IETF-67 IP over Digital Video Broadcast (IPDVB) WG Internet Area ipdvb WG Chair: Gorry Fairhurst (GF) gorry@erg.abdn.ac.uk ipdvb Note-Taker: Martin Stiemerling (MS) 1. Agenda Bashing (5 minutes) - Chair Agenda accepted with small change; the following was not presented, because there were no changes to the draft. ULE Security Extension (draft-cruickshank-ipdvb-sec-02.txt) 2. Document Status (5 minutes) - Chair * Milestones. * Documents in Last Call - None. * Documents in IESG/AD Review: http://www.ietf.org/internet-drafts/draft-ietf-ipdvb-ar-05.txt * Documents in RFC Editor Queue - None. * Published RFCs - None. Gorry Fairhurst (GF) gave an overview of working group status: - 2 pub RFCs - 1 doc in RFC last call - List of individual drafts - Milestones (most done; security requirements, ULE security requirements and address resolution are behind schedule) - GF reminders about re-charterting 3. AR Draft (Marie-Jose Montpetit, MJM) Started this work two years ago, now complete. This is now in IETF last call. There were questions. 4. Header Compression over ULE (John Border, JB) John presented a first draft on header compression issues, this was designed to stimulate discussion on this topic - there had already been some comments on the mailing list, including the position for IPv6 and multicast, features that will be added by the authors. More ideas and comments were invited. Carsten Bohrmann (CB): ROHC over X spec needs: A method is needed to communicate between compression and decompression to setup the context. We also need do define an encapsulation encoding, which could be as simple as a ULE-Type. This lets the receiver know what is being received (compressed or uncompressed). The more complicated question is negotiation: - You may not wish to negotiate (just announce using a simple UDP format, or using SI - both possible). - Actual negotiation needs a negotiation protocol, but not obvious for broadcast, so use-case would be the first thing to establish. GF: Are you negotiating profiles? CB: Yes the sender could offer 1,3,5 and the receiver only support 1,3. Another issue for bidirectional HC is how to bind the feedback to the flows. Finally, we could define ROHC profile that could include some headers from the ULE-level, if that were useful. JB: There was an I-D for ROHC applicable to ULE 5 IETFs ago, this was a very simple one, what is the status? CB: This used LLC encapsulation, aimed at WiFi. It was not published, and is not optimal for ULE. JB: Is someone willing to help on this draft? CB: Yes, I am willing to help on ROHC over ULE. GF: We also have interest in this topic. GF: Have other methods in the IETF been considered, beside ROHC? e.g. eCRTP. John: We are not currently taking other methods into account, but it may be "nice" if this were generic. GF: We are not short of ULE-Types, so we could use an alternate type for other methods. CB: ROHC is both a framework and a set of profiles. ROHC supplies this framework, into which other things can be put, e.g. support for "legacy" methods, such as eCRTP, but this currently needs a small amount of work. It would then be easy to take your compression scheme of the month and make your ROHC profile. CB: We need input from people designing related products, saying that they want to have this compression work done. JB: We are interested. CB: We work with people who are interested. GF: Yes we have interest and I know people who are also interested, see mailing list. We can't do much right now, the WG is not charted to work on header compression. We could can the IESG for this in a future charter. GF: John, do you intend to revise this draft? JB: Yes, we already have some corrections to obvious stuff and would like to add multicast support. We shall probably wait for more inputs before we discuss mechanisms. CB: I can certainly write an I-D on some of these, but more important to get the goals correct. I can also send an email to start the Chartering discussion by suggesting what documents should be written. GF: I'd like to see more discussion of this on the list. 5. ULE Security Requirements (presented by MS on behalf of authors) There had been two revisions since the last IETF, it now addressed the issues originally requested by the Security Directorate. The document received support from those who had read it the last IETF meeting, but the AD then questioned if the scope was included in the Charter. He has since noted that this document does reflect the current milestone, and could now be adopted if the WG chose. If we proceed, a Security Directorate review will be requested during the WGLC. GF: Would anyone like to speak about adopting this? Martin Stiemerling (MS): This is in pretty good shape and in my opinion, should be adopted. GF: I am starting a formal two-week call for feedback. Comments on whether this should proceed, or not proceed, should be sent to the ipdvb list. Jabber (Michael Noisternig): Comments were sent to the list after rev 03 of the draft had been published. I see some comments have not been taken into account. GF: Comments received on the list should be included, or addressed by the authors. WG items need to reflect what is said on the list. (Prashant in reply to Michael via Jabber): We think we have updated all references and editorial changes according ti the email. We will look again at the email to see if we may have missed any points. GF: Yes - please do, and talk to Michael via the list to resolve this. GF: I have comments too, on the latest rev. GF: I urge the WG to read document and send comments to list. 6. Extension Formats for the ULE Encapsulation (Gorry Fairhurst) This I-D is an individual submission, aimed to provide useful features for MPEG-2, but also good for second-generation, that do not necessarily wish to use Transport Streams. CB: When sending several PDUs in a single SNDU the last PDU-element's length info is redundant, since the length may be implied from the total SNDU Length. GF: Yes, could be optimized, but not sure if needed, this is easier to implement, and already saving bytes. CB: Yes, I was thinking of Header Compression again! GF: There are other extensions for MPEG-2 SI, more could be defined. I'd really like a co-author to volunteer, especially one with implementation experience. 7. ULE Implementation Status (Gorry Fairhurst) The University of Salzburg provided an implementation report about the ULE decapsulator in Linux (see slides). This implementation is now seeking RFC compliance (see bug fixes), and there is a further interest in developing flexible efficient support for extensions. 8. GF provided a quick summary of the current WG documents: One document is in IETF last call (AR). A new document is now requested for adoption (Sec Req). The new subjects will require a revised charter, and this needs to first be discussed by the WG. Is there interest in and energy to progress these new work items? The WG meeting closed at 4pm.