IETF
precis@jabber.ietf.org
Friday, August 2, 2013< ^ >
yone has set the subject to: précis wg :) | http://tools.ietf.org/wg/precis/ | IETF 86 | slides at https://datatracker.ietf.org/meeting/86/materials.html#session.group-precis | audio at http://ietf86streaming.dnsalias.net/ietf/ietf861.m3u
Room Configuration
Room Occupants

GMT+0
[06:07:55] yone joins the room
[06:17:52] yone has set the subject to: précis wg :) | http://tools.ietf.org/wg/precis/ | IETF 87 | slides at https://datatracker.ietf.org/meeting/87/materials.html#precis | audio at http://ietf87streaming.dnsalias.net/ietf/ietf877.m3u
[06:56:31] Florian Zeitz joins the room
[06:57:37] Florian Zeitz leaves the room
[06:58:04] Florian Zeitz joins the room
[07:01:32] hildjj joins the room
[07:01:33] m&m joins the room
[07:01:53] <m&m> taking minutes at http://tools.ietf.org/wg/precis/minutes
[07:02:18] stpeter joins the room
[07:03:27] Andrew Sullivan joins the room
[07:03:35] <Andrew Sullivan> sorry for my late arrival.  Scribing here
[07:03:46] <Andrew Sullivan> On WG Doc Status Update slide
[07:04:26] <Andrew Sullivan> Proposed Agenda
[07:04:49] <Andrew Sullivan> Peter St Andre to mic on Framework
[07:04:59] <Andrew Sullivan> Slide 2
[07:06:05] <Andrew Sullivan> Some Ruby code from IDNAbis has been patched.  StPeter started code in Python, but not all pretty.  
[07:06:29] <Andrew Sullivan> Code work indicates problems in framework spec.
[07:07:07] <Andrew Sullivan> Code allowed detection of issues.
[07:07:15] <Andrew Sullivan> Slide 3.  
[07:07:38] <Andrew Sullivan> Slide 4
[07:08:03] lef_jp joins the room
[07:08:09] nemo joins the room
[07:08:13] <Andrew Sullivan> The issue seems to be the order in which rules are checked
[07:08:31] <Andrew Sullivan> framework draft updated to match proposed changes
[07:08:41] <Andrew Sullivan> (back on slide 2)
[07:08:51] resnick joins the room
[07:09:08] <Andrew Sullivan> Slide 5
[07:09:29] <Andrew Sullivan> Slide 6
[07:09:52] <Andrew Sullivan> Slide 7
[07:10:25] <Andrew Sullivan> PrecisMaker.py doesn't have output in XML format yet
[07:11:56] <Andrew Sullivan> Goal is to make as compatible as possible with stringprep + Unicode 3.2
[07:12:23] <Andrew Sullivan> Marc Blanchet at mic as individual
[07:12:41] <Andrew Sullivan> did some comparisons with IDNA
[07:13:00] <Andrew Sullivan> quite a few differences, maybe not appropriate for this work
[07:15:46] <Andrew Sullivan> That was me at mic asking about ordering  compared to IDNA.  Sounds like not relevant.
[07:15:54] <Andrew Sullivan> slide 7 again
[07:18:11] <Andrew Sullivan> Question from Chair: do we need to wait for output of tables before LC or not?
[07:19:03] <Andrew Sullivan> Peter would like bullet 2 complete before moving ahead
[07:19:27] <Andrew Sullivan> Chairs: beg of August, so WGLC in a few weeks?
[07:20:04] <Andrew Sullivan> Proposing to start WGLC now
[07:21:12] <Andrew Sullivan> No, 3d week of August
[07:21:23] <Andrew Sullivan> So, plan for WGLC 3d week of August
[07:22:01] <Andrew Sullivan> But if the response is low, it may move to 1st week of Sept.
[07:22:46] <Andrew Sullivan> Back on agenda.  No significant issues on nickname.  Need to complete write up then go to IETF LC
[07:23:47] <Andrew Sullivan> No comments on this.  Ask for AD ok on going ahead given framework not out
[07:23:52] <Andrew Sullivan> make a normative reference?
[07:23:58] <Andrew Sullivan> AD Resnick @ mic
[07:24:19] <Andrew Sullivan> Question: would people who are doing LC review want both docs?  
[07:24:31] <Andrew Sullivan> Plan to hold IESG review until framework is available
[07:24:38] <Andrew Sullivan> Willing to do AD review
[07:24:44] <Andrew Sullivan> and put in wait state
[07:25:31] <Andrew Sullivan> next, Mapping
[07:25:43] nemo leaves the room
[07:25:44] <Andrew Sullivan> Takahiro Nemoto presenting
[07:25:52] <Andrew Sullivan> slide 1
[07:26:03] <Andrew Sullivan> slide 2
[07:27:02] <Andrew Sullivan> Slide 3
[07:27:36] <Andrew Sullivan> Argument is about where mapping is handled
[07:28:24] <Andrew Sullivan> requesting comment
[07:28:33] <Andrew Sullivan> and requesting WGLC
[07:28:47] <Andrew Sullivan> Peter St Andre at mic
[07:28:59] <Andrew Sullivan> framework just says "could be other mappings"
[07:29:14] <Andrew Sullivan> supports putting in this doc
[07:29:25] <Andrew Sullivan> Pete Resnick
[07:29:32] <Andrew Sullivan> might as well WGLC
[07:29:36] <Andrew Sullivan> since silence
[07:30:15] <Andrew Sullivan> Chairs asking for sense of room: objection to WCLC?
[07:30:32] <Andrew Sullivan> none raised
[07:31:30] <Andrew Sullivan> timing with framework?  Rev then WGLC
[07:32:05] nemo joins the room
[07:32:18] <Andrew Sullivan> Next: Peter St Andre and saslprepbis
[07:32:26] <Andrew Sullivan> slide 2
[07:33:03] <Andrew Sullivan> case mapping from MUST to MAY
[07:35:47] <Andrew Sullivan> I was at mic suggesting maybe "SHOULD"
[07:36:00] <Andrew Sullivan> Pete Resnick at mic asking for real examples
[07:36:15] <Andrew Sullivan> much talk of foot-guns and the danger of them
[07:37:45] <Andrew Sullivan> Oops, that was me at mic, sorry for missing my name
[07:37:55] <Andrew Sullivan> I said I will send text
[07:38:11] <Andrew Sullivan> Slide 3
[07:39:30] <Andrew Sullivan> More stuff that should be up in application. That is, authorization IDs controlled by application protocol.
[07:39:40] <Andrew Sullivan> again a layering argument
[07:39:46] <Andrew Sullivan> slide 4
[07:40:09] <Andrew Sullivan> slide 5
[07:40:28] <Andrew Sullivan> questions about Non-SASL application protocols
[07:40:47] <Andrew Sullivan> Suggestion to use same advice for non-SASL cases
[07:41:27] <Andrew Sullivan> is it SASLprep then?  
[07:41:38] <Andrew Sullivan> (Joe Hildebrand at mic)
[07:42:05] <Andrew Sullivan> question is basically whether draft ought to make clear that this isn't just for SASL
[07:43:34] <Andrew Sullivan> no more slides
[07:43:52] <Andrew Sullivan> Question from chair: profile from framework
[07:44:20] <Andrew Sullivan> what's the status of the review by people interested.  Asked on kitten list.  
[07:44:42] <Andrew Sullivan> Yutaka (sp?) at mic
[07:44:54] <Andrew Sullivan> confused about organization.
[07:45:10] <Andrew Sullivan> is it for non-SASL things?  Also SASL-specific things
[07:46:17] <Andrew Sullivan> St Peter: tried to organize into layers, in order to make easy for "customers" of framework
[07:46:47] <Andrew Sullivan> ask how it gets used
[07:47:14] nemo leaves the room
[07:47:40] <Andrew Sullivan> need for volunteers for http auth
[07:47:52] nemo joins the room
[07:48:10] <Andrew Sullivan> we call these usages, even though perhaps should be "profile"
[07:48:56] <Andrew Sullivan> so far, says "just UTF-8" and thinking need maybe normalization
[07:49:05] <Andrew Sullivan> not so good that
[07:49:15] <Andrew Sullivan> Pete Resnick
[07:52:32] Suz joins the room
[07:52:53] Suz leaves the room
[07:53:13] Andrew Sullivan joins the room
[07:53:31] yone joins the room
[07:53:32] <Andrew Sullivan> sorry about that, jabber blew up (and I can't see anyone else here now.  So maybe still broken)
[07:53:53] <Andrew Sullivan> discussion at mic: application protocols vs. other layers
[07:54:59] <Andrew Sullivan> Need to capture general questions to ask profilers
[07:55:33] <Andrew Sullivan> Pete Resnick, for review purposes by customers, make sure this is good for _you_
[07:57:18] <Andrew Sullivan> Joe Hildebrand at mic
[07:57:21] lef_jp joins the room
[07:57:40] <Andrew Sullivan> St Peter again.  We understand docs, but maybe others don't
[07:58:27] <Andrew Sullivan> telling note about Spotify blog post
[08:05:56] <Andrew Sullivan> That was me at mic ranting about how people need help
[08:06:08] <Andrew Sullivan> with knowing what they want in i18n
[08:06:38] <Andrew Sullivan> new rev of the saslprepbis draft
[08:07:01] <Andrew Sullivan> need to look at httpauth to ensure their needs are met
[08:10:10] <Andrew Sullivan> I agreed to send some text
[08:10:35] <Andrew Sullivan> Next, oiwa-precis-httpauthprep
[08:11:05] <Andrew Sullivan> Yutaka ÔIWA
[08:11:27] <Andrew Sullivan> Background for httpauthprep
[08:11:32] <Andrew Sullivan> slide 2
[08:12:09] <Andrew Sullivan> slide 3
[08:14:26] <Andrew Sullivan> slide 4
[08:15:16] <Andrew Sullivan> slide 5
[08:16:36] <Andrew Sullivan> slide 6
[08:18:04] <Andrew Sullivan> Peter St Andre: not clear in xmpp who has to enforce the rules
[08:18:17] <Andrew Sullivan> so 6122bis tried to be clear
[08:18:28] <Andrew Sullivan> slide 7
[08:19:44] <Andrew Sullivan> slide 8
[08:20:52] stpeter joins the room
[08:21:02] <stpeter> https://datatracker.ietf.org/doc/draft-oiwa-precis-httpauthprep/ is Yutaka's I-D
[08:21:28] <Andrew Sullivan> Peter St Andre
[08:21:51] <Andrew Sullivan> comment: better not to have httpauthprep different from saslprepbis.
[08:22:46] lef_jp leaves the room
[08:23:51] <Andrew Sullivan> slide 11 (boy I'm hating this interruptive server)
[08:24:15] <stpeter> which interruptive server?
[08:24:42] <Andrew Sullivan> It may be my client.  I'm having some trouble with my ocnnection
[08:24:50] <Andrew Sullivan> slide 12
[08:24:50] <stpeter> I got disconnected, too
[08:25:12] <stpeter> I blame jabber.org
[08:25:21] <Andrew Sullivan> those slackers
[08:26:45] <Andrew Sullivan> slide 13
[08:27:52] <Andrew Sullivan> slide 14
[08:28:30] <Andrew Sullivan> slide 15
[08:29:57] <Andrew Sullivan> Peter St Andre: can't ask users to provide UTF-8
[08:30:56] <Andrew Sullivan> (BTW, Peter St Andre is properly spelled Peter Saint-Andre, but I can't type that fast.)
[08:31:36] <Andrew Sullivan> Marc Blanchet concerned about "simple" cases
[08:33:00] <Andrew Sullivan> Jonathan Lennox: send prepared string, but indifferent to how you got the prepared string
[08:33:41] Jonathan Lennox joins the room
[08:34:07] <Andrew Sullivan> Marc Blanchet: if software prepared the config, then you have in fact prepared the string
[08:34:22] <Andrew Sullivan> So you should _always_ send prepared
[08:37:01] <stpeter> Andrew Sullivan: you can type "PSA" :-)
[08:37:03] <Andrew Sullivan> (That was me at mic with advice worth what one paid for it) Jonathan Lennox: looking at this as normative/informational
[08:37:08] <Andrew Sullivan> thanks
[08:37:35] <Andrew Sullivan> Joe Hildebrand: don't think should do what IDNA did simply.  Need better advice
[08:38:48] <Andrew Sullivan> me: agree with Joe, just try to draw bright line between proto and what apps do
[08:38:50] <Andrew Sullivan> slide 16
[08:40:39] resnick joins the room
[08:42:35] <Andrew Sullivan> PSA: slide 12.
[08:42:49] <Andrew Sullivan> "client side only" concerned about it
[08:43:03] <Andrew Sullivan> slide 10
[08:43:32] <Andrew Sullivan> need to say use of usernames and passwords as input to hash functions
[08:43:39] <Andrew Sullivan> but why is this linked to not doing on server side?
[08:43:56] <Andrew Sullivan> server can do it in plain text case
[08:44:00] <Andrew Sullivan> (answer)
[08:44:15] <Andrew Sullivan> other cases, server _cannot_ do it because they don't have access
[08:44:52] <Andrew Sullivan> Marc Blanchet: touching "store" problem?
[08:45:36] <Andrew Sullivan> i.e. string when stored should be prepared
[08:46:29] <Andrew Sullivan> Pete Resnick
[08:46:43] <Andrew Sullivan> whose head hurts
[08:46:54] <Andrew Sullivan> working out scenario at mic
[08:48:09] <Andrew Sullivan> working out storage implications; worked
[08:48:14] <stpeter> :-)
[08:48:56] nemo joins the room
[08:49:07] <Andrew Sullivan> Joe Hildebrand: different from xmpp
[08:49:17] Jonathan Lennox leaves the room
[08:49:25] <Andrew Sullivan> servers do prep selves when used as address info
[08:49:56] <Andrew Sullivan> PSA: capture in questions: "What are you doing this for?"
[08:50:13] <Andrew Sullivan> Joe: "are you looking this up, indexing it, full text, &c &c"
[08:50:43] Florian Zeitz joins the room
[08:51:44] <Andrew Sullivan> chairs ask to consider whether saslprepbis doc is compatible.
[08:51:51] resnick leaves the room
[08:52:37] <Andrew Sullivan> PSA: now, individ. draft.  Is to become WG draft?
[08:52:49] <Andrew Sullivan> asking path forward
[08:52:50] resnick joins the room
[08:53:35] <Andrew Sullivan> Yutaka Ôiwa: better here
[08:54:11] <Andrew Sullivan> Pete Resnick (as AD):
[08:54:31] <Andrew Sullivan> technically, charter says may do replacements of old profiles
[08:54:33] john.levine joins the room
[08:54:42] <Andrew Sullivan> already violated once
[08:54:48] <Andrew Sullivan> no objection to discussion
[08:55:04] <Andrew Sullivan> chartered item?
[08:55:25] <Andrew Sullivan> likely going to need  a precis directorate
[08:56:53] <Andrew Sullivan> suggest to leave off the milestaones for now
[08:58:07] <Andrew Sullivan> PSA at mic
[08:58:32] <Andrew Sullivan> Marc Blanchet: replacement of stringprep profiles is a long list
[08:58:50] <Andrew Sullivan> need to have enough exercise to be confident
[08:59:28] <Andrew Sullivan> Pete Resnick: may become long lived WG.  sounding not too keen
[08:59:28] <Andrew Sullivan> may look more like DISPATCH
[08:59:39] <Andrew Sullivan> it'll be a special case WG/Directorate
[09:01:02] <Andrew Sullivan> PSA suggests a support group instead of a working group
[09:01:18] <Andrew Sullivan> Joe Hildebrand seemed to be +1ing
[09:01:22] <Andrew Sullivan> winding up meeting
[09:01:28] <stpeter> thank you to the scribes!
[09:01:59] resnick leaves the room
[09:02:02] Andrew Sullivan leaves the room
[09:02:14] nemo leaves the room
[09:02:30] Florian Zeitz leaves the room: offline
[09:02:45] john.levine leaves the room
[09:03:26] yone leaves the room
[09:10:49] stpeter leaves the room
[09:15:16] m&m joins the room
[09:17:56] Jonathan Lennox joins the room
[09:18:03] Jonathan Lennox leaves the room
[09:20:26] m&m leaves the room: Disconnected: connection closed
[09:21:22] stpeter joins the room
[09:22:15] stpeter leaves the room
[10:30:25] m&m joins the room
[10:33:45] m&m leaves the room: Disconnected: connection closed
[10:53:50] hildjj joins the room
[11:07:22] hildjj leaves the room
[11:07:25] hildjj joins the room
[11:12:28] m&m joins the room
[11:12:42] m&m leaves the room: Disconnected: connection closed
[11:12:49] m&m joins the room
[11:17:52] hildjj leaves the room
[13:54:34] m&m leaves the room: Disconnected: connection closed
[13:54:34] m&m joins the room
[13:54:44] m&m leaves the room: Disconnected: connection closed
[13:54:56] m&m joins the room
[14:17:10] m&m leaves the room
[16:00:48] hildjj joins the room
[16:17:18] hildjj leaves the room