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

Re: Intercepting proxies - yet again

From: Nicolas mailhot <nicolas.mailhot@laposte.net>
Date: Thu, 7 Mar 2013 11:07:11 +0000 (UTC)
To: ietf-http-wg@w3.org
Message-ID: <loom.20130307T115424-976@post.gmane.org>
Hi Adrien,

You raise a lot of the same concerns than me yesterday. So I can only confirm
the need to clean the Augean stables at the protocol level now and stop
relegating it to some future date or other workgroup.

The reason people deploy intercepting proxies now is not because there is any
wish for massive surveillance, but because they work and alternatives don't. So
please make the alternatives work.

Till discovery, configuration and negotiation of intermediaries is properly
specified in the protocol, the only working solution will be interception.
Because interception is the technical solution that complies with the
"intermediaries do not exist" situation created by not specifying how to manage
intermediaries in the first place.

Explicit proxies should not mean 'proxy was configured somewhere outside the
protocol'. Client roaming is too complex to work with a static proxy definition
and the only communication channel that is sure to exist between the client and
the intermediary is http. Intermediary setup really needs to be dynamic and at
the http level (and negotiated, so the user can decide if he accepts to use an
intermediated network path or not).

Thank you

-- 
Nicolas Mailhot
Received on Thursday, 7 March 2013 11:07:49 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 7 March 2013 11:07:53 GMT