does this case that would trigger your scenaior: if you sent X (could be s STREAM frame or something else) and then a PTO fires and you retransmit X, but shortly after you receive an ACK for the first transmission. But then PTO fires for the retransmission.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/quicwg/base-drafts/pull/2212#pullrequestreview-186826809