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

Re: Moving forward on improving HTTP's security

From: Mike Belshe <mike@belshe.com>
Date: Wed, 13 Nov 2013 14:32:19 -0800
Message-ID: <CABaLYCs6-s+Q0rBPSxGVy3X6VBDcWuG_dATJjcOvt52Jz=rd-g@mail.gmail.com>
To: Willy Tarreau <w@1wt.eu>
Cc: Stephen Farrell <stephen.farrell@cs.tcd.ie>, "William Chan (?????????)" <willchan@chromium.org>, Tao Effect <contact@taoeffect.com>, Tim Bray <tbray@textuality.com>, James M Snell <jasnell@gmail.com>, Mark Nottingham <mnot@mnot.net>, HTTP Working Group <ietf-http-wg@w3.org>
On Wed, Nov 13, 2013 at 2:07 PM, Willy Tarreau <w@1wt.eu> wrote:

> On Wed, Nov 13, 2013 at 09:57:46PM +0000, Stephen Farrell wrote:
> > > Huh ? No. I mean "The TLS model is fine for me as long as it's used
> where
> > > needed and if it's not abused because I expect all actors in the chain
> to
> > > care about security". Let's ensure we don't break that weak link from
> the
> > > root CAs to me by making its use mandatory for all no-value stuff that
> > > nobody cares about and which will make it normal for everyone to deploy
> > > broken configs and rogue CAs everywhere for the sake of simplicity.
> >
> > Break the link by making it mandatory sounds like wild supposition.
>
> Well, TLS was supposedly unbreakable till it became the norm to break
> it on MITM proxies in companies. When there's a good reason for doing
> it, the adequate methods are deployed. Whether they are "you just need
> to install the attached certs in your browser to get rid of the warnings
> when you're browsing" or "you may only use the browser preinstalled on
> the PC".
>
> Right now there's no motive for doing so. When ISPs with small links and
> big caches will see they have two choices :
>   - send a cert to all their customers
>   - multiply their bandwidth by 10
>
> Do you really think they'll pick the second one ? No, they'll do the first
> one and only multiply the pipe by 2 do handle the few users who accept to
> pay more for getting rid of the cache without sacrificing the security. It
> is very simple, users will definitely accept this en masse because they
> don't care. It already works perfectly in large companies and everyone is
> happy with that. And better, with most of the bandwidth going to
> smartphones,
> themselves massively sold by mobile providers, it will be transparent for
> the
> user, the phone will come preinstalled with the "valid certs" and it will
> be
> mentionned in the contract that the ISP reserves the right to see the
> traffic
> in cleartext for law enforcement and everyone will accept except a few,
> just
> the same that absolutely want to get the sources of every component in
> their
> phones and which no ISP wants to have as customers.
>
> I don't see how hard it is to understand in fact :-/
>

Look, we've had this debate time and time again and its always the people
with vested interests that are against TLS.  I have yet to hear from a
single person that is against TLS who isn't either a hacker, a government
agent, or a seller of software which relies on unsecured traffic.  Not one.
 Actually, the hackers don't care that much.

I do hear what you're writing, that you think use of more TLS will somehow
cripple existing TLS, but you're ignoring that it is hackable now...  Our
use of it doesn't change that.  Despite shortcomings, we do need to raise
the bar -  there is real, documented evidence of that.  And TLS will evolve
too, and we (http) will evolve with it.

Upwards and onwards!

Mike




>
> Willy
>
>
Received on Wednesday, 13 November 2013 22:32:47 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 11:11:19 UTC