Narrative Minutes
INTERNET ENGINEERING STEERING GROUP (IESG)
Narrative Minutes for the November 1, 2007 IESG Teleconference
Narrative Scribe: Spencer Dawkins <spencer@mcsr-labs.org>
With corrections from Jari Arkko, Michelle Cotton, Russ Housley
1. Administrivia
1.1 Roll Call
Yes, ending daylight savings time is still not an exact science ... Synchronization of the start and stop of daylight savings time around the globe would avoid confusion.
1.2 Bash the Agenda
No additions, no changes...
1.3 Approval of the Minutes
2007 10 18 Minutes - approved with no discussion
2007 10 18 Narrative Minutes - approved with no discussion, Spencer to submit as final.
1.4 Review of Action Items
o Lisa Dusseault to find an author to update RFC 3406.
Cullen and Lisa are still discussing
o Jari Arkko to write an explanation of IESG policy of when ADs can request documents be considered by the IESG before the Last Call has ended.
Not finished yet...
o Dan Romascanu to find a designated expert for draft-ietf-ipfix-info-15.txt. [IANA #106002]
Have management item for this on today's agenda
o Chris Newman to draft a policy on Autoresponse Messages sent to any IETF mailing list.
Have drafted and will review on IESG list.
Russ - thought it looked pretty good - management item for next telechat to approve this
o Cullen Jennings to get a response from the AVT WG to help solve an IANA Registration issue for wave-avi-codec-registry [IANA # 97962]
No Cullen/no update
2. Protocol Actions
2.1 WG Submissions
2.1.1 New Item
o draft-ietf-xcon-framework-09.txt
A Framework for Centralized Conferencing (Proposed Standard) - 1 of 7
Note: Adam Roach is proto shepherd
Token: Cullen Jennings
Was deferred this morning...
o Four-document ballot: - 2 of 7
- draft-ietf-simple-xml-patch-ops-03.txt
An Extensible Markup Language (XML) Patch Operations Framework Utilizing XML Path Language (XPath) Selectors (Proposed Standard)
- draft-ietf-simple-partial-pidf-format-08.txt
Presence Information Data format (PIDF) Extension for Partial Presence (Proposed Standard)
- draft-ietf-simple-partial-notify-09.txt
Session Initiation Protocol (SIP) extension for Partial Notification of Presence Information (Proposed Standard)
- draft-ietf-simple-partial-publish-06.txt
Publication of Partial Presence Information (Proposed Standard)
Token: Jon Peterson
Several DISCUSSes, Jon isn't with us.
Russ - guessing "revised ID needed" with five DISCUSSes, of course...
o draft-ietf-avt-rtp-evrc-wb-04.txt
RTP payload format for EVRC-WB codec and media subtype updates for EVRC-B codec (Proposed Standard) - 3 of 7
Note: Roni is shepherd
Token: Cullen Jennings
Russ - draft 05 has been posted. A little concerned that people prepared based on 04.
No Cullen, so no run-down.
Russ - AD followup
Magnus - 05 mostly addressing my comments, haven't checked yet.
Chris - registration ownership - "owned by WG as delegated by IESG", told by previous IESG. Correct? then won't DISCUSS based on this, worse to be inconsistent.
Sam - please discuss offline with Lisa, DISCUSSed on SIEVE based on this, she pushed back.
Magnus - authoring guidelines for how to write these documents, please tell me if there's a change.
Lisa - just didn't want to turn down documents that did what we told them to do, would love to discuss.
Russ - who gets the action item?
Chris - just cleared my current action item, so I can take this one.
Michelle - Just want to make sure the template has been reviewed on the mailing list and/or by the expert.
Russ - still goes to AD followup for this document.
Tim - IANA DISCUSS needed? They looked at version 02
Chris - assume AD's writeup is correct. will check the IETF types list.
Magnus - my DISCUSS position already covered the IANA concern.
Russ - just don't clear until you know the review happened.
o draft-ietf-avt-rtp-vorbis-07.txt
RTP Payload Format for Vorbis Encoded Audio (Proposed Standard) - 4 of 7
Note: Colin is shepherd
Token: Cullen Jennings
Number of DISCUSSes with no Cullen.
Revised ID needed? Yes.
o draft-ietf-behave-nat-icmp-06.txt
NAT Behavioral Requirements for ICMP protocol (BCP) - 5 of 7
Token: Magnus Westerlund
Was deferred last night.
o draft-ietf-avt-avpf-ccm-10.txt
Codec Control Messages in the RTP Audio-Visual Profile with Feedback (AVPF)
(Proposed Standard) - 6 of 7
Note: Roni Even is the Proto shepherd.
Token: Cullen Jennings
DISCUSS on this document with no Cullen.
Tim - suspect that mine can be cleared with RFC Editor note.
AD Followup.
Magnus - I'm the author, so...
Tim - SecDir Review went to you?
Sam - Magnus wrote long mail that Tim hasn't seen about why this is hard to fix -... please come back to this discussion once Tim gets a chance to review the note (his email server was down).
Tim - complexity is great, I agree (after seeing Magnus mail), will clear.
Document is approved.
o draft-ietf-tsvwg-ecn-mpls-02.txt
Explicit Congestion Marking in MPLS (Proposed Standard) - 7 of 7
Note: Lars Eggert (lars.eggert@nokia.com) is document shepherd.
Token: Lars Eggert
Was deferred.
2.1.2 Returning Item
NONE
2.2 Individual Submissions
2.2.1 New Item
o draft-saintandre-rfc4622bis-01.txt
Internationalized Resource Identifiers (IRIs) and Uniform Resource Identifiers (URIs) for the Extensible Messaging and Presence Protocol (XMPP) (Proposed Standard) - 1 of 1
Token: Lisa Dusseault
Approved with no discussion
2.2.2 Returning Item
NONE
3. Document Actions
3.1 WG Submissions
3.1.1 New Item
o draft-ietf-tsvwg-diffserv-class-aggr-06.txt
Aggregation of DiffServ Service Classes (Informational) - 1 of 6
Note: Shepherd by WG chair Magnus Westerlund
Token: Magnus Westerlund
DISCUSS - need to discuss today?
Tim - thought security considerations complete for diffserv but didn't address the aggregation issue. Would be nice to say whether this imposes new security issues, can point off to diffserv for previous considerations.
Sam - We are looking for a determination that there are no new security issues. If that is the case, simply saying that there are no new issues plus a reference would be good enough. Right?
Tim - yes, that would be fine.
AD followup on this document.
o draft-ietf-ipfix-implementation-guidelines-07.txt
IPFIX Implementation Guidelines (Informational) - 2 of 6
Note: Nevil Brownlee is the PROTO shepherd
Token: Dan Romascanu
Couple of DISCUSSes - did Dward hit DISCUSS button? yes, but wasn't e-mailed to IESG.
Wasn't sure about Dward's second point about little deployment experience.
Dward - MUST-implement for SCTP but it's flaky and you don't make a recommendation between TCP and UDP - just tell us what to do, give them a fallback.
Dan - Got it.
AD followup, revised ID needed for 2119 comments.
o draft-ietf-v6ops-scanning-implications-03.txt
IPv6 Implications for Network Scanning (Informational) - 3 of 6
Token: Ron Bonica
Number of DISCUSSes.
Ron - lots of things are wrong with this document, revised ID is needed.
Sam - Abstained because the quality of this document isn't high enough to be worth reading. This document specifically has unrealistic stuff - "privacy addresses for main servers" but they need to be reachable. Onlink discussion doesn't have enough detail. Just not a good treatment of the subject. I'm checking out, not trying to block publication.
Revised ID needed.
o draft-ietf-eap-netsel-problem-08.txt
Network Discovery and Selection Problem (Informational) - 4 of 6
Token: Mark Townsley
Was deferred this morning.
o draft-ietf-behave-p2p-state-05.txt
State of Peer-to-Peer(P2P) Communication Across Network Address
Translators(NATs) (Informational) - 5 of 6
Token: Magnus Westerlund
Was also deferred this morning.
o draft-ietf-avt-topologies-07.txt
RTP Topologies (Informational) - 6 of 6
Note: Roni Even is the document shepherd
Token: Cullen Jennings
(Some confusion about web agenda here)
No DISCUSSes - approved for Informational with no discussion.
Russ - please put this in Writeup is Needed so Cullen can look at this before it's announced.
Magnus - comments are fixed
Russ - The Gen-ART Review comment is most important. Did you also look at the SecDir Review?
Magnus - Gen-ART Review issues were handled. Don't know about SecDir Review issues.
Russ - then let it go.
3.1.2 Returning Item
NONE
3.2 Individual Submissions Via AD
3.2.1 New Item
NONE
3.2.2 Returning Item
NONE
3.3 Independent Submissions Via RFC Editor
3.3.1 New Item
o draft-malis-sonet-ces-mpls-09.txt
SONET/SDH Circuit Emulation Service over MPLS (CEM) Encapsulation (Historic) - 1 of 1
Token: Mark Townsley
DISCUSS was cleared on the call, approved with no DISCUSSION, needs an IESG Note.
Point raised, write-up needed.
3.3.2 Returning Item
NONE
4. Working Group Actions
4.1 WG Creation
4.1.1 Proposed for IETF Review
NONE
4.1.2 Proposed for Approval
o Performance Metrics for Other Layers (pmol) - 1 of 1
Token: Dan
Dan - open DISCUSS by Russ.
Russ - e-mail we exchanged didn't line up with charter text.
Dan - last e-mail says not referring to rechartering is by design - limited scope, framework plus one example. Can add text about rechartering or shut down.
Russ - fine with that, will withdraw DISCUSS with that change.
Dward - chairs?
Lisa - pessimistic expectation is that we get a framework standardized and a lot of profiles written by people who write monitoring software and this will become an implementation tax for developers. A lot of APPS is really light-weight (couple of weeks). This work really increases the development load for APPS protocols.
Dan - framework document mandates nothing. Doesn't mean this will be mandatory for every protocol to define performance metrics. Don't see why this is of concern, it's an area where a number of protocols have had problems. If WGs don't think they need profiles, they don't have to do one.
Russ - can you add wording saying what Lisa said?
Lisa - can't justify blocking this work based on possible future harm. Language won't help a lot, but it might help. Won't insist on any language in particular.
Chris - also have concerns about work on a protocol without domain expertise. SMTP - store and forward, performance problems aren't problems. have to understand protocol deeply.
Dan - "work is started at initiative of WG developing a protocol and can be done in that WG - agnostic".
Chris - same concerns as Lisa, not comfortable blocking, but...
Lisa - how to ensure participation of experts?
Dan - this charter, or future work?
Lisa - framework in absence of anything in the framework makes designing a framework that nothing fits in is a problem.
Dan - APPS advisor? or early reviews with APPS area?
Lisa - thinking more about RAI/SIP, more likely to have their performance tested. Would be good to work on couple of specific pieces in this charter.
Dan - has SIP performance draft in charter now.
Lisa - OK.
Dward - BMWG issues recently - could have been covered by early review/cooperation with WGs in charter. Please circulate text, don't want to repeat same issues we've encountered in the past.
Not approved today? Maybe by end of the call - Dan will put text in jabber, can come back.
(resumed discussion later in the call, after text in jabber)
Does text resolve Lisa/Chris concerns?
Chris - acceptable.
Lisa - if everyone agreed to start working on metrics for IMAP, would that happen in PMOL? text seems to say this.
Dan - will strike this text.
No other objections, so Dan will send text - approved on this call.
4.2 WG Rechartering
4.2.1 Under evaluation for IETF Review
o IP Flow Information Export (ipfix) - 1 of 2
Token: Dan
Number of incremental changes to initial charter, don't modify core set of documents, extend them. Configuration model, status, performance modeling. IPFIX file format. Small extension dealing with type for vendor-specific extensions. Stream reporting capability and guidelines. Flow aggregation and normalization.
Magnus - just approving this, or sending it out for IETF review?
Russ - sometimes we approve a recharter without IETF Review, I think he is asking if you want to do so in this case.
Amy - standard question is "objection to just making the changes" - do you want to go for external review?
Magnus - yes. Would also like answer to e-mail question.
Dan - questioning whether item should be part of the charter?
Magnus - mediation between which entities are we talking about? not clear in charter.
Dan - think I know the answer, but can we send charter as query to the list?
Magnus - went out to the list, I think. Fine with sending for IETF review, just need answer before approval.
Will go for external review. Dan will fix a couple of typos after the call.
o Network Configuration (netconf) - 2 of 2
Token: Dan
Lisa - needs to go for external review.
Chris - especially for NETCONF over TLS - compared to BEEP/single BEEP channel?
Dan - last thing we did was put client authentication into charter.
Sam - review at a point where we can still change the approach. Sometimes WGLC is very late in the process. If they're ready, that would be fine. Like the milestones you're proposing.
Dan - send current text to IETF review and conduct discussion about security over next week or so?
Sam - suspect there are several people in NETCONF who think NETCONF over TLS will solve all their problems and don't agree on what that means.
Dan - will check for Russ' nits and provide final version of text.
4.2.2 Proposed for Approval
o Common Control and Measurement Plane (ccamp) - 1 of 1
Token: Ross
Already been to external review. Objection to rechartering?
Magnus - text still needs to be cleaned up so it's readable - formatting.
Russ - yes, still. Content is OK.
Amy will fix the format since the content is correct.
5. IAB News We can use
Loa - we have T-MPLS design team, currently under IAB review, plan to announce on end of business day Friday (probably Saturday).
Have new NGN liaison, sending out information now. (Monique Morrow)
6. Management Issue
6.1 Update URL for application/wspolicy+xml media type (Chris Newman)
W3C media types, registered without final version, this is just changing the URL in the IANA registry to point to final version.
Should be a fast-track approval process for small changes, but that's another discussion for another time.
6.2 Approval of application/sparql-query media type (Chris Newman)
New W3C media type for query mechanism standard. Didn't see any obvious problems with registration, Chris recommends approval.
Objections? none on the call, also approved.
6.3 IANA Expert for IPFIX IEs (Dan Romascanu)
Two experts approved (primary and secondary) - primary - Nevil Brownlee, backup - Juergen Quittek
6.4 Approval of application/sparql-results+xml (Chris Newman)
Chris - Not sure everyone had a chance to look at it yet.
Russ - any concerns since Chris added this one fairly late? I am OK with talking about it on this call
Cullen - OK with contents.
Approved.
6.5 Approval Questionnaire for IAOC Nominations (Russ Housley)
Have a call for nominations out now, have several nominations, last time we did this we sent people a questionnaire, suggest use same questionnaire this time, but want formal approval.
No objections - approved.
6.6 IESG and AD statistics (Jari Arkko)
Added since there's a fair bit of statistics available, been discussing what makes sense (and does not). Cullen pointed out results are highly variable depending on how you do the calculations. Want group feedback on what makes sense and how they should be used. Comparison between ADs is probably problematic, looking at improvement for a specific AD probably makes sense.
Dan - like interpretation you guys used. Trying to compare ADs, especially between areas, is apples and oranges. Find this useful for myself, setting goals, looking at deviations and understanding reasons for deviations. Useful to maintain, not a common denominator that we can use for whole IESG.
Cullen - we don't have anything, and that creates a vaccuum for people to create something that I like less.
Sam - put everything you know out there with caveats and make sure NomCom liaisons can either talk about statistics or point to someone who can. Not "sending to NomCom", but public information will probably be provided to NomCom.
Jon - have concerns about publishing this, because we will change our behavior just because of how things look on metrics - gaming.
Sam - putting stuff in external review when it really is, is not a bad thing.
Cullen - problem is that tracker has no way for external party to update.
Ross - if you are in external review for five minutes, the information becomes obsolete quickly.
Jari - current states don't map well to show you who is really responsible for time in various states.
Cullen - even most well-defined state change is variable - Sam can be best AD or worst AD based on how we do the calculation.
Jari - but it's about what the usage of the stats is. If you use them for improving your performance from what it is currently, you need data. Comparing ADs is the wrong goal
Cullen - taking your number and trying to improve by 10 percent is realistic, other uses may not be. I've already had questions from NomCom about my stats - my answer was "this stuff makes no sense".
Sam - what are people's concerns about misuse?
Jon - IESG slate, get rid of "two worst performers" - why is this not the natural thing to do?
Ross - NomComs may be inconsistent - value more DISCUSSes, value less DISCUSSes
Cullen - also concerned about non-NomCom, wanting documents finished before PUB-REQUESTED, but that makes harder for authors to understand what's going on because most of the work isn't being tracked. I work at a place where you get what you measure, need to be careful what you incent. Agree with Jari about usefulness of this data, I look at my own data and respond to yellow-red transitions.
Lisa - is underlying data public already? then why make available one chosen way of munging this data over another?
Cullen- not saying should or should. If you play with this data for more than 20 minutes you realize how lumpy the data is.
Jari - we all agree on that, not the debate. One debate is "make public", but most of it has been for years with Bill, and I've added more. Is preventing access the solution, or explaining, or stopping producing the data ...
Sam - one reason to publish is to explain the danger of having 20-minute analysis to others. IESG consensus is fine, but individual ADs should be free to show what they are comfortable showing.
Jari - OK with resticting access to IESG, I'm listening to what you say about this.
Ross - NomCom could use this information inappropriately, but they could also ask appropriate questions ("why was this document in one state for 18 months?" is not inappropriate).
Cullen - suspect Bill and Ted had this discussion a long time ago. We make this information public but no one knows about it.
Jari - difference is publicizing versus making available in some obscure web page. My opinion is "public with explanantions".
Lars - agree. Bill has had this online for years, all the same stuff. If lines are straight, that's good, if not, that's bad.
Jon - not sure what latest incarnation is, but reads like a value judgement. I'm an outlyer in all these things, but if you're saying "here's your name, here's your number", you're making a value judgement. I never looked at Bill's reports that way.
Jari - I found Bill's graphs problematic for my own information, to be used for my own improvements as an AD.
Cullen - looked useful, actually useless...
Magnus - don't care about public or not. We know how this works from the inside, anyone else will have problems using it, will assume data actually means something. Do whatever you want, not sure it's even useful to me.
Lars - would people really do something different to make numbers look different? We're all pretty busy, assume this would be true for all ADs.
Jon - I would do things differently to make myself look better - why wouldn't I?
Jari - have compromise proposal. If we make data available to IESG only would that remove concerns?
Jon - concerned this will be used by NomCom to kick people off. IESG only would remove this concern.
Cullen - with Sam's comment - hope our liaison can explain the uses of this data. If they are discussing this data, you should be talking about concerns. Is this within your instructions? If they are analysing data and you can't say anything, that's a problem.
Lars- do we all agree that I can bring this up (without being asked)?
Jari - Would be traditional IESG communication to NomCom. He would be OK to talk about this if the IESG agrees that the data has to be looked at very carefully. And I think we do.
Cullen - also focus on data that's useful for ADs, not data for comparing ADs.
Jon - when I was motivated to do something by data, that's not bad, but this isn't a consensual report card.
Jari - agree this could be improved.
Russ - doesn't reflect all aspects of the job.
Jon - measuring document lifecycles without looking at how many documents are in queue is problematic.
Cullen - think public view will make large areas serialize work more, that's bad. Data is public and has to remain public, guided use is better than unguided use.
Jari - make the stats iesg-only, instruct lars to guide the nomcom if they talk about this, rework metrics - is everyone OK with that?
(silence)
7. Agenda Working Group News
Jari Arkko
Multicase Mobility quiet, RRG list now has question about whether LISP guys can have time during lunch break for LISP tutorial. Eager to give time but need to make sure this isn't a marketing slot.
- Russ - what day?
- RRG is meeting on Monday and Friday
- Russ - WG chair training on Wednesday, approved something on BOF call
- We're going to collide with something, no matter what
- Cullen - BOF call was SIP interop, not a bad collision because there's not much overlap
- do people like this? as long as they pay for lunch, or provide time to get a lunch before they start the program
Looking for chairs for SAVI ...
Larger topic is IPv4 status/IPv6 deployment and RIPE presentations. What should we do? already initiated V6OPS - what else? V4NAT space? planning to send mail.
- Ross - understand why there's concern, but what are we talking about - v6 deployment? v4 bandaids?
- Russ - expect we need to do both
- Jari - need to know what IETF message is
EAP WG, trying to approve last document and close quickly.
SHIM6 AD reviews are complete, documents are coming.
Lisa Dusseault
Chris and I should mention workshop on APP Architecture workshop (Feb 11-12) in Bay Area - IAB declined to host, we're doing this ourselves. A lot of things appear over and over again in our queue - looking at design patterns. Asking people to prepare position papers.
Lars Eggert
Two fewer working groups - ??? and ??? closing
Chris Newman
LEMONADE is having interop event this month in Munich.