Hi,

> On 19 Mar 2016, at 11:20, Mark Risher <[email protected]> wrote:
> 
> Dear UTA:
> We have just submitted a proposal entitled SMTP Strict Transport Security, 
> which details a mechanism for protecting MTA-to-MTA email traffic against TLS 
> downgrade attacks and interception.
> 
> The initial draft is at 
> https://datatracker.ietf.org/doc/draft-margolis-smtp-sts/ and we hope to 
> discuss this at the Buenos Aires meeting next month. While we have deployed a 
> prototype/reference implementation among the authors, we are very open to 
> feedback and suggestions from the broader group and look forward to your 
> input.

As already mentioned in two previous threads to this mailing list and a GitHub 
issue: I see a big problem with deployment of the webpki authentication part 
outside of big hosting environments. But I really do like the idea of getting 
feedback from/to MTAs. Can you guys think of a way we can get in-band reporting 
within e.g. SMTP (an extension)?

EFF is currently picking up STARTTLS-Everywhere again (mind: it's been idle 
since before Let's Encrypt, a lot of new proposals like this one for mail came 
out - so the project, scope and even it's name are subject to change in the 
near future). One idea I had was to provide anonymised log information 
regarding TLS-failures / MITM attempts to distributed monitoring servers and 
the receiving party. If there were a more general document for error-reporting 
(as some suggested) - we could maybe make use of that as well. If it requires 
every small mail-OP to set up a webserver on the same machine, it's unlikely 
that we can push for that, and adoption number might turn out low.

As for authentication/TOFU: I see no better proposal than TACK around. HPKP was 
an utter failure [0] and I fear that, for the very same reasons, SMTP-STS might 
not see a lot of deployment nor use in single-domain-setups. And there're quite 
a few around. While it's important to secure the majority of mail traffic 
between large providers - they're already doing well - we should not forget 
that there's still a lot of self-hosted MXs on the Internet.  I could imagine 
an updated (e.g. using EdDSA over ECDSA etc) version of TACK fitted to SMTP. 
One of the original authors might be interested in working on that he told me 
earlier this year, unfortunately I did not have any time to pursue this idea 
further.

Thanks,
Aaron

[0] 
http://news.netcraft.com/archives/2016/03/30/http-public-key-pinning-youre-doing-it-wrong.html

Attachment: signature.asc
Description: Message signed with OpenPGP using GPGMail

_______________________________________________
Uta mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/uta

Reply via email to