Re: [perpass] Unauthenticated, ephemeral keying in HTTP/1.0 without TLS

[Stephen Farrell <[e-mail]>]

Folks,

On 11/16/2013 09:47 PM, Phillip Hallam-Baker wrote:
> I like TLS everywhere with strong authentication. The idea of weakening the
> authentication requirements further and calling the result TLS worries me a
> lot.

TLS has include anon-DH from the get go. Self-signed certificate
deployments with TLS amount to 10% of all web sites (says [1]).
Only about 3 times that number use certs that chain up to a
browser-trusted CA of one sort or another. So a premise that
all TLS deployments have strong server authentication today is
wrong. And I've objected to PHB saying "weakening" elsewhere, [2]
so I won't repeat that here.

And to the extent this discussion is about http/tls, that
belongs on the httpbis wg list, where there's been a firestorm
of discussion on exactly that topic. So, please let's not start
another http/tls thread here, which is what this seems to be turning
in to, even though Phill asked something else.

Other foo/tls protocols will also soon have a separate venue [3]
and we have a TLS working group. So I see little left to discuss
about TLS on this list to be honest.

Finally, I note the subject line here said "without TLS" and
not "debate TLS on yet another list" :-)

S.


[1] http://w3techs.com/technologies/overview/ssl_certificate/all
[2] http://lists.w3.org/Archives/Public/ietf-http-wg/2013OctDec/0928.html
[3] https://datatracker.ietf.org/doc/charter-ietf-uta/