Re: draft-kohler-dccp-ccid3-drops-01.txt Re: Remote participation in DCCP WG, IETF-69

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Ian -

Many thanks for the feedback.

I have a couple of comments I'd like raised if possible (not sure I'll
be attending via jabber):
- is this actually needed as a separate RFC? Given how small it is,
perhaps it is more of a start of CCID3bis? I don't really know
anything about IETF processes though - so it may be appropriate as it
is - it just seems it would be better in core CCID3 document.

Whenever CCID3bis is started, this would certainly become part
of it.  But for now, I think this is appropriate for a separate RFC,
that will get incorporated into the core CCID3 document whenever
CCID3bis is undertaken.  (As an example, it would be a drag
if whenever one wanted to add new functionality to TCP congestion
control, one had to start a new bis of a single document that
encompassed all of TCP congestion control (e.g., initial windows,
timers, highspeed TCP, SACK options, responding to dup acks, etc...)

- Given that it is a standalone document I think it could be clarified
a little. In Section 3.1 it discusses how the content should basically
match loss intervals, which does make sense. I think it should make
explicit reference to section 8.6.1 of RFC4342 though. In particular
it should discuss how many loss counts to send.

That sounds ok to me.  Section 3.1 says the following:

    "Dropped Packets options SHOULD be sent in tandem with corresponding
    Loss Intervals options...  When this receiver sends a feedback
    packet containing information about the N most recent loss intervals
    (packaged in one or more Loss Intervals options), it SHOULD include
    on the same feedback packet one or more Dropped Packets options
    covering exactly those N loss intervals."

So all that it needs is an explicit reference to Section 8.6.1 of RFC 4342.

I was wondering
whether we needed to potentially retransmit loss counts, but seeing
that loss intervals in effect keep on getting sent until we realise we
have an acknowledgment the same applies to loss counts in effect - I
just think this should be made explicit somehow.

Sure.

- Sally
http://www.icir.org/floyd/



[Index of Archives]     [Linux Kernel Development]     [Linux DCCP]     [IETF Annouce]     [Linux Networking]     [Git]     [Security]     [Linux Assembly]     [Bugtraq]     [Yosemite]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Linux SCSI]     [DDR & Rambus]

  Powered by Linux