Re: Securing Internet mail at the MTA level

SSH, while a quite useful tool, is not the right long-term solution for
transport layer security - IP security is.

It's also clear to me that for E-mail, you don't want transport level
security for the system; you want "object" security, that is, digital
signature and encryption of the mail message. That way, no matter how many
MTAs the message passes through (and pretty much regardless of how
trustworthy they are) you have end-to-end authentication, integrity and
privacy.

        Erik Fair <fair@clock.org>

"Erik E. Fair" writes:
> SSH, while a quite useful tool, is not the right long-term solution for
> transport layer security - IP security is.

Agreed. (After all, I'm one of the creators of the spec, so how
couldn't I agree :) However, SSH is a neat hack for today -- I use it
for connecting over the net to machines where I can't hack the
operating system but can drop SSH on. An SSH like interface (actually,
Berkeley r-command like interface) will still be needed into the
future, btw.

> It's also clear to me that for E-mail, you don't want transport level
> security for the system; you want "object" security, that is, digital
> signature and encryption of the mail message.

Yup. This is a frequently missed point. Link security and object
security have different uses at different times -- and people confuse
them way too often.

Perry

you really need both object and transport level security; the transport 
level stuff helps protect against traffic analysis; the real 
authentication and privacy coming from the object level stuff.

Simon

---
Cause maybe  (maybe)		      | In my mind I'm going to Carolina
you're gonna be the one that saves me | - back in Chapel Hill May 16th.
And after all			      | Email address remains unchanged
You're my firewall -    	      | ........First in Usenet.........

While I agree in general, I think I'd rather see effort spent on
getting everyone on message security first, and worry about traffic
analysis later. My fear is that, having attempted (since it's unlikely
we'll ever manage to get 100%) to secure the transport, people will
stop worrying about message security, and let that slide.

To put it another way, the first order of business is to prevent
everyone from reading the mails. We can afford to worry about second
order effects like traffic analysis after message security is well
under way (say, 50% of all nominally private message traffic is
encrypted).

Or, put it yet another way: time for everyone to host a PGP key
signing party for your friends, neighbors, and co-workers!

	Erik E. Fair    fair@clock.org