[09:35:48] --- Balazs has joined
[09:39:00] --- Balazs has left
[09:39:01] --- Balazs has joined
[09:46:51] --- Balazs has left
[10:15:53] --- Balazs has joined
[10:34:17] --- Balazs has left
[12:20:25] --- Michael Kirkham has joined
[12:28:57] --- rstory has joined
[12:29:24] --- Balazs has joined
[12:45:07] --- bert has joined
[12:47:22] --- simon has joined
[12:53:50] <bert> are any of you remote participants?
[12:53:51] <simon> Discussion about the number of namespaces
[12:53:56] * Michael Kirkham is
[12:54:04] <simon> Michael, do you have MP3 feed?
[12:54:53] <Michael Kirkham> yes.
[12:55:00] <Michael Kirkham> (do now, didn't notice what time it was :)
[12:57:45] --- simon has left
[12:58:26] --- simon has joined
[12:58:41] --- GuyD has joined
[12:58:55] <simon> Try this for the MP3 stream: http://videolab.uoregon.edu/events/ietf/ietf688.m3u
[12:59:15] --- ric@jabber.org.au has joined
[13:01:29] --- GuyD has left
[13:02:15] --- m_ersue has joined
[13:07:20] <simon> Show of hands shows no clear preference for one vs. three additional namespaces
[13:10:01] <simon> Dave H. and Lada Lhotka argue for the simplicity of a single (additional) namespace in this case
[13:10:20] --- ric@jabber.org.au has left
[13:10:38] --- atarashi has joined
[13:10:44] <Michael Kirkham> debate referring to namespace URIs or URIs in general?
[13:11:02] <simon> Namespaces
[13:11:48] <Michael Kirkham> i'd concur then. they don't even have to actually point to anything existing.
[13:17:08] <simon> Current discussion: camelCase versus dashed-iden-tifiers
[13:21:52] <simon> No consensus either way
[13:27:10] <simon> Current discussion: Does the single notification defined here <replayCompleteNotification> really require an "eventClass" attribute?
[13:27:42] <simon> Andy claims this is unwarrantedly constraining data modeling for notifications (which we're not supposed to be doing in this document)
[13:29:17] <simon> ...and "numberEventsReplayed" attribute - Sharon claims this is useful for detecting that notifications have been dropped, e.g. due to throttling
[13:30:00] <simon> [Correction - eventClass and numberEventsReplayed are elements, not attributes, under replayCompleteNotification]
[13:34:30] <simon> Rough (almost complete in fact) consensus to remove the elements under numberEventsReplayed.
[13:35:57] <simon> Discussion: granularity of errors in response to <create-subscription> method
[13:36:35] <simon> Andy argues for avoiding to define new error-tags - would require revving the protocol document
[13:36:49] <simon> Sharon agrees, nobody complains
[13:38:00] <simon> Andy suggests to move the section about replay closer to the description of <create-subscription>
[13:38:37] <simon> Sharon explains that this is an artifact from replay being an optional feature (capability) in previous versions. Could be moved.
[13:42:05] <Balazs> put specific error message into app-error-tag. I hate when I have to guess what the real error is based on some generic error message.
[13:49:00] --- dlpartain has joined
[13:50:16] <simon> Phil Shafer on <create-subscription>
[13:50:30] <simon> Claims "create-subscription" is a bad name for what this is really doing.
[13:52:34] --- m_ersue has left: Replaced by new connection
[13:52:39] --- m_ersue has joined
[13:53:34] <simon> But apparently we had already discussed this before and decided to keep "create-subscription"
[13:55:00] <simon> Andy raises issue of <notification> structure - thinks the <data> element is not necessary.
[13:55:29] <simon> Sharon agrees that it might not be necessary, if we are sure we never want to define headers (as elements)
[13:58:04] <simon> Does the use of dateTime (without restrictions) create problems with timezones?
[14:01:01] <simon> Discussion on whether all imports are really necessary in the XSD schemas
[14:04:37] <simon> Discussion about namespaces - current thinking is to avoid version numbers in the namespace names, but rev the locations
[14:05:28] <simon> Sharon suggests: remove the version number for the content-related namespaces (for consistency with RFC4741), keep it for the protocol-related ones
[14:05:40] <simon> *ouch* was that the inverse
[14:05:42] <simon> ?
[14:09:05] <simon> Sharon will present her slides about notification transport
[14:10:01] <simon> No, sorry, Sharon will FIRST present her slides about Netconf monitoring
[14:10:15] <simon> (sorry the slides aren't online, they will be soon AFTER her talk.)
[14:12:00] <simon> Discussion whether monitoring is in scope
[14:12:30] <simon> Sharon claims that it would behoove us to define monitoring capabilities for our protocols
[14:13:03] <simon> Dan thinks the scope of the WG ends with the current notification work.
[14:16:39] <simon> Andy: open question whether we need to work on transport-specific text about notifications.
[14:17:32] <simon> Sharon thinks the ssh mapping must be updated because it only talks about <rpc> currently
[14:17:52] <simon> Andy doubts that the SOAP mapping can handle notifications
[14:18:18] <simon> Sharon argues that Web Services have a notification features
[14:20:03] <simon> Process discussion on whether to publish a "diff" spec, or revise the existing documents (RFC4742 etc.)
[14:21:00] <simon> Possible to publish as errata?
[14:21:33] <simon> Dan: errata not suitable for adding functionality
[14:28:07] <simon> Hideki presenting on his data modeling draft
[14:28:58] <simon> Diff-Serv configuration
[14:29:12] <simon> draft-okita-ngo-diffservdatamodel-00.txt
[14:29:37] --- Balazs has left
[14:31:03] <simon> based on the RFC3289 informal model of a diff-serv device
[14:31:17] <simon> this draft is an XML-based variant of RFC3290 (diffserv MIB)
[14:34:57] --- dlpartain has left
[14:40:29] <simon> Andy likes the fact that the schema has a distinct top-level node (<diffserv>)
[14:43:15] <simon> Disagreement on the interaction of filters vs. namespaces
[14:51:49] <simon> Discussion of schema languages
[14:56:27] --- m_ersue has left
[15:01:17] --- lengyel has joined
[15:06:46] --- bert has left
[15:07:36] --- lengyel has left
[15:09:57] --- Michael Kirkham has left
[15:09:57] --- atarashi has left
[15:10:04] --- lengyel has joined
[15:10:10] --- lengyel has left
[15:23:49] --- simon has left
[15:29:31] --- rstory has left: Logged out
[19:08:28] --- dlpartain has joined
[19:08:35] --- dlpartain has left