IETF
scim@jabber.ietf.org
Wednesday, July 23, 2014< ^ >
stpeter has set the subject to: SCIM WG | http://tools.ietf.org/wg/scim/agenda?item=agenda-85-scim.html
Room Configuration
Room Occupants

GMT+0
[16:50:13] Slide Feed joins the room
[16:50:26] Room Feed joins the room
[16:52:28] Paolo Saviano joins the room
[16:54:05] ראש תחום joins the room
[16:58:28] Erik Wahlstrom joins the room
[16:58:29] frodek joins the room
[16:59:43] Erik Wahlstrom leaves the room
[16:59:46] Erik Wahlstrom joins the room
[17:00:32] <Erik Wahlstrom> hmm... no sound in meetecho. yet at least.
[17:01:52] Nik Tomkinson joins the room
[17:02:53] Erik Wahlstrom leaves the room
[17:02:55] Erik Wahlstrom joins the room
[17:03:19] Erik Wahlstrom leaves the room
[17:03:34] Erik Wahlstrom joins the room
[17:04:11] Erik Wahlstrom leaves the room
[17:04:13] Erik Wahlstrom joins the room
[17:04:36] wmills_92105 joins the room
[17:04:50] Kelly Grizzle joins the room
[17:05:07] <ראש תחום> How about now?
[17:05:28] <Room Feed> is audio ok?
[17:05:30] <ראש תחום> If you want to be spoken in the room, prefix your comment with "MIC", please.
[17:05:41] <Erik Wahlstrom> not yet. no audio. i will try to restart browser.
[17:05:45] Erik Wahlstrom leaves the room
[17:05:59] <Nik Tomkinson> I don't hear any audio either
[17:06:19] Erik Wahlstrom joins the room
[17:06:26] <ראש תחום> Meetecho guy just bolted from the room.
[17:06:35] <ראש תחום> I presume that means he's going to try to fix something.
[17:06:51] resnick joins the room
[17:07:12] <Erik Wahlstrom> Cool. Hope any of you guys know sign language :)
[17:07:15] Nik Tomkinson leaves the room
[17:07:59] <Erik Wahlstrom> Audio Streaming worked.
[17:10:46] frodek joins the room
[17:11:20] Nik Tomkinson joins the room
[17:13:01] <Erik Wahlstrom> meetecho sound just stared working again :)
[17:13:06] frodek joins the room
[17:13:35] <Nik Tomkinson> For me too - thanks - I assume no Meetecho slides?
[17:13:58] <Erik Wahlstrom> i've got slides :) so looks like everything is up and running ok from my side.
[17:15:08] <wmills_92105> blue! let's paint it blue!
[17:15:41] wmills_92105 leaves the room
[17:15:45] <resnick> Oh, stop it.
[17:15:47] <resnick> It's obvious it needs to be red.
[17:16:19] <ראש תחום> I'm a purple guy, mese'f.
[17:16:24] Bill Mills joins the room
[17:16:55] Nik Tomkinson leaves the room
[17:17:14] Simon Romano joins the room
[17:17:20] <Bill Mills> anyone have the meetcho link?
[17:17:45] <Erik Wahlstrom> http://www.meetecho.com/ietf90/scim
[17:17:51] Simon Romano leaves the room
[17:17:51] <Bill Mills> tx
[17:18:00] <Erik Wahlstrom> np
[17:18:16] Nik Tomkinson joins the room
[17:18:51] Paolo Saviano leaves the room
[17:19:22] Paolo Saviano joins the room
[17:19:43] <Bill Mills> MIC:  let's make sure we are clear on the difference between SCIM client/peer authentication and end user auth where SCIM is the data sink.  We're mixing them a bit.
[17:21:17] Nik Tomkinson leaves the room
[17:21:39] Bill Mills_1 joins the room
[17:23:47] <Paolo Saviano> is audio working in meetecho?
[17:24:57] <Erik Wahlstrom> MIC: my comment around the override default attributes on server that is that the server can decide what it does with all request. it can ignore clients requests. So how about something like "clients may try, and server can listen if it wants to".
[17:25:09] Paolo Saviano leaves the room
[17:25:31] <Erik Wahlstrom> yes, audio works.
[17:27:20] frodek leaves the room
[17:28:20] frodek leaves the room
[17:31:56] <Erik Wahlstrom> mmmm with that provision :)
[17:35:34] <Bill Mills> yes, bigger font
[17:35:56] <Kelly Grizzle> better bil?
[17:38:16] Paolo Saviano joins the room
[17:38:31] <Erik Wahlstrom> mic. does immutable and errors don't bleed some info?
[17:38:58] <Erik Wahlstrom> thinking about passwords.
[17:40:46] <Erik Wahlstrom> ooo maybe it's writeOnly. ignore my MIC request.
[17:40:53] Paolo Saviano leaves the room
[17:45:17] <Bill Mills> +1 for "the behavior shoudl all be the same on all methods". I also like the way we have mutability right now
[17:50:20] frodek leaves the room
[17:52:51] Paolo Saviano joins the room
[17:53:13] Paolo Saviano leaves the room
[17:56:35] <Erik Wahlstrom> mmm for SHOULD.
[17:57:00] Tobia Castaldi joins the room
[17:58:24] Tobia Castaldi leaves the room
[17:58:59] Tob Cast joins the room
[18:06:59] <Bill Mills> MIC: this isn't going a good way
[18:07:38] <Bill Mills> MIC: because a BULK might update the same user multiple times under this and the server is expected to get the ordering right
[18:09:29] <Erik Wahlstrom> MIC: Then the clients order must always be followed by SP. And the use of bulkID requires a user to be added before it's added to a group. Client take must take notice of that and construct the request in the right order.
[18:10:10] <Bill Mills> agreed that DWIM is bound for failure
[18:14:53] <Bill Mills> MIC: Super huge transactions are not HTTP friendly.  A bulk of 10k updates or more may well cause you unexpected problems
[18:15:52] <Erik Wahlstrom> MIC: we did a bulk inteorp between unbound and scimproxy at IETF meeting in Paris,
[18:16:04] Paolo Saviano joins the room
[18:16:43] <Erik Wahlstrom> no more
[18:17:05] <ראש תחום> I misread the comma, then.  :-D
[18:17:10] <Erik Wahlstrom> hehe
[18:18:47] <Erik Wahlstrom> +1 for stay in spec and not doing async
[18:19:02] ראש תחום joins the room
[18:21:51] <Bill Mills> the error case is where the client isn't careful and takes an update for the same user that came in while the bulk update was being put together.  Yes, it's a client being stupid.
[18:21:51] ראש תחום leaves the room
[18:22:21] <Bill Mills> no problem with leaving it in the current spec and extending it later
[18:22:24] Test User joins the room
[18:22:39] Test User leaves the room
[18:24:52] <Erik Wahlstrom> mmm for keep bulk in core spec.
[18:27:17] <Erik Wahlstrom> mic: Phil, I can help out fixing bulk.
[18:27:24] Ian Glazer joins the room
[18:27:53] <Kelly Grizzle> thx erik!
[18:37:49] <resnick> Why are they using URIs? Just for uniqueness, or are these URIs being resolved to something?
[18:38:19] <Bill Mills> I don't think they are resolving to anything
[18:38:26] <ראש תחום> They're using URIs as names.
[18:38:51] <ראש תחום> They're doing it to make it easy for providers to define their own — they can just use a URI with their own domain in the authority.
[18:39:00] <Bill Mills> mmmmmmmmm for no freaking clue
[18:42:29] <resnick> :-/
[18:46:20] Tob Cast leaves the room
[18:48:22] <ראש תחום> http://www.iana.org/assignments/params <http://www.iana.org/assignments/params/params.xhtml#urn-subnamespaces>
[18:48:29] <ראש תחום> IETF URN Sub-namespace for Registered Protocol Parameter Identifiers
[18:48:51] <ראש תחום> URNs have the form "urn:ietf:params:scim:<whatever-you-want>"
[18:49:16] <ראש תחום> The document would register "scim" as a new registered parameter identifier.
[18:51:25] Paolo Saviano leaves the room
[18:56:50] <Erik Wahlstrom> mic: interop: agree.. can try to figure out some resources to improve it.
[18:58:32] <ראש תחום> (I'm logging off jabber)
[18:58:36] ראש תחום leaves the room
[18:59:11] Paolo Saviano joins the room
[19:01:06] Ian Glazer leaves the room
[19:01:36] Bill Mills leaves the room
[19:01:36] <Erik Wahlstrom> Thanks! Great progress today.
[19:01:46] Slide Feed leaves the room
[19:02:16] Room Feed leaves the room
[19:02:31] resnick leaves the room
[19:02:44] Bill Mills_1 leaves the room
[19:02:49] Erik Wahlstrom leaves the room
[19:02:49] Kelly Grizzle leaves the room
[19:03:14] Paolo Saviano leaves the room
Powered by ejabberd Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!