Routing Area Open Meeting (rtgarea) IETF 99 (Prague, CZ) =============================================================================== Area Directors: Alia Atlas (akatlas@juniper.net) Deborah Brungard (db3546@att.com) Alvaro Retana (aretana@cisco.com) Area Secretary: Jonathan Hardwick (jonathan.hardwick@metaswitch.com) Wiki: https://trac.tools.ietf.org/area/rtg/trac/wiki/WikiStart Scribe: Jonathan Hardwick (jonathan.hardwick@metaswitch.com) Location: Congress Hall II, Hilton Prague, Prague, CZ Time: July 19, 2017, 0930-1200 (9:30am-12:00pm) ------------------------------------------------------------------------------- 1. Administrivia ---------------- Alia Atlas: There are new guidelines for YANG model authors - the Network Management Datastore Archutecture (NMDA). This has been discussed in RTGWG and NETMOD earlier this week by Robert Wilton. YANG models should transition to be compliant with the NMDA guidelines ASAP. New YANG models SHOULD follow NMDA from their inception. ------------------------------------------------------------------------------- 2. The New Note Well and Sanctions Available for Application to Violators of IETF IPR Policy ------------------------------------------------------------------- Lou Berger & Adrian Farrel Adrian Farrel: I am not a lawyer - I am expressing an opinion as an individual. For proper legal advice, please see a lawyer. Lou Berger: When am I obligated to disclose IPR? Adrian Farrel: As SOON as possible. Lou Berger: George says "Before making a contribution, or as I am making a contribution." Adrian Farrel: That would be safest. Daniele Ceccarelli: What if I don't think my patent affects something but the IETF does? Who can I check with? How can I know what the IETF thinks? Lou Berger: Should never occur. The IETF should not make any judgment on whether IPR applies. Not for the IETF to decide. Individuals may make a disclosure but the IETF does not judge - that is the job of the courts. Loa Andersson: So should we do an IPR call on every doc sent to the WG? Adrian Farrell: See RFC 6702, which makes it clear that you don't ever need to do an IPR poll on anything. WG chairs apply this differently depending on the WG's history. Some don't do polls, because the rules are already clear. Others do polls at key times as a reminder. But to poll for every doc that is sent in the WG's direction would probably be too disruptive to our process. Lou Berger: The shepherd write-up template includes a section covering whether there is IPR and whether an IPR poll has been issued to the WG. There is some onus on the shepherd to make sure everyone has followed the disclosure process. Lou Berger: If you contribute on a topic, then you are obliged to follow the disclosure process. Loa Andersson: If we ban someone from posting to a mailing list as a sanction, then we stop them from doing future IPR disclosures. It is a blunt instrument. Adrian Farrel: That's not true. Disclosures are made through a web page, not on the list. Anyone can read a mailing list, even if they are banned from posting. Alvaro Retana: We think that polling for IPR and encouraging participants to remember the disclosure process is a good practice to follow. Keyur Patel: It would be good to call for IPR earlier in the WG than during the last call. Implementers need to know. Adrian Farrell: Some WGs do an IPR poll at both WG adoption and WG last call. Jeff Tantsura: What does a WG do if IPR is disclosed? How do they discuss whether they like it or not? Lou Berger: The IETF does not judge the validity of an IPR claim, only a court can do that. WG participants can take any step to voice opinion on the work covered by an IPR disclosure. Some groups push towards open licensing and avoid encumbered technologies. In a particular WG, if there is an IPR disclosure, the group can decide whether to proceed with the technology or draft based on the disclosure. It is a consensus process. Adrian Farrel: It is impossible to engineer a solution around IPR that you can't see. One factor is knowing what the licensing terns are. If a licensor has strict terms then people may engineer around it; if terms are liberal then the decision may be not to work around it. But if the details of the IPR are hidden then it is a problem. Lou Berger: This was just two people giving their personal opinions. Talk to your own legal department if you need actual legal advice. ------------------------------------------------------------------------------- 3. Working Group and BoF Reports -------------------------------- 3.1 WG Chair Reports -------------------- See the wiki for the content of the reports. Some comments below that were not reflected in the wiki. https://trac.ietf.org/trac/rtg/wiki/IETF99summary BABEL report (Russ White) Discussed unicast hellos; a few updates to base spec; presentation on BIER but we side-lined it until we see requirements. Simple meeting, lots of energy and cool ideas. L2TPEXT (Deborah for chairs) Not meeting; have 1 new RFC on IPv6 tunnel MANET (Alvaro for chairs) Not meeting this week. There is steady progress on DLEP extensions. Lou Berger: Neither chair is here but there are still corridor conversations about DLEP etc. SIDR (Alvaro for chairs) Sandy is not here, but Chris is here. SIDR is not meeting; their work is pretty much finished. 8 or 10 documents are with the RFC Editor in AUTH48 state. They are hoping to publish everything as RFCs in the next few months. Alvaro Retana: It's great to do this status update because it is a facilitator for WG cross-pollination. 3.2 Routing Directorate Report ------------------------------ Jon Hardwick Adrian Farrel: Is it reasonable that 58% of drafts have issues when they are reviewed by the directorate? Alvaro Retana: Many of these are early reviews and it is expected that more issues will be found at that stage. Jon Hardwick: I could present stats on issues raised on early reviews versus last call reviews. They may break down differently. Alvaro Retana: Nevertheless, there are too many +1s for WG documents during last call. It would be good if WGs reviewed their docs more thoroughly. Loa Andersson: Routing directorate work has improved in the last year or so. We are getting good quality, proper reviews. Thank you to the directorate. Chris Hopps: Can we examine the stats for issues raised on reviews from outside the WG versus in? Jon Hardwick: Yes, but to make it statistically significant we need to look over a long period as we don't get too many of those reviews. ------------------------------------------------------------------------------- 4. Open Discussion / Any other business ---------------------------------------- John Scudder: Code point squatting. I suggest that, maybe at the periodic chairs meeting, we talk about code point allocation. Surprised by IETF old timers who differ on how the IANA allocation policies work. I'm concerned we're not making enough use of IANA to do work for us. We are trying to do it ourselves and are not as good at it as they are. Alvaro Retana: Jeff Haas did a presentation on IANA allocation policies a few IETFs ago. Please look up the materials. Beyond that, the RFC for IANA considerations was just made a -bis so perhaps we should get IANA to come and give us a refresher. John Scudder: Great. I remember Jeff's talk. We need to focus education on the IANA part. Rick Taylor: MANET. We recently published RFC 8175 (DLEP) which has applicability outside of MANET. DLEP is of interest to BFD, OSPF and ISIS once you get outside the fixed network. Please have a look at it. As Alvaro says, some WGs are doing work beyond their original niche. and other WGs can benefit from this Alvaro Retana: Justin & Stan did a presentation on DLEP a few IETFs ago. Please find those slides. In ROLL, they have been discussing multicast wireless environments. 6LOWPAN are discovering mechanisms that could solve some of our problems. The more we talk to each other and re-use solutions, the better. Alvaro: What is in the future for the routing area? Rick Taylor: As chair of delay tolerant networking (Transport area). We are looking at how you move packets around in heavily delayed and disrupted networks. We are transport focussed, but need to solve routing problems and we would love to see routing guys come and help us solve these problems. Alia Atlas: Thanks for thiis suggestion. Talk to Chris Bowers and Jeff Tantsura, as it would be a great discussion to bring to RTGWG. Alvaro Retana: There is a WG called ipwave (vehicle-to-vehicle) which originally had routing items in their charter. They are considering how to connect cars to the network infrastructure etc. The ipwave charter eventually did not include routing but there will need to be a place to discuss those issues. Alvaro Retana: In concusion, please take a look at DTN and IPWAVE Jeff Tantsura: In RTGWG, we are discussing new approaches to data centre routing. We will need to work with ETSI. Alvaro closed the meeting at 11:15am.