W3C home > Mailing lists > Public > ietf-http-wg@w3.org > April to June 2013

Re: Design Issue: Which frames are allowed to reference Stream ID #0 and which aren't?

From: Martin Thomson <martin.thomson@gmail.com>
Date: Fri, 26 Apr 2013 11:38:59 -0700
Message-ID: <CABkgnnUpwM8aOzwt=Yb7_TsvNheotMYb+BD17eE7P4QaZSCdeA@mail.gmail.com>
To: James M Snell <jasnell@gmail.com>
Cc: "ietf-http-wg@w3.org" <ietf-http-wg@w3.org>
On 26 April 2013 11:28, James M Snell <jasnell@gmail.com> wrote:
> One can look at the various frame types and determine which are
> obviously intended for use with streams and which are for the session
> but we need to be clear on exactly what happens when various types of
> frames reference Stream ID #0.
>
> For instance, if a DATA frame reference Stream ID #0, I would say that
> the result MUST be a protocol error. (Likewise with HEADERS frames)

DATA on stream 0 is an error, we need to make that clear.

> Likewise, what exactly does it mean for a PING Frame to specify a
> Stream Identifier? Or are all PING Frames required to use Stream ID
> #0?

PING on non-stream 0 is an error.  The problem there is that PING
demands a PONG and you cannot know whether it is possible for the peer
to do that.
Received on Friday, 26 April 2013 18:39:26 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 11:11:12 UTC