Re: [TLS] Protocol Action: 'IANA Registry Updates for Transport Layer Security (TLS) and Datagram Transport Layer Security (DTLS)' to Proposed Standard (draft-ietf-tls-iana-registry-updates-05.txt)

Well, this is a bit premature because the document hasn't actually been
published, just approved.

In any case, I don't think we should assign code point 26 to this
extension. I recognize that you have existing implementations that happen
to use it, but that's a result of the unfortunate decision to squat on a
code point which was right in the way of near future assignments, and those
implementations can change to the new code point. Of course, it might be
useful to add a note to implementations of the compression draft as well.

-Ekr

On Sun, May 27, 2018 at 5:49 AM, Peter Gutmann <pgut001@cs.auckland.ac.nz>
wrote:

> The IESG <iesg-secretary@ietf.org> writes:
>
> >The IESG has approved the following document:
> >- 'IANA Registry Updates for Transport Layer Security (TLS) and Datagram
> >   Transport Layer Security (DTLS)'
> >  (draft-ietf-tls-iana-registry-updates-05.txt) as Proposed Standard
>
> Now that it's been published, I'd like to request assignment of an
> extension
> ID for tls_lts under the Specification Required terms.
>
> In terms of the conflict between the existing use of 26 = tls_lts and the
> new
> 26 = compress_certificate, this can probably be managed by noting that
> compress_certificate will only be sent by TLS 1.3 implementations and
> tls_lts
> will only be sent by TLS 1.0-1.2 implementations.  In addition
> compress_certificate has a nonzero size (a list of algorithms is sent)
> while
> tls_lts has a zero size (it's an indicator flag).  This means that the two
> different uses are unlikely to occur in the same implementation, and even
> if
> they do can be distinguished by looking at the extension data size.
>
> Having compress_certificate moved to 27 would be a better solution in
> terms of
> existing use, but I can add a note to the LTS draft mentioning the change.
> The
> certificate compression draft should probably also be updated to indicate
> that
> a zero-size compress_certificate, if encountered, isn't a
> compress_certificate
> but a tls_lts.
>
> Peter.
>
> _______________________________________________
> TLS mailing list
> TLS@ietf.org
> https://www.ietf.org/mailman/listinfo/tls
>