On Mar 10, 2008, at 12:52 AM, Jonathan Rosenberg wrote: > I read through the new invfix draft. I think this new format allows > it to be understandable as a standalone document, which is good. > > Just a couple of comments: > > 1. section 7 (which has the standalone description) has normative > language in it. Thus, there is the possibility of conflict with the > actual formal spec, which is done via diff. One of the two needs to > be made non-normative.
Some conversation has happened around this point on list already - could you reply to the thread it was in? > > 2. 7.1 and 7.2 are a bit confusing in that they talk specifically > about UAC and UAS but of course the transaction machines apply to > proxies. Section 7.3 has proxy considerations but is very thin. I > think 7.1 and 7.2 should be restructured around client and server > transactions Hrm - is this just a word-choice change rather than a restructure? > > > I personally suspect the actual formal diffs in section 8 will be > too hard to apply for anyone to use them; everyone will work off of > the descriptive text in section 7. So I don't object to it being > there, but I think we should realize it will be ignored and section > 7 better be complete and accurate. Having watched many implementors approach the specs, particularly updates to specs, I disagree that 8 will be ignored and that interoperability will suffer without it. I do agree that 7 needs to be complete and accurate. Other than the one comment you have above, do you see a place where it isn't? > > Thanks, > Jonathan R. > > Robert Sparks wrote: >> Folks - >> Lets try to get the Essential Corrections and -invfix- discussion >> off the agenda since there are clearly gnarlier things to talk about. >> I sent in a revision to inv-fix attempting to address Jonathan's >> concern about stand-alone readability. >> Jonathan - is the new format better? Can we go with this? >> I sent a separate message showing what it would look like to build >> a diff against 3261 that we could put in the document. You'll >> remember that on looking at it, its my opinion that going down >> that path is a very bad idea. >> So far, there's only been one response on list to those messages. >> Please, everyone (especially those that have expressed strong >> opinions in the past), look those over and let us know if we're >> past the primary concerns, or if we still have something to work >> through that needs face time to do so. >> RjS >> On Feb 29, 2008, at 5:00 AM, DRAGE, Keith ((Keith)) wrote: >>> (As SIP WG cochair) >>> >>> As Dean already said, we spend 3 hours last night trying to plan >>> a 2 1/2 >>> hour session based on input we do not completely have. We don't >>> think we >>> finished, but this was the best we could reach so far in the time >>> available. Things can move up, but they don't solve the fundamental >>> problem that we don't have enough time. It was not our idea to >>> limit SIP >>> to a single session - this was forced on us by the ADs. >>> >>> Modification of this agenda is more likely to occur not by >>> complaining >>> about the time allocated, but by the presence of technical >>> discussion on >>> list over the next 8 days. >>> >>> Some generals: >>> >>> We have taken the principle that if discussion time is needed on >>> chartered items specifically, then they come first, and more >>> specifically highest priority for any discussion needed to get >>> something >>> out of WGLC and into publication request. This is then followed by >>> documents that we envisaged could hold up chartered items (in >>> this and >>> other groups) if not discussed. Then follow the other items where >>> time >>> has been requested. >>> >>> Some specifics based on comments received so far: >>> >>> Domain-certs and eku are in WGLC which ends of Friday next week. >>> These >>> are a dependency from other chartered items where we are ready to >>> submit >>> the publication request. We will not have a view of what >>> specifically >>> needs discussion on these in the meeting until the end of the >>> WGLC, when >>> all the comments are in. Many of the comments made will not need >>> discussion in the SIP session and can be discussed on list by the >>> editors after the end of last call. If any comment does need SIP WG >>> session time, then we will give it on the above basis. >>> >>> Outbound is in WGLC and has no time. We are still waiting for >>> input to >>> know if it needs time - not helped by the authors dropping yet >>> another >>> new version in the inbox 3 minutes before the -xx deadline when it >>> should have been there a month ago. Our current view is to issue a >>> refreshed WGLC on this following the meeting but that can change. >>> >>> Location conveyance is in WGLC and has no time. Again author >>> drops new >>> version in inbox 3 minutes before -xx document deadline. This >>> document >>> is impacted by: >>> >>> http://www.ietf.org/internet-drafts/draft-peterson-geopriv- >>> retransmissio >>> n-00.txt >>> >>> Which will be discussed in the GEOPRIV group, but if followed >>> reverses >>> the decision we made in the location-conveyance breakout at the last >>> meeting. We hope the GEOPRIV discussion does not force >>> rediscussion in >>> SIP, but please bear in mind that it could do so. >>> >>> RPH in responses is not chartered. 2 meetings back the WG >>> expressed a >>> wish to work on this. The AD pushed back to our request to >>> charter. As >>> the AD makes the decision, it is not chartered and has no >>> milestones. >>> That's why we had the further discussion in the last meeting - where >>> again we came to no conclusions, and I assume the AD is still >>> unconvinced. Fundamental problem - convince the AD. Put him in a >>> corner >>> somewhere in Philadelphia and kick hell out of him, but we don't >>> need >>> the SIP session to watch you do that - although it may be >>> entertaining! >>> >>> Media security requirements is holding up other chartered items >>> in both >>> SIP and AVT. When we considered this for inclusion, we believed >>> there >>> were two items that may need agenda item. Overnight Dean seems to >>> have >>> taken the time away at the moment, but I remain unconvinced that >>> this >>> does not need time. >>> >>> This is the media-security requirements milestones: >>> >>> Sep 2007 Requirements for media keying to WGLC (Informational) >>> Nov 2007 Requirements for media keying to IESG (Informational) >>> >>> Which provides the requirements for the mechanism in these >>> charter items >>> in SIP. >>> >>> Dec 2007 Establishment of secure media sessions using DTLS- >>> SRTP to >>> WGLC (PS) >>> Feb 2008 Establishment of secure media sessions using DTLS- >>> SRTP to >>> IESG (PS) >>> >>> And these chartered items in AVT: >>> >>> Dec 2007 Submit in band keying mechanism for SRTP draft for >>> Proposed >>> Standard >>> >>> >>> Regards >>> >>> Keith >>> >>>> -----Original Message----- >>>> From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On >>>> Behalf Of Cullen Jennings >>>> Sent: Friday, February 29, 2008 5:33 AM >>>> To: Dean Willis >>>> Cc: IETF SIP List >>>> Subject: Re: [Sip] Revised agenda for SIP -- needs more work yet >>>> >>>> >>>> On Feb 28, 2008, at 8:57 PM, Dean Willis wrote: >>>> >>>>> >>>>> On Feb 28, 2008, at 5:00 PM, Jonathan Rosenberg wrote: >>>>> >>>>>> 10 minutes on the phone number dichotomy thing isn't going to be >>>>>> even close enough to cut it. It should be either zero or a more >>>>>> significant number. >>>>>> >>>>>> We have 30 minutes on media security. The only chartered >>>> item there >>>>>> is the requiremetns document which I thought was mostly >>>> done. Why is >>>>>> there so much time dedicated to this? I would rather move that to >>>>>> INFO or the identity mess. >>>>> >>>>> Me too. But hey, the media security requirements document is >>>>> CHARTERED. We have permission to work on it and a commitment to a >>>>> deliverable, with a published milestone. We don't have that >>>> for INFO >>>>> or the identity mess (even though I think media security is >>>> blocked on >>>>> the identity mess -- our AD currently disagrees). >>>>> >>>> >>>> Dean - I think we must be having some serious >>>> miscommunications - many times in the past I have had enough >>>> typos in my emails that no one could understand what I might >>>> have been thinking but .... >>>> >>>> I don't think I said that. The two things I thought I said >>>> that might have been confused are: >>>> >>>> 1) I don't think the E.164 discussion is holding up milestone >>>> items in MEDIACTRL or SIMPLE >>>> >>>> 2) I don't think a problem in some mechanism such as 4474 >>>> should be holding up the media security requirements draft. >>>> (If I could bold requirements in the previous sentences I would) >>>> >>>> I certainly do think it is important that we have a clear >>>> understanding of how E.164 numbers fit into the overall SIP >>>> security picture. Sorry if I said something that made you >>>> think otherwise. >>>> >>>> >>>> _______________________________________________ >>>> Sip mailing list https://www.ietf.org/mailman/listinfo/sip >>>> This list is for NEW development of the core SIP Protocol Use >>>> [EMAIL PROTECTED] for questions on current sip >>>> Use [EMAIL PROTECTED] for new developments on the application of sip >>>> >>> _______________________________________________ >>> Sip mailing list https://www.ietf.org/mailman/listinfo/sip >>> This list is for NEW development of the core SIP Protocol >>> Use [EMAIL PROTECTED] for questions on current sip >>> Use [EMAIL PROTECTED] for new developments on the application of sip > > -- > Jonathan D. Rosenberg, Ph.D. 499 Thornall St. > Cisco Fellow Edison, NJ 08837 > Cisco, Voice Technology Group > [EMAIL PROTECTED] > http://www.jdrosen.net PHONE: (408) 902-3084 > http://www.cisco.com _______________________________________________ Sip mailing list https://www.ietf.org/mailman/listinfo/sip This list is for NEW development of the core SIP Protocol Use [EMAIL PROTECTED] for questions on current sip Use [EMAIL PROTECTED] for new developments on the application of sip
