W3C home > Mailing lists > Public > ietf-http-wg@w3.org > October to December 2007

Re: NEW ISSUE: Monitoring Connections text

From: Jamie Lokier <jamie@shareable.org>
Date: Thu, 22 Nov 2007 14:07:54 +0000
To: David Morris <dwm@xpasc.com>
Cc: ietf-http-wg@w3.org
Message-ID: <20071122140753.GA13759@shareable.org>

David Morris wrote:
> Also, I thing the notion of a chunk length of -1 deserves further
> discusion. Seems to me that if a recipient understands the new value as
> abort, then the connection can continue to be used. If the recipient
> doesn't understand the signal, it should be treated as a serious error
> resulting in an abort of the request and connection. If I've not missed
> something, that leaves older recipients protected while allowing a useful
> tweak to the protocol.

I wouldn't be surprised if some implementation out there actually
parses it as a -1 length, leading to goodness knows what mayhem :-)

Any reason why "0;reason=aborted" cannot be used?  It's valid in RFC
2616 chunk-extension syntax, so anything which can accept chunked
requests really ought to accept this.

-- Jamie
Received on Thursday, 22 November 2007 14:07:59 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 11:10:43 UTC