It is no longer true that tinkering with the Internet at the edges is easier
than the middle. It has not been true for at least a decade.
It is utterly impractical to change the core Internet. Changing the edge is
very very slow.
The only place where change is practical is the interface between the network
and the Internetwork.
This means abandoning the myth that hosts connect to the Internet, they don't
and they won't in future either.
If we are to have security we have to apply Butler Lampson's concept of a
security reference monitor and realize that in the network context this is a
firewall or other edge security device.
We could change the S/MIME spec but that would eliminate the advantage of using
S/MIME and create even more problems as legacy S/MIME clients misbehave when
they see the new S/MIME. S/MIME does not cope with upgrades gracefully.
Introducing a parallel spec is far more effective and simpler.
-----Original Message-----
From: Yakov Shafranovich [mailto:research(_at_)solidmatrix(_dot_)com]
Sent: Wednesday, January 28, 2004 7:17 PM
To: Hallam-Baker, Phillip
Cc: ASRG
Subject: Re: [Asrg] Re: 2. Uselessness of C/R
Hallam-Baker, Phillip wrote:
While we are on the topic of S/MIME: currently majority of
MUAs have
S/MIME support built-in including root certificates. Why is that no
banks or financial companies that are suffering from "phishing"
attacks, consider signing their email via S/MIME?
I know several banks that are considering it. The
disadvantage is that there
are email users with MUAs that don't handle S/MIME. The big
problem is that
Eudora is effectively an orphan code-base with little
serious development
work.
Any ideas on what is the percentage of users that do not have
S/MIME? If
MSFT, Mozilla, etc. and the other MUAs cover a virtual
majority of the
market, and would cover a majority of users affected by the phishing
attacks, why aren't the banks deploying it? It would be
easier to tinker
with the edges of the network, rather than the center.
There is a private working group looking at this. Yahoo!
Domain keys looks
like a better fit for what it is intended to achieve.
Wouldn't a profile of S/MIME that stores keys in DNS achieve
essentially
the same thing?
Yakov
-------
Yakov Shafranovich / asrg <at> shaftek.org
SolidMatrix Technologies, Inc. / research <at> solidmatrix.com
"Power tends to corrupt, and absolute power corrupts
absolutely" (Lord
Acton)
-------
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg