[saag] TLS IETF 105 report

["Christopher Wood" <[e-mail]>]

The TLS WG session met on Tuesday and Thursday. There are a few issues remaining for the delegated credentials draft. An update will be issued by September, which includes renaming to Delegated Authentication Keys, at which point we will start WGLC. Formal analysis will proceed in parallel, and the draft will not advance to IESG until analysis is complete.

The WG signalled interest in adopting draft-lvelvindron-tls-md5-sha1-deprecate and draft-nir-tls-flags. Both will be confirmed on the list. The followup draft-thomson-tls-sic draft needs more discussion, development, and experimentation before being considered for WG adoption. The chairs decided that draft-camwinget-tls-use-cases was not appropriate for the WG, and advised the authors to seek AD sponsorship or an alternative publication stream. 

During the second meeting, the DTLS 1.3 was updated based on review from WG participants. There are three interoperable implementations. The draft will go through WGLC and move forward to IESG publication. Open issues in the Encrypted SNI draft were discussed. A small design team will be formed to focus on changes needed to resolve open issues and conduct formal analysis. The Connection ID presentation led to a plan to update the DTLS 1.3 and DTLS 1.2 connection ID documents. Document authors will work to update these drafts so both can move forward. The WG will consider draft-tschofenig-tls-dtls-rrc afterwards.

cTLS was presented for informational purposes. Discussion, interaction, and collaboration with the parallel LAKE effort will continue. Hybrid key exchange was presented and will continue discussion on the list. TLS Metadata for Load Balancers was presented. It is unclear if the document should be worked on in TLS or elsewhere as a generic application protocol. Related work was discussed in the past in the HTTP workshop. The IESG and IAB will need to discuss where the work should take place. The HTTPSVC record was presented and received good feedback and support of a generic mechanism. 

Best,
Chris