Minutes IETF97: perc
Privacy Enhanced RTP Conferencing
||Minutes IETF97: perc
Privacy Enhanced RTP Conferencing
Session 2016-11-15 1330-: Studio4
Chairs :- Richard Barnes, Suhas Nandakumar
Scribe: Mo Zanaty
Jabber: Matt Miller
* draft-ietf-perc-double, draft-ietf-perc-srtp-ekt-diet,
- Authors discussed open issues and pending items from IETF 96
- WG decided to progress these drafts to WGLC (to be finalized on the list)
- Authors presented updated spec with TLS as the tunnel protocol
- Authors to resolve pending open issue on Conference ID usages
- A design team consisting of Cullen Jennings, Adam Roach, Roni Even,
Christian Groves has been asked to explore the interactions between the
signaling flows and the tunnel spec - The design team is expected to produce
recommendations on the same for IETF 98 - WG decided to adopt this spec as
Working group draft. (to be confirmed on the list)
* Chairs to send out WGLC calls for double, ekt-diet and the
private-media-framework specs * Chairs to send out WG adoption call for tunnel
Notetaker: Mo Zanaty
No major issues or decisions.
SRTP Double (Cullen Jennings)
Cullen: No major open issues.
EKT on Diet (Cullen Jennings)
Cullen: Any objection to adding explicit (rather than implicit/computed) size
in EKT long field? Russ Housley: More straightforward to have explicit sizes.
Cullen: Ready for WGLC? Chair: Defer WGLC to end of session.
DTLS Tunnel (Adam Roach for Paul Jones)
Sean Turner: Frivilous comments in jest.
Richard Barnes: Is there only UnsupportedVersion?
Adam: At the top there is a version field for which version you do support.
Nils: Conference ID establishment depends on who needs to know it.
Option 1 is MDD sends Conference ID to KMF.
Option 2 is KMF assigns Conference ID and sends to the MDD.
Option 3 is no Conf ID at all. Requires a participant to use a different
certificate in each conference. Cullen: Is there an Option 4 to allow all of
1-3? Adam and Nils: That adds too much complexity. Roni Even: There is a case
where the KMF is also an endpoint in the conference. Whoever creates the
conference and knows who is allowed in it should set the Conference ID, i.e.
the KMF in option 2. Richard (from floor mic): Inclined to avoid imposing
requirements here. Cullen: Argument against option 3 and 4. Agree with Roni
suggesting option 2. Richard (as chair): Adopt as WG doc? Unanimous hums to
adopt. Will confirm on the list. DECISION: Adopt draft as WG doc.
PERC Framework (Christian Groves)
Richard: Framework refers to tunnel draft for Conference ID details. Is it
consistent with the prior discussion? Christian: It just refers to the tunnel
draft without assuming how the Conference ID is established. So it should be
consistent with whatever outcome is decided for this.
Chairs: What do we want to work on, and when are we done?
Adam: Before finalizing, we first need to understand all the signaling and data
flows. Richard: What about interest in implementing this? Adam: We started work
to put a KMF in Firefox. EKT needs to be finalized. There is interest to
implement, but no resources committed. This is the client perspective. We would
like to see MDD implementers come forward. Sean: Strive for completion. Cullen:
Would like to move the docs forward now, even if they need to be revisited
later. Where do we park them? With editors, chairs, ADs, RFC queue? Russ: Park
them before IETF LC to avoid pulling back and having multiple IETF LCs. Alissa
Cooper: IESG should review once, so park them with WG chairs. Cullen: EKR
supports moving forward with WGLC, but may want to review EKT more critically
during WGLC. Adam: IETF process should not leak into wire protocols. Richard:
We have a version field, use it! Cullen: Current docs support interop between
MDD and KMF from one vendor and endpoints from a different vendor. But not MDD
and KMF from different vendors, which needs signaling specs for interop. Adam:
Cullen and I will sign up to write a signaling draft (for SIP or WebRTC
contexts) for IETF 98. Christian: Need WebRTC signaling to flesh this out.
DECISION: Progress EKT, double, and framework drafts. Wait for signaling draft.
Leave tunnel open for now until signaling draft is available. Chairs: Need
reviews for: EKT and Double: Sean and Russ Tunnel: Roni and Cullen Framework:
Roni and Adam