W3C home > Mailing lists > Public > ietf-http-wg@w3.org > January to March 2012

Re: HTTbis spec size, was: Rechartering HTTPbis

From: Carsten Bormann <cabo@tzi.org>
Date: Mon, 6 Feb 2012 23:02:13 +0100
Cc: Ted Hardie <ted.ietf@gmail.com>, Henrik Nordström <henrik@henriknordstrom.net>, Julian Reschke <julian.reschke@gmx.de>, Mark Nottingham <mnot@mnot.net>, HTTP Working Group <ietf-http-wg@w3.org>
Message-Id: <E2E208A8-0E8A-43EC-BAC0-A50A35BEFFE5@tzi.org>
To: "Poul-Henning Kamp" <phk@phk.freebsd.dk>
On Feb 6, 2012, at 20:26, Poul-Henning Kamp wrote:

> Obviously, HTTP over UDP would be very exposed to DoS, so it should
> probably be confined to behind firewalls and other walled gardens,
> but I can certainly see the utility for a number of cases.

If you are interested in a REST-style transfer protocol that runs over datagram transports (such as UDP and DTLS) at a low level, you may want to look at CoAP.

http://tools.ietf.org/html/draft-ietf-core-coap-08

CoAP might work well for favicon.ico :-)
(It is designed for constrained node/networks.)

Also have a look at coap-block (for payloads larger than what you like to have in a single datagram) and coap-observe (for going beyond original REST).

CoAP is different enough from HTTP that it may not be a drop-in, but if all you need is a back-end protocol behind, say, varnish, that runs on datagrams, it may be worth a look.

Grüße, Carsten
Received on Monday, 6 February 2012 22:04:39 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 06:51:55 GMT