W3C home > Mailing lists > Public > ietf-http-wg@w3.org > April to June 2014

Re: #540: "jumbo" frames

From: Poul-Henning Kamp <phk@phk.freebsd.dk>
Date: Wed, 25 Jun 2014 17:47:28 +0000
To: Greg Wilkins <gregw@intalio.com>
cc: "K.Morgan@iaea.org" <K.Morgan@iaea.org>, Mark Nottingham <mnot@mnot.net>, Willy Tarreau <w@1wt.eu>, HTTP Working Group <ietf-http-wg@w3.org>, Martin Dürst <duerst@it.aoyama.ac.jp>
Message-ID: <1156.1403718448@critter.freebsd.dk>
In message <CAH_y2NEWTUjdsLr-tDbF7fRLkNWsojCR7imKMc-9fBU3cT1g7Q@mail.gmail.com>
, Greg Wilkins writes:

>Jumbo frames could easily be achieve without and extension bit if we change
>the frame header to:

I agree that it would be cleaner to totally redesign the frame-header,
but we have to pay some respect to the fact that some people think
we're ready for Last Call.

My proposal was intended to slot as seamlessly in as possible for
that reason:

* Define one of the reserved bits as adding an 8 byte network
  byte order length field in front of the payload.  (No need to
  make it terribly complicated, 8 bytes out of 16KB+ is < epsilon.)

* Add SETTINGS defaulting to 16KB, and mention that using the
  extention length is only legal if SETTINGS > 16KB has been
  received.

* Remove every mention of CONTINUATION

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk@FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.
Received on Wednesday, 25 June 2014 17:47:55 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:14:31 UTC