IETF
Arcing
arcing@jabber.ietf.org
Tuesday, April 5, 2016< ^ >
Room Configuration
Room Occupants

GMT+0
[12:49:08] bortzmeyer joins the room
[12:50:52] DanYork joins the room
[12:52:31] hildjj joins the room
[12:53:15] Meetecho joins the room
[12:53:25] Barry Leiba joins the room
[12:55:58] KY Hong joins the room
[12:56:20] Andrew McConachie joins the room
[12:56:42] Ted.h joins the room
[12:56:44] Antoin Verschuren joins the room
[12:57:01] philippe Fouquart joins the room
[12:57:13] Brian Haberman joins the room
[12:58:25] Wolfgang Beck joins the room
[12:58:32] Yoshiro Yoneya joins the room
[12:58:48] Yoshiro Yoneya leaves the room
[12:58:56] Yoshiro Yoneya joins the room
[12:59:04] Lyman Chapin joins the room
[12:59:35] Peter Koch joins the room
[12:59:42] ken carlberg joins the room
[13:01:06] Alexander Mayrhofer joins the room
[13:02:05] Michael Richardson joins the room
[13:02:35] mcr2 joins the room
[13:02:48] <mcr2> *TWO* IABs people at BOF chairs. Oh My.
[13:03:08] mcr2 is now known as mcr
[13:03:28] DanYork has set the subject to: ARCING BOF at IETF95
[13:03:32] <hildjj> Yes.  Oh my indeed.
[13:04:29] <hildjj> however, this is an *ietf* BoF, not an IAB meeting.  Brian is the overseeing AD.
[13:04:31] Roger Carney joins the room
[13:04:45] Suzanne (co-chair) joins the room
[13:05:08] Dave Crocker joins the room
[13:05:13] m&m joins the room
[13:05:25] <Brian Haberman> @hildjj : And I packed my whip in case we get out of hand. ;)
[13:05:32] rlb joins the room
[13:05:34] Andrew Sullivan joins the room
[13:05:38] John Levine joins the room
[13:05:58] Ted.h leaves the room
[13:06:07] Ted.h joins the room
[13:06:09] <Andrew Sullivan> Suzanne reports that she's seeing illusory faces and opening the meeting
[13:06:09] Brian Trammell joins the room
[13:06:10] <Ted.h> http://www.azlyrics.com/lyrics/temptations/smilingfacessometimes.html
[13:06:28] naptee joins the room
[13:06:32] <DanYork> Opening up the discussions
[13:06:43] Eliot Lear joins the room
[13:06:50] <Brian Haberman> Dan thought the stars were a joke.
[13:06:50] Matthew Pounsett joins the room
[13:06:52] shollenbeck joins the room
[13:06:54] <bortzmeyer> Pronounciation: "arking" or "arssing" ?
[13:07:02] <Matthew Pounsett> arking.
[13:07:08] <Barry Leiba> Pity.
[13:07:11] <DanYork> https://www.ietf.org/proceedings/95/slides/slides-95-arcing-3.pdf
[13:07:16] <Matthew Pounsett> :)
[13:07:16] <Ted.h> As an "we're hoping to get grounded"
[13:07:30] <DanYork> Slide 4 - Agenda
[13:07:31] tale joins the room
[13:07:55] <DanYork> IF YOU ARE REMOTE AND WANT SOMETHING BROUGHT TO THE MIC, please put "mic:" at the beginning of you comment.
[13:08:10] <DanYork> Otherwise I'll assume everything else here is backchannel conversations
[13:08:12] <tale> Hahaha @ arssing
[13:08:49] <DanYork> Slide:  Intro: Internet Names
[13:08:55] <DanYork> Slide: How is this a problem?
[13:09:07] <Wolfgang Beck> thanks Dan!
[13:09:20] tim wicinski joins the room
[13:09:23] philippe Fouquart leaves the room
[13:10:40] Tomohiro Fujisaki joins the room
[13:11:21] <DanYork> Slide: Boundaries
[13:12:02] <DanYork> "We are not here to fix DNS at any layer" - you have been warned! ;-)
[13:12:24] Steve Olshansky joins the room
[13:12:37] <DanYork> Slide: Possible discussion questions
[13:13:18] <DanYork> Ralph Droms at mic
[13:13:33] philippe fouquart joins the room
[13:14:41] Olafur joins the room
[13:14:51] <DanYork> Ed Lewis presenting https://www.ietf.org/proceedings/95/slides/slides-95-arcing-1.pdf
[13:15:17] <DanYork> Draft is https://datatracker.ietf.org/doc/draft-lewis-domain-names/
[13:15:27] RjS joins the room
[13:16:27] <mcr> the mic seems to be picking up whistling... maybe turn down the gain?
[13:16:53] <hildjj> better?
[13:16:54] <mcr> still whistling.
[13:17:07] <mcr> a bit better... I think it's echo from the room on higher freq.
[13:17:24] <hildjj> it's fixed in the room, sorry.  maybe meetecho can look at it, if that's what you're using.
[13:17:26] <Suzanne (co-chair)> these rooms have that problem
[13:17:40] <mcr> it's better now.
[13:17:45] <hildjj> cool.
[13:17:51] <mcr> maybe Ed it just backing off more...
[13:17:53] <hildjj> it may be that you just have 30s lag.
[13:18:43] <Suzanne (co-chair)> meetecho there might be a problem to sort with the sound, could you check?
[13:18:50] <Meetecho> will do, just a sec
[13:18:59] <Meetecho> is it a remote audio issue, or local?
[13:19:01] John Levine leaves the room
[13:19:07] <Ted.h> remote
[13:19:19] <Ted.h> (Source local, issue perceived remotely)
[13:19:30] <mcr> remote people hear really loud "s" sounds.
[13:19:40] <mcr> (there is word for that kind of sound....)
[13:19:48] <hildjj> sibilant
[13:19:49] <Ted.h> Sibiliant
[13:20:04] <Ted.h> Joe's spelling is better than mine
[13:20:37] <Matthew Pounsett> I can hear it in the room as well.  I think turning down the mic a bit, or lowering the high frequencies in the EQ would help.
[13:20:37] <mcr> maybe Ed speaks Parsetongue?  (maybe I hear it...)  anyway, it's tolerable.
[13:20:44] Paolo Saviano joins the room
[13:20:47] John Levine joins the room
[13:20:50] <mcr> what Matthew said: the eq for the room is wrong.
[13:21:15] nllz joins the room
[13:21:21] <hildjj> he's not going to talk that much longer.  let's just listen the best we can.
[13:21:32] <mcr> sure :-)
[13:21:58] <m&m> mcr: FWIW, I can hear it too (sensitive to it due to hearing aids)
[13:22:00] <Meetecho> sending someone there to the mixer
[13:22:02] <m&m> (in the room)
[13:23:19] <nllz> No jabber scribe? Or active note taking on the pad?
[13:23:46] <hildjj> At the moment, Ed is presenting his slides.
[13:24:06] <Suzanne (co-chair)> Dan York is jabber scribe
[13:24:22] Lorenzo Miniero joins the room
[13:24:33] <Suzanne (co-chair)> (silde 10 in Ed's deck)
[13:24:41] <nllz> tx :)
[13:25:49] wu teng joins the room
[13:26:13] <Meetecho> playing with the mixer, can you give feedback on the audio?
[13:26:19] <DanYork> Slide:  Step 3
[13:26:19] <Meetecho> so that we know when it's ok
[13:26:26] <DanYork> Slide: Permission-less Innovation
[13:26:52] <Matthew Pounsett> That sounds a lot better.
[13:26:52] Steve Olshansky leaves the room
[13:26:58] <Matthew Pounsett> in the room, anyway.
[13:27:12] <DanYork> nllz: Sorry... I had something urgent that pulled me away from Jabber for a moment
[13:27:15] <Meetecho> and for remotes?
[13:27:21] <DanYork> Slide: Rubber meets the Road
[13:27:42] <DanYork> /help
[13:27:49] <Dave Crocker> remote audio sounds ok to me.
[13:28:09] <Andrew Sullivan> "Time to clarify" slide
[13:28:14] <DanYork> Slide: Time to Clarify?
[13:28:28] <Meetecho> (y)
[13:28:55] <DanYork> Slide: Definition of Domain names
[13:29:46] <DanYork> Slide: Knowing the resolution system?
[13:29:51] Lorenzo Miniero leaves the room
[13:30:04] <DanYork> Slide: Why this is architectual
[13:30:16] <DanYork> George Michaelson at mic
[13:30:54] <DanYork> George has a problem with the definitions of all these things as domain names
[13:31:16] Dave Thaler joins the room
[13:32:14] Petr Špa?ek joins the room
[13:33:01] <DanYork> John Levine at mic
[13:33:30] <Dave Thaler> (previous speaker was Jon Crowcroft at mic)
[13:34:17] <Dave Crocker> Meetecho:  Hmmm.  Speakers are clear enough at higher volumes, but I keep missing what they are saying at lower volumes.  Distortion.
[13:34:43] mellon joins the room
[13:34:50] <Meetecho> we'll check that
[13:34:54] <DanYork> Ted Hardie presenting https://www.ietf.org/proceedings/95/slides/slides-95-arcing-2.pdf
[13:34:57] <mcr> 1080p was 4-pounds...?
[13:35:04] <mcr> oh... I get it now.
[13:35:50] <DanYork> Slide: "Identifiers occur in contexts"
[13:36:02] <DanYork> Slide: "The Internet is a context"
[13:36:10] <DanYork> Ahhh... he reuses titles...
[13:36:32] <DanYork> okay... numbers... oh, wait... slides aren't numbered....   well... we'll do some combo.
[13:36:48] <DanYork> Slide: "Internet names are part of directed graphs"
[13:37:16] <Dave Thaler> reuse of titles means you have to resolve them in different slide contexts :)
[13:37:24] <hildjj> heh
[13:38:04] philippe fouquart leaves the room
[13:38:17] <DanYork> HA!
[13:38:36] philippe fouquart joins the room
[13:38:36] <DanYork> "Ted Hardie is a context"
[13:38:36] <Dave Crocker> If www.nasa.gov is unique, so is gov.
[13:38:36] <DanYork> Slide: Not all names are with the global DNS"
[13:38:36] Ted.h leaves the room
[13:38:54] <Matthew Pounsett> Dave: "gov." vs "gov"
[13:39:00] <Barry Leiba> One expects there to be many Andrew Sullivans.  But one might not really expect it for Joe Hildebrand, and yet…
[13:39:07] <DanYork> Slide: Not all names are with the global DNS - second slide
[13:39:17] <Suzanne (co-chair)> Ted Hardie and Ed Lewis are actually both named Edward, but the disambiguation is so good we never notice :)
[13:39:23] resnick joins the room
[13:39:29] <DanYork> Slide: Remember nsswitch.conf
[13:39:41] <Dave Crocker> Matthew, maybe.  But for the way that slide was done, not really.  www.nasa.gov isn't scoped, so gov doesn't need to be either.
[13:40:12] <hildjj> the slide had "gov" and "www.nasa.gov <http://www.nasa.gov>." (carefully)
[13:40:14] <DanYork> yes, I remember nsswitch.conf
[13:40:26] <Dave Crocker> So far, the presentations seem to be trying to distinguish name registration from name lookup.
[13:40:29] <Brian Haberman> @Dan : I feel for you.
[13:40:42] <rlb> can we all agree that the final dot is a useless distraction?  kthx
[13:40:44] <mcr> reports of the death of nsswitch.conf is exagerated. It's still around. It's just that few have an NIS...
[13:40:50] Eliot Lear leaves the room
[13:40:52] <hildjj> rlb: not in this context
[13:40:52] Eliot Lear joins the room
[13:40:59] <hildjj> mcr: not on OSX
[13:41:13] <Dave Crocker> rlb, depends on whether it's a dot or a question mark.
[13:41:23] <Dave Crocker> oh.  you mean for a domain name...
[13:41:35] <mcr> hildijj, of course, it's not on OSX.  But I'll bet ubuntu on win10 has it too :-)
[13:41:40] <DanYork> Slide: "An NSSWITCH indicator for the Internet"
[13:42:07] <rlb> am i the only one who's never heard of NSSWITCH?
[13:42:12] <hildjj> yes
[13:42:15] <rlb> k
[13:42:24] <Eliot Lear> but this was necessary.  Sun and friends needed a means to transition people from host files and YP to the DNS.
[13:42:24] <DanYork> "how do we make it so that different resolution contexts can coexist on the Internet without collision?"
[13:42:29] <Matthew Pounsett> rlb: I'm pretty sure you are.
[13:42:42] <DanYork> Slide: But how?
[13:42:43] rlb goes back to the security ghetto
[13:42:55] <mcr> http://man7.org/linux/man-pages/man5/nsswitch.conf.5.html
[13:43:05] <mcr> first invented by Sun around 1990 in SunOS 3.
[13:43:16] <mcr> or at least, that's when *I* first used it...
[13:43:20] <rlb> https://en.wikipedia.org/wiki/Name_Service_Switch
[13:43:31] <rlb> so, historical name resolution cruft :)
[13:43:45] <Brian Haberman> @rlb: yes.
[13:43:47] <Eliot Lear> no it wasn't in 3, nor was it in 4.0.  we had to hack libc back then in very ugly ways.
[13:43:59] <Suzanne (co-chair)> @rlb please don't be scared away though. Part of the challenge here is the silos among apps people, DNS people, security people, UI people, ….
[13:44:02] <DanYork> rlb: https://en.wikipedia.org/wiki/Name_Service_Switch#nsswitch.conf
[13:44:33] <DanYork> Yes, some of us spent too many years in Solaris and other early Unix flavors...
[13:44:52] <DanYork> And yes... "historical name resolution cruft"
[13:44:56] <Eliot Lear> yeah
[13:45:04] <Olafur> old draft on this problem https://tools.ietf.org/html/draft-ogud-appsawg-multiple-namespaces-00
[13:45:08] <resnick> You got Solaris? Some of us were stuck on AIX on IBM RTs.
[13:45:12] rlb is not actually in the room, just trolling on xmpp
[13:45:19] <DanYork> Paul Hoffman at mic
[13:45:52] Steve Olshansky joins the room
[13:45:53] <Brian Haberman> @resnick: Wow… I used to develop routing code on AIX…
[13:46:02] <Eliot Lear> AIX: it will remind you of UNIX
[13:46:24] <Matthew Pounsett> I'd like to point out that Centos 7 still comes with nsswitch enabled.
[13:46:45] <mcr> so, Paul is saying that obiwan.sandelman.ca can be assigned in a DAG, but I could look it up linearly in /etc/hosts.
[13:47:31] <DanYork> Eliot Lear at mic
[13:47:47] <mcr> for /etc/passwd, pam has replaced nsswitch, but I think that there is a pam module to read /etc/nsswitch now.
[13:48:00] <John Levine> I have an nsswitch right here on my laptop
[13:48:55] Steve Olshansky leaves the room
[13:49:14] <DanYork> Brian Trammel up to present https://www.ietf.org/proceedings/95/slides/slides-95-arcing-0.pdf
[13:49:34] Ted.h joins the room
[13:49:41] <DanYork> Slide: Why am I here?
[13:49:43] <mcr> I think it's scarrier than what Ted said:  "if you resolve name FOO outside of context BAR, then you may get spoofed"
[13:50:03] <hildjj> yes, and i may leak info to the wrong context.
[13:50:32] <Andrew Sullivan> haha, he said root
[13:50:41] <Suzanne (co-chair)> I think Brian just said he personally is out of context as a non-DNS geek
[13:50:42] <DanYork> Draft here: https://datatracker.ietf.org/doc/draft-trammell-inip-pins/
[13:51:06] mcr wonders if buzzword bingo sheets were handed out.
[13:51:17] <Andrew Sullivan> alas, no
[13:51:35] <hildjj> no fair if you use the person's draft to generate the sheets.
[13:52:05] <mcr> isn't that point: you take all the drafts and put them through the grinder.  That way people who play have to pay attention.
[13:52:22] <Matthew Pounsett> But extra funny if you engineer it so that the entire room shouts "bingo" at a key moment in the talk.
[13:52:22] <DanYork> Slide: Insights
[13:53:03] <Eliot Lear> the day after the sun explodes
[13:53:20] <tale> for liberal defintions of "turn off"
[13:53:51] <DanYork> I suspect everyone in the room today will be retired by that time.
[13:54:10] <hildjj> this is another place in the discussion where it may be easier to deploy a new naming system.  it would be nice to know when you're using the newer one.
[13:54:15] <DanYork> I have heard the word "context" mentioned more in this session than I have in a long time.
[13:54:20] <Brian Haberman> @Dan : are you including young'uns like RLB who doesn't know nsswitch? ;)
[13:54:24] <tale> context matters, Dan
[13:54:45] <Andrew Sullivan> I now believe that one piece of context that was _supposed_ to be in the DNS was "CLASS".  
[13:54:54] <Andrew Sullivan> Didn't happen
[13:54:56] <Eliot Lear> Hey Stephane, didn't you just talk at PROTOSEC about the API that Brian needs?
[13:55:13] <tale> sure it did, andrew.  version.bind!
[13:55:23] <DanYork> Andrew Sullivan - yes, that did seem to be the plan
[13:55:32] <DanYork> Slide: Application to ARCING
[13:55:35] <tale> the tragedy of the success of IN
[13:55:48] <DanYork> Andrew - are you going to get up and argue we should bring back CLASS? ;-)
[13:55:59] <Andrew Sullivan> @tale, but A and AAAA
[13:56:00] <hildjj> DNS CLASS only helps in the DNS context, for creating sub-contexts.
[13:56:01] <Ted.h> @Andrew is here to bury class, not praise it.
[13:56:03] <Alexander Mayrhofer> Even though the INternet seems to be pretty CHOAtic
[13:56:08] <Matthew Pounsett> Class is great if you can present that to the user somehow.
[13:56:11] <Andrew Sullivan> which probably should have been a class difference
[13:56:19] <hildjj> Matthew Pounsett: yes
[13:56:29] <DanYork> Ted.h: Yes, I'm aware of Andrew's draft to kill class.
[13:56:32] <mcr> LONG LIVE HESIOD!
[13:56:46] <tale> I think it's worth exploring, even if Andrew's tool of exploration would be an axe
[13:56:47] <Suzanne (co-chair)> I think he actually argues that DNS has no class and we just need to admit it.
[13:57:01] <DanYork> we are aiming for a class-less society
[13:57:17] <mcr> Long live the Marxist DNS revolution!
[13:57:18] <Ted.h> There can be only N.
[13:57:22] <Matthew Pounsett> I wonder if there are any implementations of Hesiod still alive.
[13:57:24] <Suzanne (co-chair)> "Go home DNS, you're drunk and you have no CLASS."
[13:57:27] <rlb> Suzanne: what's a class?
[13:57:34] <rlb> (j/k)
[13:57:37] <Andrew Sullivan> Sometimes, when you're exploring a thicket, an axe is needed.  But more generally, I was trying to point out that explicit context is hard
[13:57:42] <Andrew Sullivan> self-awareness is tricky
[13:57:50] <mcr> rlb: "IN" is a class in DNS. We had other classes: CHAOS, HESIOD, DECNET (I think?)
[13:58:01] <tale> Can't be that tricky.  You can win Go without it.
[13:58:14] <rlb> There is no Class but IN
[13:58:26] <DanYork> Chairs are back starting discussion
[13:58:37] <Ted.h> Yeah, but those attempts didn't have Brian
[13:58:40] <DanYork> Back in chair's slides:  https://www.ietf.org/proceedings/95/slides/slides-95-arcing-3.pdf
[13:58:59] <DanYork> Slide 9 - Possible discussions questoins
[13:59:01] <mcr> mic: is the goal here to head off/redirect the next .onion discussion?   I don't mean this in a negative way, but rather in a permissionless innovation positive way.
[13:59:03] Jim Galvin joins the room
[13:59:09] <DanYork> Dave Thaler at mic
[13:59:20] nllz joins the room
[13:59:27] <DanYork> mcr - will relay
[13:59:27] nllz leaves the room
[13:59:42] <mcr> I guess I could also try the put my hand up.
[14:00:10] <hildjj> i can't see the screen with the meetecho queue
[14:00:23] <hildjj> you're in queue after stewart, either way.
[14:00:29] <mcr> oh well. Dan can ask my question.
[14:00:50] <Meetecho> hildjj: there should be a second screen in the room displaying the queue
[14:01:03] Steve Olshansky joins the room
[14:01:05] <Meetecho> if not something's wrong, let us know and we'll check
[14:01:06] <hildjj> oh, i see it now.  i was looking across the room.  thx.
[14:01:07] <mcr> chairs can't see it...
[14:01:12] <DanYork> @meetecho - it's there
[14:01:27] <Meetecho> chairs have a small dedicated laptop
[14:01:28] <DanYork> And now we see @mcr standing in the queue
[14:01:34] <Meetecho> the one with the "big red button"
[14:01:38] <hildjj> yup. everything's fine now.
[14:01:40] <Meetecho> (y)
[14:01:41] <Suzanne (co-chair)> sorry @mcr, laptop is behind stuff
[14:01:56] <DanYork> mcr: do you want to come in via audio?  Or do you want me to ask your question here?
[14:02:15] <mcr> let's try the tech! WOOHOO.  
[14:03:19] <DanYork> Ted Hardie at mic
[14:03:20] Ted.h leaves the room
[14:05:14] <DanYork> Ted speaking about the different between resolution and naming
[14:05:17] <hildjj> although i'll note, there are people that believe that URNs *are* resolvable in general.
[14:05:19] Brian Trammell leaves the room
[14:05:21] <mcr> I think that URNs are underused due the lack of universal resolution for them.
[14:05:35] <hildjj> ask Keith.
[14:05:38] <DanYork> Stephane Bortzmeyer at mic
[14:05:53] <rlb> mcr: isn't the point of URNs that they don't have resolution?
[14:05:57] <rlb> otherwise they'd be a URI
[14:06:03] <rlb> er, URL
[14:06:09] Ted.h joins the room
[14:06:50] <mcr> URLs are kinds of "absolute", in that they use the DNS, while URNs may map to many different places.  Imagine a URN for (nearest) mens-washroom.
[14:06:52] tim wicinski leaves the room
[14:07:28] <DanYork> Stuart Cheshire at mic
[14:07:36] naptee leaves the room
[14:07:41] <DanYork> mcr: Sounds like a plan for you to come in via audio.
[14:07:53] <DanYork> mcr: I can stand by to relay your question if the tech doesn't work
[14:08:11] <DanYork> @Meetecho: we have @mcr up next in the speaking queue
[14:08:17] <Ted.h> @rlb you might want to look at https://www.isoc.org/inet99/proceedings/4m/4m_2.htm
[14:08:30] <Meetecho> (y)
[14:08:57] <DanYork> @Meetecho - worked nicely!
[14:09:14] <Meetecho> cool! big red button rulez :)
[14:09:39] <DanYork> Eliot Lear at mic
[14:09:46] shollenbeck leaves the room
[14:09:54] <mcr> I think video wasn't enabled at my end. or was it?
[14:10:02] <hildjj> it was.  we saw you.
[14:10:40] Olafur leaves the room
[14:11:01] <mcr> if our URL Location bar was more sophisticated, would we even be having this discussion?
[14:11:48] <DanYork> Andrew Sullivan at mic
[14:11:49] <Ted.h> @mcr they are in many ways too sophisticated, as they incorporate heuristic search.  
[14:12:12] Steve Olshansky leaves the room
[14:12:24] <mcr> Ted.h: I'm not denying that, but in the days before it had that heuristic search, we might have gone in a different direction.
[14:12:27] <Dave Crocker> hildjj: second screen showing queue:  That was tried once and the result was quite a bit of distraction in the room, with a couple of guys turning it into a game.  My conclusion was that the queue needs to be shown to remote attendees and the chair but not the room.
[14:13:05] <mcr> game?  what game?
[14:13:13] <DanYork> Andrew: "The universe has people creating naming systems faster than we can adapt to that."
[14:13:16] <hildjj> Dave: the barrier to entry is high enough that hasn't happened.  also, this isn't the plenary.
[14:14:16] <Dave Crocker> hildjj: ack. this suggests that we need a range of experiments that look for useful modes/mixes of queue display.
[14:14:26] <DanYork> Ted Lemon at mic
[14:14:43] <hildjj> of note: you don't get to put random text on the display, just your name.
[14:15:47] Ted.h leaves the room
[14:16:03] Ted.h joins the room
[14:16:29] <Ted.h> So, to boil that down:  a context-sensitive solution should not be limited to URI contexts?
[14:16:34] <hildjj> yes
[14:17:00] <Dave Thaler> yes since the problem is not limited to URIs
[14:18:09] naptee joins the room
[14:18:21] <DanYork> Lucy Lynch at mic
[14:18:26] Steve Olshansky joins the room
[14:18:27] mcr proposes worldwide /etc/hosts. Disk space is cheap.
[14:18:38] <Meetecho> Dave Crocker in virtual queue as well
[14:18:44] Timothy Morizot joins the room
[14:19:06] <hildjj> dave, were you in queue before i said "people after ralph"?
[14:19:14] <DanYork> John Levine at mic
[14:19:53] <Dave Crocker> hildjj, before, and my comment is probably more for the general discussion now than the targeted 'solution' dicussion you want to get going (though it pertains to it.)
[14:20:10] <Dave Crocker> hildjj:  and btw, thanks for asking.
[14:20:24] <hildjj> ack.  you're after ralph, but allowed to not talk about solutions (although i'd prefer it!)
[14:20:46] <Matthew Pounsett> I have to disagree with John.   Sometimes it's important to disambiguate which pizza place you're talking about.
[14:21:33] <bortzmeyer> mcr: would create interesting privacy issues (no more need for zone file access processes)
[14:21:38] Petr Špa?ek leaves the room
[14:22:12] Ted.h leaves the room
[14:22:16] <DanYork> "bring back the bang-path" :-)
[14:22:38] <Suzanne (co-chair)> @dan: "save the bangs"?
[14:22:44] <mcr> I like SPKI's naming. IETF's John Levine's Joes Pizza's Customer:Joe Hildebrand.
[14:22:57] <Dave Thaler> http://www.example.com.globaldns.contexts.arpa/
[14:23:17] Alexander Mayrhofer leaves the room
[14:23:23] <DanYork> Brian Trammell at mic
[14:23:26] <bortzmeyer> mcr:  GNUnet allows that. (And of course UUCP did a long time ago)
[14:23:30] <mcr> (... I had a test case change because example.com's AAAA record changed...)
[14:23:49] <Peter Koch> people are lazy, so browsers will start adding 'globaldns.contexts.arpa' by default ... (or any other context indicator)
[14:23:52] <John Levine> @matthew yes, sometimes it's important but in natural language we have ways to do it
[14:24:13] <mcr> What Brian said. It's important. It's what SPKI said about naming when it discarded DNs.
[14:24:24] <mcr> (20 years ago)
[14:24:34] <rlb> in tls just now "have you tried to put something in the DNS that's not already there??"
[14:25:26] <DanYork> Ralph Droms at mic
[14:25:42] Steve Olshansky leaves the room
[14:25:42] <hildjj> we call him "Ted"
[14:25:52] Paolo Saviano leaves the room
[14:26:10] Brian Trammell joins the room
[14:26:15] <Barry Leiba> "Bruce"
[14:26:23] <hildjj> Thanks, Bruce.
[14:26:37] <mcr> Ralph's wife probably calls him "Honey", except when sailing, then it's Captain.  And his children have other names for him.  All of those are examples of local-context names.
[14:27:32] <bortzmeyer> Six persons on line to propose actual solutions :-)
[14:27:53] <hildjj> ok, relay?
[14:28:00] <Meetecho> Dave we can't hear you in Meetecho either
[14:28:00] <Dave Crocker> ack
[14:28:01] <DanYork> dave - do you need a relay?
[14:28:07] <Dave Crocker> i'll type.
[14:28:30] <DanYork> Ted Hardie at mic
[14:28:57] <hildjj> MT gets a shout-out from Ted.
[14:29:21] nllz joins the room
[14:29:40] nllz leaves the room
[14:30:04] <hildjj> so, .alt is a proposed solution.
[14:30:47] <Andrew Sullivan> Every meeting I've been in this week, the _second_ remote person couldn't be heard in the room.  Is this a bug?
[14:30:53] <tale> worked brilliantly for usenet.
[14:31:08] <mcr> clearly, we need to reserve bork.bork.chef.swedish.alt as our example.com.
[14:31:14] <Andrew Sullivan> I think alt is a proposed solution of one type.  I think there's probably more than one
[14:31:17] <DanYork> PHB at mic
[14:31:32] Ted.h joins the room
[14:31:42] <DanYork> PHB wants to separate space into two spaces: machine names and human names
[14:32:28] <mcr> PHB is describing public key anchored names, which is  https://tools.ietf.org/html/rfc2693#section-2.6
[14:32:51] <Dave Crocker> The problem with looking for solutions at this point is that we don't yet have a community that is clear and comfortable with the issues being attacked.  There are probably several.  Some concerns are merely distinguishing name registration vs. name lookup and wanting different lookup.  Others want to augment domain names but still use domain name registrations as a base.  Others want a completely independent name space.  The question for these is why?  What is the problem with their using domain names?  And this last leads question leads to the concern that sometimes merely adding a meta-naming context specifier suffers on usability:  people simply won't adopt the convention. We need a more robust community sense of this problem/discussion/solution space before trying to pursue a solution.
[14:32:59] <Dave Crocker> hildjj: see above.
[14:33:19] <DanYork> hildjj: I'd like to get in the queue ... I guess behind Ed Lewis at this point
[14:33:48] <DanYork> (but as I'm scribing I would like to remain seated until it's my point)
[14:33:56] Steve Olshansky joins the room
[14:34:11] <hildjj> dan: ack
[14:34:47] rlb leaves the room
[14:34:56] <DanYork> Dave Thaler at mic
[14:35:04] <hildjj> dave: i'll get you in after thaler.
[14:35:43] <mcr> that's a good context @Dave Thaler.
[14:35:49] Ralph joins the room
[14:35:52] <Andrew Sullivan> The hole between Ed and me is Dan
[14:35:55] <mcr> we need existing names that can go into old placeholders....
[14:35:59] <Dave Crocker> hildjj: tnx.
[14:37:43] <Ralph> Dave - I agree 100%.  
[14:37:53] Timothy Morizot leaves the room
[14:37:58] Antoin Verschuren leaves the room
[14:37:59] <DanYork> Stuart Cheshire at mic
[14:38:05] <Dave Crocker> hildj: very nice reading job.  tnx! (and sorry for the typo; yes you read what i meant.)
[14:38:20] <hildjj> :)
[14:39:02] <Dave Crocker> meetecho:  anyone wanting to 'participate' in a session needs to have access to a sandbox for at least testing audio -- making sure their microphone works for meetecho.
[14:39:23] <mcr> and then we'd AS112 alt. ?
[14:39:26] <Meetecho> that's what the self test page is for :)
[14:39:28] <Meetecho> https://self-test.conf.meetecho.com/self-test/
[14:39:46] <Dave Crocker> ralph, et al: how to have the needed discussion?  I fear this is an irtf topic, for now.
[14:40:05] <Meetecho> we documented some instructions here, if you think they're lacking we'll try and improve them: https://ietf95.conf.meetecho.com/index.php/Remote_Participation
[14:40:13] <Ralph> Dave … this solution vs. problem understanding is, of course, a classic IETF problem.
At the same time, getting to a community understanding of the problem is another classic IETF problem.  Seems to me we don't have good ways to interact that can quickly get to a rough shared understanding of the problem.
[14:40:20] <Dave Crocker> meetecho: tnx.  did I miss that link in the documentation?
[14:40:25] <bortzmeyer> mcr:  suggested in Yokohama https://www.ietf.org/proceedings/94/slides/slides-94-dnsop-8.pdf
[14:40:59] <Meetecho> just noticed it only mentions the self-test for presenters, and not for participants in the queue as well, we'll update it, sorry about that
[14:41:09] <DanYork> Ted Lemon at mic
[14:41:13] <Suzanne (co-chair)> @dave it;s also not necessarily an accident that the people kicking this around include IAB members, who were thinking about itneresting architectural problems but also what Ralph just said about community engagement
[14:41:26] <Dave Crocker> meetecho: tnx!
[14:41:31] Steve Olshansky leaves the room
[14:41:53] <Dave Crocker> suzanne: ack.
[14:42:02] <Ralph> I'm not convinced IRTF is a good process for getting to the rough community understanding.
If I were in a small group of physically co-located people, I'd get us all in a room, apply some well-known collaboration techniques involving postit notes, a white board, and perhaps even a trained guide.
Dunno how we do it in our distributed, asynchronous environment.
[14:43:03] <DanYork> Ralph... agreed
[14:43:06] <Ted.h> Yay, query minimization!
[14:43:16] <Ted.h> Thanks, dprivers!
[14:43:16] <mcr> two-part DNS queries.
[14:43:38] Dan York 2 joins the room
[14:43:58] <DanYork> Ed Lewis at mic
[14:44:41] <Dave Crocker> Ralph: this realm of topic/s/ is complicated and seems not to be well-understood across participants.  It needs quite a bit of discussion, and yes, some of it needs to be with a highly responsive medium, like a physical room, with or without beer or wine.  But my guess is that this won't resolve with any one such meeting.  It's gonna take time.
[14:44:58] ken carlberg leaves the room
[14:45:57] <DanYork> Andrew Sullivan at mic
[14:46:57] <DanYork> "hey, IAB, go do your job"
[14:47:09] <DanYork> Ted Lemon at mic
[14:48:22] <Andrew Sullivan> The IAB document was the sort of "considerations for which of these one does".  I have no terribly strong desire for the IAB to "hold on" to work, though
[14:48:42] <Andrew Sullivan> if we think there is energy to do one of those "considerations" document as an IETF product, that's how we ought to do it
[14:49:50] Ted.h leaves the room
[14:50:30] <DanYork> Dan York at mic
[14:50:39] <DanYork> Ted Hardie at mic
[14:50:45] <Ralph> Dave: agreed that it will take time.  In this case, I wonder if there are new ways of working that we could devise to help us make progress faster.
[14:52:03] <DanYork> Ted argues for URI registrations
[14:52:03] <Andrew Sullivan> I don't care whether we say it's part of the domain name space.  My point is that alt fits the "domain name pattern" which is different from the URI scheme pattern or the .well-known pattern or whatever
[14:52:48] <DanYork> "blazed goat tracks"
[14:53:17] <hildjj> that's the name of my music studio
[14:53:41] <Ralph> I thought it was the name of the state road department in Massachusetts.
[14:54:02] Ted.h joins the room
[14:54:05] <Ralph> .alt solves what *some people* think is the problem.
[14:54:07] <Dan York 2> Paul Hoffmann at mic
[14:54:16] <Andrew Sullivan> yeah, and I didn't actually say that :)
[14:54:32] <Suzanne (co-chair)> .alt probably does solve some problems some people have
[14:54:58] Steve Olshansky joins the room
[14:55:02] <Dave Thaler> re "Ted argues for URI registrations", more closer to the opposite... the bar has to be low to create contexts if you want to carve out namespace
[14:55:06] <Ted.h> @Dan York I was not arguing for URI registrations.  I was pointing out that collisions went down when we lowered the friction to register and stopped trying to use protocol policing at the registration step.  Sorry that I was not clear
[14:55:13] <Dave Thaler> was Ted's point (which I agree with)
[14:55:27] rlb joins the room
[14:55:27] <Dave Crocker> Ralph: administratively, I suspect this initially needs to be in the style of a design team, if only to avoid IETF administrative overhead.
[14:55:59] <Dave Crocker> Ralph: the obvious first step is to in fact schedule a day in a rooom.  Absent that, a day in a virtual room.
[14:56:08] <Ralph> Suzanne: certainly .alt is considered to solve some problems some people believe to comprise the entire Problem.  Some people disagree.
[14:57:16] <Dan York 2> Aaron Falk at mic
[14:57:16] <bortzmeyer> Ralph: Suzanne: c
[14:57:28] Brian Trammell leaves the room
[14:57:36] <bortzmeyer> Ralph: Suzanne: sorry, I meant the biggest problem with .alt is that it does not exist (and I'm not sure it will)
[14:58:46] <Dave Crocker> Once upon a time, the IRTF served the purpose of gestating work for the IETF.  Apparently it doesn't have that goal these days, as Falk just noted.  So perhaps we need the IGTF (g=gestation)?
[14:59:01] <DanYork> Andrew Sullivan at mic
[14:59:39] <Brian Haberman> @Dave : No, I think Aaron is saying that the IETF just can't throw issues over the wall and expect the IRTF to figure it out.
[14:59:43] <DanYork> Ted.h: Yes... I started writing trying to take notes... and realized only after I typed that that you were reversing the position
[14:59:57] <DanYork> and then didn't follow up.  Thank you for clarifying
[15:01:03] <DanYork> Brian Haberman: Yes, I agree that was what Aaron said
[15:01:08] aaron joins the room
[15:01:12] <DanYork> Wes Hardaker at mic
[15:01:21] naptee leaves the room
[15:01:43] <DanYork> Wes suggests some form of the nsswitch.conf notion
[15:02:33] <DanYork> Wes notes that DNS is already fractured with split view DNS
[15:03:00] <aaron> Dave: there are examples of gestation but there are also examples of researchers mostly sharing their work.  It's hard to get gestation.  And, fWIW, this doesn't sound like research to me.  
[15:03:00] Ralph leaves the room
[15:03:19] <Matthew Pounsett> Did Wes just propose the Domain Context System?
[15:03:29] Brian Trammell joins the room
[15:03:55] <Brian Haberman> @aaron : Research would be investigating a new "system" that handles names, labels, contexts, etc.
[15:04:11] <Matthew Pounsett> My browser doesn't support gopher://
[15:04:16] Ralph joins the room
[15:04:32] naptee joins the room
[15:04:44] <aaron> My Chrome doesn't either
[15:05:27] <DanYork> PHB at mic
[15:05:48] <Ted.h> That's not what I mean when I say resolution context
[15:06:17] <aaron> @Brian: research works best when you don't have a clear idea of what you trying to accomplish.  Exploration.  IMO, the IETF has a clear enough idea that problem solving process could happen in the iETF.  The benefit of that would be the emphasis on consensus on the solutions
[15:06:20] <Ted.h> I really mean "what software stack do I invoke to get this resolved"
[15:06:45] <tale> The point someone made a while back about getting context wrong both leaking information and making spoofing more possible is salient.
[15:06:50] Timothy Morizot joins the room
[15:06:58] <Ted.h> @tale very true.
[15:07:08] <DanYork> me at mic
[15:07:12] <Dave Thaler> @Ted: just protocol stack not sufficient.  mdns has multiple rsolution contexts (one per link).  So does DNS if you're multihomed to both split dns (e.g. a VPN) and the global DNS
[15:07:18] <DanYork> Tim Shepard at mic
[15:08:39] <Ted.h> @dthaler  Fair enough.  I was thinking about the protocol stack having the ability to identify contexts within its own resolution contexts from the identifier passed to it.  That may not be universal.
[15:08:53] <DanYork> Ted Lemon at mic
[15:09:03] wu teng leaves the room
[15:09:27] <Dave Thaler> @Ted: yes.  so the resolution context conceptually has (at least) two layers... protocol and then resolution context within protocol
[15:09:30] <tale> btw, we need a bash for ietf mtgs.  "For a while everything had a colon!" - Ted.h
[15:10:06] <DanYork> "users want what they want" - Andrew
[15:10:24] <DanYork> Andrew: "Can we do some practical things for developers?"
[15:12:29] Eliot Lear leaves the room
[15:12:39] monica joins the room
[15:12:58] KY Hong leaves the room
[15:13:09] RjS leaves the room
[15:13:17] KY Hong joins the room
[15:13:41] <Dave Crocker> Obvious suggestion:  non-wg mailing list.  just to provide a common forum unless/until activities converge.
[15:14:01] <Dan York 2> Dave Crocker: We have the "arcing" list now… is that sufficient?
[15:14:18] monica leaves the room
[15:14:42] Wolfgang Beck leaves the room
[15:14:51] <Suzanne (co-chair)> there is a BOF llist. There's also a public list associated with the IAB Names & Identifiers program, where we've been trying to look at some of these things (hence the BOF), inip-discuss
[15:14:56] Steve Olshansky leaves the room
[15:15:08] <Brian Haberman> I approved the arcing mailing list so that we could have focused discussion on these topics.
[15:15:58] Ted.h leaves the room
[15:15:59] <Suzanne (co-chair)> @Brian yes I was assuming the list will remain available, thanks
[15:16:14] <DanYork> Leif Johansson at mic
[15:16:49] Ted.h joins the room
[15:17:24] Ralph leaves the room
[15:17:27] <Dave Crocker> suzanne: tnx.
[15:17:28] Ralph joins the room
[15:17:34] <bortzmeyer> "Eh, software authors, do you want to swap your TLD to a strange, unproven, untested and unimplemented system?"
[15:17:39] rlb leaves the room
[15:17:52] <Ted.h> @bortzmeyer:  "Sign me up!"
[15:17:58] <Dan York 2> bortzmeyer: yes, that's an issue
[15:18:09] <Dan York 2> Ted Lemon at mic
[15:18:42] <Dan York 2> Ted mentions his homenet draft
[15:19:00] <Dan York 2> Ted at mic
[15:19:10] <Dan York 2> Dave Thaler at mic
[15:19:14] <bortzmeyer> Dan York 2: too many Teds
[15:19:24] <Dan York 2> (er… Dave *will* be at mic)
[15:19:25] <bortzmeyer> Dan York 2: Ted in which context?
[15:19:39] <Matthew Pounsett> I kinda like the punycode example.   Pardon my jabber client which keeps changing double hyphens to EM-dashes, but … Assigning [\w\w]— prefixes to resolution contexts would mean that you could sideline resolution at any point in the tree.   If xz— indicates a name to be resolved in some other scheme, you could have www.xz—foo.example.com <http://www.xz—foo.example.com> or www.example.xz—foo <http://www.example.xz—foo>.
[15:20:34] m&m leaves the room: Disconnected: closed
[15:20:59] <Ted.h> Do not confuse a negotiating position for a reasonable solution.
[15:21:23] <Dave Crocker> MIC:  Ted H has the premise that our inventing a mechanism for avoiding/resolving naming contexts will be useful.  I suspect that we do not have a basis for expecting that it will be sufficienty appealing and usable for real-world adoption.
[15:21:30] aaron leaves the room
[15:21:43] <Dan York 2> will relay
[15:22:30] Steve Olshansky joins the room
[15:23:22] <Ted.h> If the choice is FCFS under .alt and a long slog for the root, the calculus may shift for anyone that has not already achieved wide deployment.
[15:23:34] <Dan York 2> Mark Nottingham at mic
[15:23:36] <Dave Thaler> agree with .alt for new
[15:23:52] <Ted.h> "Beef and Malbec" would be a great band name
[15:23:55] <Dave Thaler> people already do other stuff today that has to keep working
[15:24:38] nllz joins the room
[15:24:40] nllz leaves the room
[15:24:52] Steve Olshansky leaves the room
[15:24:55] <Dan York 2> WRITE TEXT :-)
[15:25:03] Brian Haberman leaves the room
[15:26:14] naptee leaves the room
[15:26:16] Brian Trammell leaves the room
[15:26:17] Yoshiro Yoneya leaves the room
[15:26:17] John Levine leaves the room
[15:26:18] Andrew Sullivan leaves the room
[15:26:21] philippe fouquart leaves the room
[15:26:24] Ted.h leaves the room
[15:26:26] Matthew Pounsett leaves the room
[15:26:35] Peter Koch leaves the room
[15:26:38] tale leaves the room
[15:26:40] Ralph leaves the room
[15:26:48] KY Hong leaves the room
[15:26:51] bortzmeyer leaves the room
[15:27:17] Andrew McConachie leaves the room
[15:27:18] Roger Carney leaves the room
[15:27:44] Tomohiro Fujisaki leaves the room
[15:27:55] Michael Richardson leaves the room
[15:27:56] bortzmeyer joins the room
[15:27:59] Barry Leiba leaves the room
[15:28:14] Dave Crocker leaves the room
[15:28:40] Suzanne (co-chair) leaves the room
[15:28:54] Timothy Morizot leaves the room
[15:29:10] Jim Galvin leaves the room
[15:30:10] hildjj leaves the room
[15:30:47] resnick leaves the room
[15:33:00] Meetecho leaves the room
[15:34:07] m&m joins the room
[15:34:48] Ralph joins the room
[15:35:02] m&m leaves the room
[15:35:21] mellon leaves the room
[15:39:07] resnick joins the room
[15:39:10] Dave Thaler leaves the room
[15:41:15] Steve Olshansky joins the room
[15:41:31] Steve Olshansky leaves the room
[15:43:30] rlb joins the room
[15:44:58] Lyman Chapin leaves the room
[15:46:59] Eliot Lear joins the room
[15:48:39] Brian Trammell joins the room
[15:50:16] resnick leaves the room
[15:51:10] tale joins the room
[15:57:06] Brian Trammell leaves the room
[15:58:06] Steve Olshansky joins the room
[15:58:40] Ralph leaves the room
[16:02:21] hildjj joins the room
[16:02:33] Suzanne (co-chair) joins the room
[16:02:52] Olafur joins the room
[16:03:36] nllz leaves the room
[16:03:57] Steve Olshansky leaves the room
[16:04:10] Olafur leaves the room
[16:04:17] Olafur joins the room
[16:04:49] bortzmeyer leaves the room
[16:06:33] Barry Leiba joins the room
[16:12:13] Steve Olshansky joins the room
[16:19:09] Eliot Lear leaves the room
[16:25:47] aaron joins the room
[16:29:48] Suzanne (co-chair) leaves the room
[16:33:38] Dave Thaler joins the room
[16:34:02] Dave Thaler leaves the room
[16:34:57] Steve Olshansky leaves the room
[16:35:59] Dan York 2 leaves the room
[16:36:22] Dan York 2 joins the room
[16:38:12] aaron leaves the room
[16:40:18] rlb leaves the room
[16:43:01] hildjj leaves the room
[16:43:09] aaron joins the room
[16:43:10] Olafur leaves the room
[16:43:43] Barry Leiba leaves the room
[16:43:50] Barry Leiba joins the room
[16:46:05] aaron leaves the room
[16:50:01] aaron joins the room
[16:50:05] Barry Leiba leaves the room
[16:51:58] aaron leaves the room
[16:57:40] RjS joins the room
[16:58:03] RjS leaves the room
[16:58:57] Barry Leiba joins the room
[17:03:12] Jim Galvin joins the room
[17:04:00] rlb joins the room
[17:07:53] Barry Leiba leaves the room
[17:12:49] rlb leaves the room
[17:13:59] Matthew Pounsett joins the room
[17:14:03] Matthew Pounsett leaves the room
[17:14:12] rlb joins the room
[17:14:13] Matthew Pounsett joins the room
[17:14:58] rlb leaves the room
[17:21:10] Jim Galvin leaves the room
[17:22:26] aaron joins the room
[17:22:40] DanYork leaves the room
[17:22:42] Dan York 2 leaves the room
[17:27:02] Dan York 2 joins the room
[17:31:42] aaron leaves the room
[17:32:08] bortzmeyer joins the room
[17:35:30] Matthew Pounsett leaves the room
[17:35:49] Eliot Lear joins the room
[17:46:50] mcr leaves the room
[17:56:08] Jim Galvin joins the room
[18:03:10] Dan York 2 leaves the room
[18:10:19] Dan York 2 joins the room
[18:12:44] Matthew Pounsett joins the room
[18:22:11] Jim Galvin leaves the room
[18:29:30] Eliot Lear leaves the room
[18:35:00] tale leaves the room
[18:39:46] Dan York 2 leaves the room
[18:54:09] Matthew Pounsett leaves the room
[18:56:19] Matthew Pounsett joins the room
[18:57:49] Jim Galvin joins the room
[19:09:40] Jim Galvin leaves the room
[19:09:55] tale joins the room
[19:11:28] bortzmeyer leaves the room
[19:14:23] Dan York 2 joins the room
[19:39:13] bortzmeyer joins the room
[19:39:39] mellon joins the room
[19:40:41] Matthew Pounsett leaves the room
[19:44:38] Matthew Pounsett joins the room
[19:46:46] Dan York 2 leaves the room: Replaced by new connection
[19:47:07] Dan York 2 joins the room
[19:49:41] Dan York 2 leaves the room: Replaced by new connection
[19:49:59] Dan York 2 joins the room
[20:16:26] tale leaves the room
[20:27:23] tale joins the room
[20:34:22] mellon leaves the room
[20:40:06] mellon joins the room
[20:40:08] mellon leaves the room
[20:40:42] mellon joins the room
[20:57:03] mellon leaves the room
[21:43:02] Matthew Pounsett leaves the room
[22:15:51] Dan York 2 leaves the room
[22:28:16] bortzmeyer leaves the room
[22:28:42] bortzmeyer joins the room
[22:30:07] Dan York 2 joins the room
[22:36:46] tale leaves the room
Powered by ejabberd - robust, scalable and extensible XMPP server Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!