[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Draft Minutes 12th December 2012
Attendees:
Carlson, Mark Oracle Voting Member
Chapman, Martin Oracle Chair
Durand, Jacques Fujitsu Limited Voting Member
Heneveld, Alex Cloudsoft Corporation Limited Member
Janssen, Gershon Individual Voting Member
Johnston-Watt, Duncan Cloudsoft Corporation Limited Member
Karmarkar, Anish Oracle Voting Member
Kunze, Tobias Red Hat Voting Member
Luster, Eugene US Department of Defense (DoD) Member (gains voting after this meeting)
Malhotra, Ashok Oracle Voting Member
Miller, Rich Cloudsoft Corporation Limited Voting Member
Mischkinsky, Jeff Oracle Voting Member
Otto, Adrian Rackspace Hosting, Inc. Secretary
Pilz, Gilbert Oracle Voting Member
Song, Zhexuan Huawei Technologies Co., Ltd. Voting Member
Tupitza, Charles JumpSoft Voting Member
Yendluri, Prasad Software AG, Inc. Voting Member
Intro:
Jacques Durand assumes scribe duties.
Roll call. Meeting is quorate 14 of 18, 77%.
Agenda: add update from editors.
agenda adopted w/o
Minutes:
5th December 2012: https://www.oasis-open.org/apps/org/workgroup/camp/email/archives/201212/msg00026.html
MOTION: Gil to adopt minutes Dec 5th, Adrian 2nd. No discuss. passes UNAN
Administrivia:
Adrian confirms Rackspace will host the next F2F in San Antonio, TX, 22-25 Jan 2013
... please gives notice attendance to Adrian (name + email + title) by COB
Editors Update:
Anish: editors generated WD #1, available in WD folder - just formatted initial contribution in OASIS template.
... no change in content. No content received so far.
... always use the latest draft # for any comment.
... Editors look into JIRA for whatever issue is resolved, and moves the resolution to latest draft.
... rev history tells what issue is in.
New Issues:
https://tools.oasis-open.org/issues/browse/CAMP-35 Spec needs to clarify the constraints on the relationship between Assembly Templates and Application Component Templates
gil: based on implementation of #24. Relationship between Assemble template and app component template.
... can such app comp templ exist separately from assembly templates?
tobias: valid question. Q1: yes, Q2: yes, Q3: yes (need flexibility)
alex: tempted to say no
MOTION: (gil) accept to open issue 35, 2nd Anish. No discussion. No objection. Passed w/o
https://tools.oasis-open.org/issues/browse/CAMP-36 Use HTTP Patch and JSON Patch for updates to resources
Anish: PUT not good for partial updates. PATCH is the way.
... JSON PATCH format is an option
adrian: can we address this in #34?
anish: still relevant separately but we can link the 2 issues
adrian: the scheme to use PATCH should be shared across issues
MOTION (adrian) open Issue 36, anish 2nd. No discuss. No object. Passed w/o
Issue Discussions:
https://tools.oasis-open.org/issues/browse/CAMP-12 Clarify scope of Application Components
gil: proposal changes def of App Component in glossary appendix
anish: glossary normative by default
anish: no need for "keywords" to be normative
gil: diff between app comp and platform comp are more about management (user role)
anish: i'll create the links between issues 34 and 36
anish: current proposal is OK for now, another issue can take care of platform comp vs appl comp
chair: incremental improvement is OK
MOTION: (gil) resolve issue 12 with proposal in JIRA, 2nd adrian. No discuss. No objection. Passes w/o
http://tools.oasis-open.org/issues/browse/CAMP-15 Use message-id and profile rather than uri and namespace
adrian: about ID for error messages
anish: i have one suggestion for the proposal: s/A string containing a unique uuid or error number that references this particular message/A string containing unique identifier that identifies this particular message/
anish: another Q I have is: do we want say *globally* unique
anish: message ID not a URL. Just need be unique not necessarily globally.
Ashok Malhotra (Oracle): unique identifier is enough
adrian: no portability issue - only context is consumuer+provider
Adrian Otto (Rackspace): Use:
| message-id | String | 0..1 | A string containing a unique identifier that identifies this particular message |
Change to:
| message-id | String | 0..1 | A unique string that identifies this particular message |
MOTION: (adrian) to resolve issue 15 with proposal updated in preceding lines latest edit in the chat. 2nd (gil)
No discussion. No objection. Passes w/o
https://tools.oasis-open.org/issues/browse/CAMP-32 CAMP needs to say something about the serialization of null attribute values and empty arrays
anish: camp attributes <--> JSON properties
gil: serialization of optional attributes
... response to a GET should omit the absent attr
tobias: the def of such attributes (here description) should mention this optionally.
... questioning the rationale behind this issue. Would motion to close the issue w no action.
adrian: 2 schools of thought: strongly typed approach (XML based), and JSON style less typed. Important to align w the style. a JSON API should not be so strict on attr presence. Conform to the style.
gil: interpretation problem. If attr has no value, be explicit about it.
Tobias Kunze: They key to this issue is that the platform implementation has no knowledge of the resource structure. As a result, it can't simply drop properties.
anish: @tobias do you mean client instead of platform impl in your comment above?
Tobias Kunze: I mean platform but applies for client as well
tobias: an impl cannot just drop an attribute.
... 3rd party providers of Platform components might ignore such attributes
chair: issue #32 needs more thoughts
Adrian: table discussion on 32. No objection
http://tools.oasis-open.org/issues/browse/CAMP-30 Parameter Scheme
adrian: create a template with a concept of variable in it (parameterized)
... such values are dynamically passed at creation time
... spares a PATCH
anish: more than just PATCH. Like this direction. Could also have another way to indicate (create a list of such parameterized attr)
anish: http://tools.ietf.org/html/rfc6570
gil: use case: creating multiple assemblies from a single template. Need to distinguish at creation time (can't just create clones then PATCH later)
... may need a default value if not provided.
martin: proposal needs more in case $value not provided
jacques: need to look at mutability indication in the more general case?
adrian: will elaborate on the issue , make link with #34, think about what happens when no value is provided and will look at RFC6570
AOB
chair: next meeting on Dec 19, Meeting on 26 dec cancelled, need to decide next week whether to have a call on 2nd Jan 2013.
Straggler roll: add jeff and rich
Meeting Adjourned.
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]