Minutes for NETCONF at interim-2014-netconf-1
minutes-interim-2014-netconf-1-1

Meeting start: 8 Sept 2014, 17:11 UTC
Meeting end: 8 Sept 2014, 19:05 UTC

Chair: Bert Wijnen
Attendees: Juergen Schoenwaelder
Ladislav Lhotka
Martin Bjorklund
Kent Watsen
Andy Bierman
Michael Scharf
Mahesh Jethanandani
Alberto Gonzalez Prieto
Benoit CLaise (AD)

Regrets: Mehmet Ersue (co-chair, vacation)

Hide quoted text
> Dear WG participants, we had our first virtual interim meeting yesterday.
> here are the draft minutes (thanks to Kent for taking notes, amended/merged
> with my own notes):
>
> The meetings run from 8 Sept - 3 Nov 2014. Biweekly on Mondays from 7pm-9pm
> CET (Berlin, Amsterdam, Paris). That is 17:00-19:00 UTC. They are webex based
> meetings. We have decided last night to record the meetings. But the WG chair(s)
> will ensure at the start of every meeting that everyone is aware and that
> nobody objects. The recording of yesterdays meeting is here:
>
> Streaming recording link:
> https://ietf.webex.com/ietf/ldr.php?RCID=2fd2a151d4ef7ebe95f310880f92fdad
> Download recording link:
> https://ietf.webex.com/ietf/lsr.php?RCID=b57f37cff1571be1061b1963a16bd472
>
> - The main objective of the call is to help us (WG) move faster on our
> chartered work items. RESTCONF is a big piece of that. Whatever we want
> to decide must be verified on the WG mailing list. Of course some discussion
> end up in changes to a document, which will always be presented to the WG
> mailing list and which will eventually be Last Called, so everyone has a say.
>
> - We had 10 participants on the call
>
> - Kent presented GitHub, which is used to track issues and which also
> is used by various editors to work together on the documents
> * DECISION: We will NOT forward all changes in GITHUB to the WG
> mailing list.
> * Pull-requests for editorial changes are OK,
> * but more substantial issues (in fact all issues) should be discussed
> on our NETCONF WG mailing list. We can use github to keep track of the
> status and to record summaries, proposed solutions and the final
> resolution.
>
> - Andy presented RESTCONF issues
> see: https://github.com/netconf-wg/restconf/issues
> We discussed them in reverse order (will try to get them properly
> discussed in order next time). I (Bert) have added all issues,
> also those that were created new as a result o the virtual meeting and
> those that were not discussed. If you want to comment or voice an
> opinion on any issue (that would be VERY GOOD and HELPFUL), then pls
> set the subject line to the issue number plus issue title.
>
> * issue # 10: Naming convention for Query Parameters
> o opened by Bert as a result of our first virual meeting
> o split off from issue # 6
> * issue # 9: Define how authentication is performed
> o opened by Andy as a result of our first virtual meeting
> * issue #7: Mandatory-to-implement encoding
> o WG chairs did a poll on the WG list for opinions:
> see: http://www.ietf.org/mail-archive/web/netconf/current/msg09218.html
> It runs till Sept 18, so those who have not expressed an opinion yet
> are encouraged to do so.
> o For now, "Optional" encoding (choice d) got more votes than "XML mandatory"
> (choice a).
> o Martin mentions technical motivations were expressed with choice 'a'
> o Bert reminded that STANDARDIZATION means making choices and making
interoperability
> easy. From that point of view, choice a seems more appropriate.
> o ACTION: Bert will send reminder to collect more opinions (done)
> * issue #6: how to identify query parameters supported by the server
> o No prefix for ietf-defined (this is naming, see new issue #10)
> o We basically have 2 issues, one is Naming,the other is discovery of which
> parameters are supported by the server. So issue #6 stays as is,
> issue #10 opened by Bert for Naming conventions
> o Capability encoding in the air: Lada is thinking a json object
> o Proposal to collapse into a single URI string
> o Opinions of other WG participants are welcome! Please speak up
> * issue #5: protocol capability URIs
> o was not discussed on this call
> o Toronto consensus: Add protocol capabilities somehow
> o so we need to see new text in a revision of the draft
> *issue #4: Defaults handling
> o Mapping to RESTCONF straight forward
> o the server MUST report what basic-mode it supports
> o Question: Should full support of with-defaults RFC be added to base RESTCONF?
> Please speak up and express opinions?
> * issue #3: add collection resource
> o Put collections media-type and query-parameters into main draft, but?
> o There seems to be consensus to put collections resource in another document.
> o Need to deal with GET on a list with no keys given:
> GET /restconf/data/interfaces/interface
> --> Maybe this will be an error unless collection resource requested
> o Need text proposals on mailing list.
> * issue #2 Netconf state monitoring
> o Need to create an issue for authentication
> (done by Andy, that is issue #9
> o Do we need to have a 6022-bis document or can RESTCONF just update the text
> in order to support non-NETCONF sessions.
> o An idea is to let this doc "update" RFC 6022
> * issue #1: Select parameter
> o What syntax should be used for the "select" query parameter? The
> current choices are "XPath" and "path-expr". Perhaps an
> additional parameter to identify the select string format is
> needed to allow extensibility?
> o Martin's proposal stands...issue seems closed
> o WG participants, please speak up if you cannot live with
> Martin's proposal, see
> https://github.com/netconf-wg/restconf/issues/1
>
> - A question was raised by Mahesh, regarding CLI type uses as ping and traceroute.
> He was asked to post the problem statement to the WG list, which he has done
now
> see: http://www.ietf.org/mail-archive/web/netconf/current/msg09254.html
>
> - Next time we will discuss restconf issues again, but also: call-home.
> It seems like a good idea to then have the accompanying documents updated as
well,
> so:
> o ACTION: Juergen should update 5539bis
> 0 ACTION: Kent to submit server-model
>
> Bert Wijnen, with special thanks to Kent and Andy.