dbound - Friday, November 14, 2014

dbound

dbound@jabber.ietf.org

Friday, November 14, 2014< ^ >

=JeffH has set the subject to: dbound Bof IETF-89

Room Configuration

Room Occupants

GMT+0
[00:13:33] Kurt Andersen joins the room
[01:04:40] Kurt Andersen leaves the room
[01:21:32] Kurt Andersen joins the room
[01:58:10] Kurt Andersen leaves the room
[02:18:53] Kurt Andersen joins the room
[02:19:20] <Kurt Andersen> checking to see if this room is live
[02:28:53] <Kurt Andersen> For the slides and audio: join https://bluejeans.com/966106876
or audio only: 408 740 7256, meeting ID 9661-06876
[02:33:47] tdraegen joins the room
[02:36:05] =JeffH joins the room
[02:37:25] <=JeffH> how do we reset the subject ?
[02:38:53] <Kurt Andersen> =JeffH - don't know, but it blames you for the current value :-)
[02:40:48] <=JeffH> yep, blame me. i can take it.
[02:45:59] zwicky joins the room
[02:49:36] Yoshiro Yoneya joins the room
[02:49:40] DanYork joins the room
[02:50:03] DanYork has set the subject to: DBOUND Bof IETF91
[02:51:33] msk joins the room
[02:53:42] dcrocker joins the room
[02:54:31] <Kurt Andersen> question in the room: what about private-private boundaries - will address later
[02:55:07] jtrentadams joins the room
[02:56:19] <=JeffH> what is the bluejeans URI ?
[02:57:19] <Kurt Andersen> =JeffH: https://bluejeans.com/966106876 but please be sure to mute your system's speaker and mic
[02:59:47] <=JeffH> thanks
[03:05:47] <dcrocker> maybe we should use this second chat room for @dbad-attitude
[03:06:04] <jtrentadams> So, wait, the text chat is happening in Blue Jean?
[03:06:09] <dcrocker> right
[03:06:13] <jtrentadams> :P
[03:06:17] <jtrentadams> Lame
[03:06:25] <DanYork> It would be better if the text chat could happen HERE instead of bluejeans
[03:06:29] <jtrentadams> I ain't installin' no silly Blue Jean plugins
[03:06:37] <dcrocker> these are multitasked times.  short attention span theatre.
[03:06:37] <=JeffH> huh, blue jeans chat just went away for me
[03:06:41] <jtrentadams> +1 Dan
[03:06:51] <DanYork> That way others of us don't have to launch bluejeans
[03:07:09] <dcrocker> well it's ok for the bj chat to just be for bj control issues.
[03:07:26] <dcrocker> lunch bluejeans?  chew on it?
[03:07:27] <jtrentadams> My corporate overlords don't like it when I run stuff like that
[03:07:27] <DanYork> sure
[03:08:25] <Kurt Andersen> the idea is just to use BJ chat for BJ control issues, but Jothan had problems logging into this room
[03:08:53] jtrentadams sighs
[03:10:47] <=JeffH> https://publicsuffix.org/list/effective_tld_names.dat
[03:11:38] <=JeffH> dc: MIC:  There is the possibility of multiple boundaries along an ancestry.  Pub/priv is merely an interesting subset.
[03:11:51] <dcrocker> tnx
[03:12:26] Barry Leiba joins the room
[03:12:58] <Kurt Andersen> Q from room: does effective include complete and accurate? A: yes
[03:13:13] <=JeffH> bluejeans just kicked me off again....
[03:13:19] <=JeffH> we should just use this
[03:13:31] <jtrentadams> +1
[03:13:58] <Kurt Andersen> =jeffh: yes - I'll babysit the chat on BJ for control issue or relay from Jothan
[03:14:50] <=JeffH> great thx
[03:15:32] <dcrocker> (I just realized that this sesssion is instead of you guys having bits 'n' bytes.)
[03:19:33] <dcrocker> MIC:  Sounds like a task might be to use the 'framework' the the PSL is using, to characterize this construct, but give it a better name?
[03:20:22] <Kurt Andersen> dcrocker: did you see =JeffH's email from this afternoon to the list?
[03:20:31] <Kurt Andersen> (the list == dbound)
[03:20:38] <dcrocker> @kurt, sorry, no not yet.
[03:20:52] <dcrocker> 'attended' too many other sessions.
[03:23:18] <Kurt Andersen> dcrocker - I think that Jeff is advocating the term "effective TLD" to characterize what the PSL actually identifies
[03:24:31] <jtrentadams> Maybe it'd help to move beyond the naming of the list?
[03:24:48] <Kurt Andersen> jeff's suggestion to forget the "public" term got a bunch of hands raised for the mic
[03:24:57] <dcrocker> MIC:  how about 'ineffective' TLD?  but seriiously, 'public' doesn't work, but I don't know what the best alternative is.  Not immediately in love with 'effective' but let's switch to it until someone comes up with something better.
[03:25:45] <tdraegen> TLDelimiter
[03:25:56] <tdraegen> TLDelineator
[03:26:25] <jtrentadams> MIC: Can we postpone this conversation so we can keep moving forward on the meat of the topic?
[03:26:39] <dcrocker> @tim:  first let's formulate a definition we like.  then let's look for a label.  starting with the label distorts things, absent a good definition.
[03:26:58] <tdraegen> agreed
[03:27:27] John Rae-Grant joins the room
[03:27:31] Olafur joins the room
[03:27:42] <dcrocker> MIC: understanding why subordinate boundaries are useful in the PSL will probably be useful. (Not now, of course.)
[03:28:03] <jtrentadams> +1 dcrocker
[03:28:29] <Olafur> The DNS term for this is "Delegation Domain"
[03:29:46] <=JeffH> +1
[03:29:47] <=JeffH> thx olafur
[03:30:05] <Olafur> The domains that you can not get delegations in are "Vanity Domains" :-)
[03:30:26] john.levine joins the room
[03:30:27] Peter Koch joins the room
[03:30:33] <John Rae-Grant> +1 BobTheWonderDog
[03:31:24] <Kurt Andersen> relay from BJ chat:
[03:31:25] <=JeffH> the PSL got us somewhere, but we need to learn from it and move on and not get distracted by its warts
[03:31:25] <Kurt Andersen> Jothan Frakes3 mins agoI think the dialog here should focus on the replacement not the problems with what the internet made in a vacuum of a solution
Dave Crocker1 min agoUnderstanding the existing stuff is not to focus on the problems but to make sure we are anchored in demonstrated need.
[03:32:02] <Kurt Andersen> =jeffH: that's why I responded on the list yesterday that we should start with problem substatement 3
[03:32:15] <dcrocker> (hmmm. I'm not finding Jeff's posting today on dbound that suggests a name other than PSL.  I see two message from him today butneither does that.)
[03:32:35] <=JeffH> yeah -- I wish i'd had time to contrib to the prob stmt work
[03:32:40] <=JeffH> by casey
[03:32:53] <dcrocker> oh.
[03:33:33] <Kurt Andersen> dcrocker: [Dbound] HTTP cookie processing wrt "public suffixes" - time stamp around 16:28 HST
[03:34:07] <Kurt Andersen> there's a later thread about multiple-PSL equivalents
[03:36:52] <=JeffH> http://www.ietf.org/mail-archive/web/dbound/current/msg00165.html
[03:37:14] <dcrocker> @jeffh - ack. tnx.
[03:39:03] <=JeffH> can we please post jfrakes' comments into here ? thx
[03:39:04] <dcrocker> @jeffh - 'effective' bothers me.  it doesn't seem a natural choice, in terms of the semantics of effective.  first alternative that occurs to me is based on thinking that they are potentially lower than a TLD and so:  Intermediate Level Domain.  ILD.
[03:39:21] <Kurt Andersen> =JeffH - yes, will do
[03:39:35] <dcrocker> Or perhaps:  Boundary Domain.
[03:39:45] <Kurt Andersen> Jothan Frakes8 mins agoso, being part of the group at mozilla, I know a little bit about this
Jothan Frakes6 mins agothose of the fellow editors of the PSL like myself all refer to it as the least awful solutions available
Jothan Frakes5 mins agoAMEN to speaker
Jothan Frakes5 mins ago+10000
Jothan Frakes3 mins agoyes , think of PSL like a buffet ... take what is interesting or helpful
[03:40:21] <=JeffH> thx
[03:40:28] Barry Leiba leaves the room
[03:42:34] Barry Leiba joins the room
[03:46:52] <=JeffH> http://tools.ietf.org/html/draft-sullivan-domain-policy-authority-01
[03:50:09] <jtrentadams> Has a draft "problem statement" been written anywhere other than in the slides being presented?
[03:50:36] <tdraegen> yes, posted to list
[03:50:37] <Kurt Andersen> jtrentadams: see Casey's email from earlier this week - Monday ish
[03:50:48] <jtrentadams> thanks
[03:50:52] <tdraegen> http://www.ietf.org/mail-archive/web/dbound/current/msg00141.html
[03:50:55] <Kurt Andersen> references a doc he wrote w/ John L
[03:51:48] John Rae-Grant leaves the room
[03:52:01] <dcrocker> @MIC:  this should move toward working group when things coalesce around an approach.
[03:52:03] John Rae-Grant joins the room
[03:52:44] <dcrocker> MIC:  That is, we are still wandering around this topic.  We don't have enough focus yet.
[03:52:46] <dcrocker> IMO.
[03:52:50] John Rae-Grant leaves the room
[03:53:11] <tdraegen> I'd like to see more fleshing out of the cross-domain use cases that are not related to cookies
[03:53:17] <=JeffH> problem stmt needs major surgery
[03:53:25] <DanYork> tdraegen: +1
[03:54:01] <Kurt Andersen> =JeffH - heart transplant?
[03:54:35] <jtrentadams> Problem statement: "Relying on the PSL sux and needs to be fixed."
[03:55:37] <dcrocker> @trent:  my reading of the room is that there is a split between a narrow scope such as you've given, versus a much more ambitious/general conceptual problem to attack.
[03:55:53] <tdraegen> is "same origin" the "use case that is not related to cookies"?
[03:56:13] <=JeffH> I claim that section 1 of draft-sullivan-domain-policy-authority-01 provides a solid basis for a more general problem stmt
[03:56:14] <tdraegen> eg, for cross-site forgery prevention?
[03:56:44] <dcrocker> I think this reduces down to "strictly ancestral" vs. "ancestral or cross-domain".  It might also distinguish between only defining boundaries vs. providing a policy specification framework.
[03:56:47] <Kurt Andersen> =JeffH ++
[04:00:05] <jtrentadams> Problem Statement: "Need to replace the PSL with a functional equivalent that defines dynamic policy boundaries between domains in an extensible way."
[04:00:12] John Rae-Grant joins the room
[04:00:33] <=JeffH> tdraegen: SOP in web UAs is a potential use case
[04:00:37] <Kurt Andersen> From the cited I-D: What appears to be needed is a mechanism to determine policy boundaries in the DNS.
[04:01:11] Olafur leaves the room
[04:02:47] Olafur joins the room
[04:03:26] <msk> ACBS
[04:04:44] <jtrentadams> Problem Statement: "Some Internet client entities on the Internet make inferences about the administrative relationships among services on the Internet based on the domain names at which they are offered. At present, it is not possible to ascertain organizational administrative boundaries in the DNS, therefore such inferences can be erroneous in various ways. Mitigation strategies deployed so far will not scale.  A solution needs to be developed that provides a means to make explicit assertions regarding the administrative relationships between domain names."
[04:06:03] Franck Martin joins the room
[04:07:18] <=JeffH> the more general use case is the origin the code was sourced from implies a policy domain.
[04:08:25] Suzanne joins the room
[04:09:15] Suzanne leaves the room
[04:12:19] John Rae-Grant leaves the room
[04:12:47] <jtrentadams> Sign up TinTin for some work!
[04:12:58] <Franck Martin> :D
[04:13:01] <=JeffH> :)
[04:13:53] <jtrentadams> Short & Sweet
[04:14:13] <jtrentadams> "Make bad things go away."
[04:14:22] <=JeffH> lol
[04:14:41] <dcrocker> "Make good things come and stay"
[04:14:58] <=JeffH> +1
[04:15:00] <jtrentadams> "Feed good & starve bad."
[04:15:22] <dcrocker> so a cold is good and a fever is bad?
[04:15:42] <jtrentadams> Cold Beer Good
[04:17:59] <Kurt Andersen> Jothan Frakes3 mins agoI am absent a corporate sponsor but can aid in answering q's about PSL
Jothan Frakes3 mins agoas I have a lot of the tribal knowledge on use cases or why something is there
Jothan Frakesjust nowI HAVE USE CASES
[04:18:31] <jtrentadams> Send use cases in using emojis
[04:19:41] tdraegen leaves the room
[04:19:46] <dcrocker> bye
[04:19:46] msk leaves the room
[04:19:48] dcrocker leaves the room
[04:20:06] Peter Koch leaves the room
[04:20:09] Yoshiro Yoneya leaves the room
[04:20:11] Olafur leaves the room
[04:20:41] Franck Martin leaves the room
[04:20:41] Kurt Andersen leaves the room
[04:20:54] jtrentadams leaves the room
[04:21:58] john.levine leaves the room
[04:23:32] Barry Leiba leaves the room
[04:26:02] DanYork leaves the room
[04:33:11] zwicky leaves the room
[05:32:30] Yoshiro Yoneya joins the room
[05:34:53] Yoshiro Yoneya leaves the room
[08:23:39] John Rae-Grant joins the room
[08:24:05] John Rae-Grant leaves the room
[08:47:02] msk joins the room
[09:05:20] msk leaves the room
[16:37:53] Kurt Andersen joins the room
[17:06:22] Olafur joins the room
[17:20:44] Olafur leaves the room
[17:26:17] Kurt Andersen leaves the room
[17:43:19] zwicky joins the room
[17:50:17] Kurt Andersen joins the room
[17:51:01] jtrentadams joins the room
[17:51:17] zwicky leaves the room
[17:52:45] jtrentadams leaves the room
[17:53:17] jtrentadams joins the room
[18:02:05] tdraegen joins the room
[18:04:29] jtrentadams joins the room
[18:04:42] jtrentadams leaves the room
[18:05:41] jtrentadams leaves the room
[18:06:49] jtrentadams joins the room
[18:16:17] tdraegen leaves the room
[18:29:35] tdraegen joins the room
[18:34:41] tdraegen leaves the room
[18:36:18] jtrentadams leaves the room
[18:38:02] zwicky joins the room
[18:38:16] zwicky leaves the room
[18:39:44] Franck Martin joins the room
[18:55:48] Franck Martin leaves the room
[20:04:48] Kurt Andersen leaves the room