This is worth working on.

When I worked on this problem in Rust, I used Cloudflare's ad-hoc Go
serialization, although it looks as though they've changed it since[1].

I wasn't really sure how these things should be stored relative to
ECHConfig ids etc, but I came up with a different answer than this draft
does (not that I feel strongly) [2].

thanks,
Rob

[1]
https://github.com/cloudflare/go/blob/7d4ce7c5377c6bf6ff5977d4734d6f6f2a10ccdb/src/crypto/tls/ech_config.go#L16
[2]
https://github.com/rustls/rustls/pull/663/files#diff-62731cbc005ac23c9805e5f34445ff0d9650b4077f8e35a3fdbe5a885986a779R65

On Fri, Nov 19, 2021 at 6:38 PM Stephen Farrell <[email protected]>
wrote:

>
> Hiya,
>
> As discussed at IETF112, I've updated the PEM file format
> draft for ECH. [1] Happy to take comments via mail or via
> that github thing:-)
>
> As and when the chairs think it's a good time to consider
> adoption, or incorporation into the ECH draft, or sending
> this somewhere else, I guess we can have that discussion.
>
> This format is supported by my OpenSSL fork [2] which has
> CLI tooling for making/consuming these and can be ingested
> by the various TLS servers (lighttpd, apache, nginx and
> haproxy) with which I've integrated that ECH-enabled build.
> (There're pointers to forks for those servers at [3].)
>
> I'll work on the well known url draft in the next while.
> It likely needs a bit more work than simple substitution.
>
> Cheers,
> S.
>
> [1] https://datatracker.ietf.org/doc/draft-farrell-tls-pemesni/
> [2] https://github.com/sftcd/openssl/tree/ECH-draft-13a
> [3] https://defo.ie/
> _______________________________________________
> TLS mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/tls
>
_______________________________________________
TLS mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/tls

Reply via email to