W3C home > Mailing lists > Public > ietf-http-wg@w3.org > July to September 2004

Re: Is forwarding hop-by-hop headers a MUST-level violation?

From: Alex Rousskov <rousskov@measurement-factory.com>
Date: Mon, 12 Jul 2004 12:59:45 -0600 (MDT)
To: Jamie Lokier <jamie@shareable.org>
Cc: Jeffrey Mogul <Jeff.Mogul@hp.com>, ietf-http-wg@w3.org
Message-ID: <Pine.BSF.4.58.0407121251000.37360@measurement-factory.com>


On Mon, 12 Jul 2004, Jamie Lokier wrote:

> Jeffrey Mogul wrote:
> >        All hop-by-hop headers (other than Connection) MUST be listed
> >        in the Connection header.
>
> Again, existing implementations don't list Transfer-Encoding in the
> Connection header.  This is very common and conforming behaviour.

Understood. Most existing implementations do not list most of the
explicitly documented hop-by-hop headers in the Connection header.

> Is this proposal meant to imply that all such existing
> implementations are no longer conforming?

Yes. Why is that a problem? We are changing the old protocol (RFC
2616) to fix a bug. Old implementations might be compliant with RFC
2616 (the vast majority is not). New or updated implementations may be
compliant with RFC XXXX. The change does not break old
implementations, and might actually help some. I do not see a
compelling reason to make an exception for Transfer-Encoding. Do you?

Note that several HTTP errata entries would have a similar effect of
making implementations based on old/buggy RFC 2616 incompliant with
the updated HTTP/1.1 standard. That fact alone does not prove me
right, of course.

Thanks,

Alex.
Received on Monday, 12 July 2004 14:59:53 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 06:49:35 GMT