Re: [quicwg/base-drafts] When to ignore coalesced packets and when to terminate the connection (#1840)

Mike Bishop <notifications@github.com> Mon, 08 October 2018 18:34 UTC

Date: Mon, 08 Oct 2018 18:34:16 +0000
From: Mike Bishop <notifications@github.com>
Reply-To: quicwg/base-drafts <reply+0166e4abea041e0527c9430a62564231f05cdb695fd6814092cf0000000117d3652892a169ce15ed11b4@reply.github.com>
To: quicwg/base-drafts <base-drafts@noreply.github.com>
Cc: Subscribed <subscribed@noreply.github.com>
Message-ID: <quicwg/base-drafts/issues/1840/427936644@github.com>
In-Reply-To: <quicwg/base-drafts/issues/1840@github.com>
References: <quicwg/base-drafts/issues/1840@github.com>
Subject: Re: [quicwg/base-drafts] When to ignore coalesced packets and when to terminate the connection (#1840)
Mime-Version: 1.0
Content-Type: multipart/alternative; boundary="--==_mimepart_5bbba3286f696_1f5b3f9282ad45b81890659"; charset="UTF-8"
Content-Transfer-Encoding: 7bit
Precedence: list
Archived-At: <https://mailarchive.ietf.org/arch/msg/quic-issues/BAa2U6dHrhLN_6bMHEzHMYCnm8U>

My interpretation would be that fatal errors can only occur *while processing* a packet, which is to say that protection has been successfully removed and you're consuming frames out of the payload.  Errors that prevent you from doing any processing in the first place, however, should cause you to keep going to other packets.

In other words, for most purposes, handle the packets independently -- if they were in separate datagrams and one was garbage, you'd move to the next-received datagram.

-- 
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/issues/1840#issuecomment-427936644

[quicwg/base-drafts] When to ignore coalesced pac… Igor Lubashev
Re: [quicwg/base-drafts] When to ignore coalesced… Mike Bishop
Re: [quicwg/base-drafts] When to ignore coalesced… Martin Thomson
Re: [quicwg/base-drafts] When to ignore coalesced… Martin Thomson