RE: [odata] Agenda for OData TC meeting on 2017-05-18

[17:04] Ralf Handl (SAP SE): 1.Roll call [8:00 am PT]

a.Self-registration link: https://www.oasis-open.org/apps/org/workgroup/odata/event.php?event_id=43978

[17:05] Ralf Handl (SAP SE): Dial-in information: https://www.oasis-open.org/apps/org/workgroup/odata/download.php/56760/latest

Online chat room: http://webconf.soaphub.org/conf/room/odatatc

Screen sharing: https://sap.emea.pgiconnect.com/OData-TC/

[17:06] Michael Pizzo: Link to minutes from last meeting should be: https://www.oasis-open.org/apps/org/workgroup/odata/download.php/60730/odata-meeting-174_on-20170511-minutes.html

[17:07] Ralf Handl (SAP SE): Voting Members: 5 of 11 (45%) (used for quorum calculation)

[17:09] Stefan Hagen: Voting Members: 6 of 11 (54%) (used for quorum calculation)

[17:09] Stefan Hagen: Sorry travelling

[17:09] Stefan Hagen: quorum should be reached

[17:09] Ralf Handl (SAP SE): Thanks!

[17:10] Ralf Handl (SAP SE): 2.Approve agenda [8:05 am PT]

[17:11] Ralf Handl (SAP SE): Gerald to share news on the Data Aggregation spec

[17:11] Ralf Handl (SAP SE): Mike: new issue

[17:11] Martin Zurmuehl (SAP SE): I'm in, sorry! Quorum reached.

[17:12] Ralf Handl (SAP SE): ODATA-1077 Include atomicityGroup in JSON Batch Response

[17:12] Ralf Handl (SAP SE): ODATA-1076 If-Match and If-None-Match: align wording with RFC7232

[17:13] Ralf Handl (SAP SE): Mike: start with these two issues before document walkthrough

[17:14] Ralf Handl (SAP SE): Geralds topic after #3

[17:14] Ralf Handl (SAP SE): Agenda is approved with these additions

[17:14] Ralf Handl (SAP SE): 3.Approve minutes from previous meeting(s) [8:10 am PT]

a.Minutes from May 11, 2017 TC meeting: https://www.oasis-open.org/committees/download.php/60691/odata-meeting-173_on-20170504-minutes.html

[17:14] Michael Pizzo: Correct link: https://www.oasis-open.org/apps/org/workgroup/odata/download.php/60730/odata-meeting-174_on-20170511-minutes.html

[17:15] Stefan Hagen: @Mike: Thanks.

[17:16] Ralf Handl (SAP SE): Minutes are approved with these changes

[17:16] Ralf Handl (SAP SE): New agenda item: Gerald on Data Aggregation

[17:20] Ralf Handl (SAP SE): 4.Review action items [Action item list: https://www.oasis-open.org/apps/org/workgroup/odata/members/action_items.php] [8:15am PT]

a.Action items due

i.None

[17:21] Ralf Handl (SAP SE): 5.V4.01 [8:20 am PT]

[17:21] Ralf Handl (SAP SE): b.Issues for V4.01_CSD02

[17:21] Michael Pizzo: Gerald introduced idea of new transformation for Data Aggregation.  Will send out a document that we can describe at a later meeting.

[17:21] Ralf Handl (SAP SE): ODATA-1076 If-Match and If-None-Match: align wording with RFC7232

[17:22] Ralf Handl (SAP SE): https://issues.oasis-open.org/browse/ODATA-1076

[17:23] Ralf Handl (SAP SE): ODATA-1076 is OPEN

[17:23] Michael Pizzo: from RFC7232:

2.3.2.  Comparison

   There are two entity-tag comparison functions, depending on whether

   or not the comparison context allows the use of weak validators:

   o  Strong comparison: two entity-tags are equivalent if both are not

      weak and their opaque-tags match character-by-character.

   o  Weak comparison: two entity-tags are equivalent if their

      opaque-tags match character-by-character, regardless of either or

      both being tagged as "weak".

Fielding & Reschke           Standards Track                   [Page 10]

RFC 7232              HTTP/1.1 Conditional Requests            June 2014

   The example below shows the results for a set of entity-tag pairs and

   both the weak and strong comparison function results:

   +--------+--------+-------------------+-----------------+

   | ETag 1 | ETag 2 | Strong Comparison | Weak Comparison |

   +--------+--------+-------------------+-----------------+

   | W/"1"  | W/"1"  | no match          | match           |

   | W/"1"  | W/"2"  | no match          | no match        |

   | W/"1"  | "1"    | no match          | match           |

   | "1"    | "1"    | match             | match           |

   +--------+--------+-------------------+-----------------+

[17:27] Michael Pizzo: From RFC7232:

An origin server MUST use the strong comparison function when

   comparing entity-tags for If-Match (Section 2.3.2), since the client

   intends this precondition to prevent the method from being applied if

   there have been any changes to the representation data.

[17:27] anonymous morphed into George Ericson(Dell)

[17:28] Michael Pizzo: Section 3.2 says use weak for If-None-Match:

 A recipient MUST use the weak comparison function when comparing

   entity-tags for If-None-Match (Section 2.3.2), since weak entity-tags

   can be used for cache validation even if there have been changes to

   the representation data.

[17:28] Martin Zurmuehl (SAP SE): from rfc 7232

[17:28] Martin Zurmuehl (SAP SE): An entity-tag can be either a weak or strong validator, with strong

   being the default.  If an origin server provides an entity-tag for a

   representation and the generation of that entity-tag does not satisfy

   all of the characteristics of a strong validator (Section 2.1), then

   the origin server MUST mark the entity-tag as weak by prefixing its

   opaque value with "W/" (case-sensitive).

[17:30] Michael Pizzo: Section 2.1:

2.1.  Weak versus Strong

   Validators come in two flavors: strong or weak.  Weak validators are

   easy to generate but are far less useful for comparisons.  Strong

   validators are ideal for comparisons but can be very difficult (and

   occasionally impossible) to generate efficiently.  Rather than impose

   that all forms of resource adhere to the same strength of validator,

   HTTP exposes the type of validator in use and imposes restrictions on

   when weak validators can be used as preconditions.

   A "strong validator" is representation metadata that changes value

   whenever a change occurs to the representation data that would be

   observable in the payload body of a 200 (OK) response to GET.

   A strong validator might change for reasons other than a change to

   the representation data, such as when a semantically significant part

   of the representation metadata is changed (e.g., Content-Type), but

   it is in the best interests of the origin server to only change the

   value when it is necessary to invalidate the stored responses held by

   remote caches and authoring tools.

   Cache entries might persist for arbitrarily long periods, regardless

   of expiration times.  Thus, a cache might attempt to validate an

   entry using a validator that it obtained in the distant past.  A

   strong validator is unique across all versions of all representations

   associated with a particular resource over time.  However, there is

   no implication of uniqueness across representations of different

   resources (i.e., the same strong validator might be in use for

   representations of multiple resources at the same time and does not

   imply that those representations are equivalent).

Fielding & Reschke           Standards Track                    [Page 5]

RFC 7232              HTTP/1.1 Conditional Requests            June 2014

   There are a variety of strong validators used in practice.  The best

   are based on strict revision control, wherein each change to a

   representation always results in a unique node name and revision

   identifier being assigned before the representation is made

   accessible to GET.  A collision-resistant hash function applied to

   the representation data is also sufficient if the data is available

   prior to the response header fields being sent and the digest does

   not need to be recalculated every time a validation request is

   received.  However, if a resource has distinct representations that

   differ only in their metadata, such as might occur with content

   negotiation over media types that happen to share the same data

   format, then the origin server needs to incorporate additional

   information in the validator to distinguish those representations.

   In contrast, a "weak validator" is representation metadata that might

   not change for every change to the representation data.  This

   weakness might be due to limitations in how the value is calculated,

   such as clock resolution, an inability to ensure uniqueness for all

   possible representations of the resource, or a desire of the resource

   owner to group representations by some self-determined set of

   equivalency rather than unique sequences of data.  An origin server

   SHOULD change a weak entity-tag whenever it considers prior

   representations to be unacceptable as a substitute for the current

   representation.  In other words, a weak entity-tag ought to change

   whenever the origin server wants caches to invalidate old responses.

   For example, the representation of a weather report that changes in

   content every second, based on dynamic measurements, might be grouped

   into sets of equivalent representations (from the origin server's

   perspective) with the same weak validator in order to allow cached

   representations to be valid for a reasonable period of time (perhaps

   adjusted dynamically based on server load or weather quality).

   Likewise, a representation's modification time, if defined with only

   one-second resolution, might be a weak validator if it is possible

   for the representation to be modified twice during a single second

   and retrieved between those modifications.

   Likewise, a validator is weak if it is shared by two or more

   representations of a given resource at the same time, unless those

   representations have identical representation data.  For example, if

   the origin server sends the same validator for a representation with

   a gzip content coding applied as it does for a representation with no

   content coding, then that validator is weak.  However, two

   simultaneous representations might share the same strong validator if

   they differ only in the representation metadata, such as when two

   different media types are available for the same representation data.

Fielding & Reschke           Standards Track                    [Page 6]

RFC 7232              HTTP/1.1 Conditional Requests            June 2014

   Strong validators are usable for all conditional requests, including

   cache validation, partial content ranges, and "lost update"

   avoidance.  Weak validators are only usable when the client does not

   require exact equality with previously obtained representation data,

   such as when validating a cache entry or limiting a web traversal to

   recent changes.

[17:40] Michael Pizzo: Potential issue: according to strict reading of RFC7232, it appears that returning the same etag when $select, $apply, or $expand is applied (which is what we would want) would imply that the etag was weak.

[17:43] Michael Pizzo: Ralf: because these are specified in the query string it's really a different URL...

[17:43] Ralf Handl (SAP SE): ETags and $expand work different than HTTP ETags in general: each expanded entity has its own ETag in the @odata.etag annotation

[17:50] Ralf Handl (SAP SE): So in a response representing a single entity with expanded entities, each entity could have a @odata.etag, which is the strong ETag, and the overall response in the ETag header could have either a weak ETag computed from the "main" entity, or a strong ETag computed collision-free from all involved @odata.etags

[17:50] Ralf Handl (SAP SE): Similar for collection-valued responses

[17:59] Michael Pizzo: https://tools.ietf.org/html/rfc7232

[17:59] Ralf Handl (SAP SE): Will prepare proposal based on this discussion

[18:00] Ralf Handl (SAP SE): Next agenda item: ODATA-1077 Include atomicityGroup in JSON Batch Response

[18:00] Ralf Handl (SAP SE): https://issues.oasis-open.org/browse/ODATA-1077

[18:04] Ralf Handl (SAP SE): ODATA-1077 is OPEN

[18:05] Mark Biamonte (Progress): I move that OData-1077 be resolved as proposed

[18:05] Michael Pizzo: I second

[18:05] Ralf Handl (SAP SE): ODATA-1077 is RESOLVED as proposed

[18:05] Ralf Handl (SAP SE): a.Document walkthrough

[18:06] Ralf Handl (SAP SE): ii.Part 2: URL Conventions  https://www.oasis-open.org/committees/download.php/60367/odata-v4.01-wd02-part2-url-conventions-2017-03-24.docx

1.Continue with section 5.1.1.5, String Functions

[18:06] Ralf Handl (SAP SE): Five minutes break until 18:11

[18:14] Ralf Handl (SAP SE): Mike: [1,1,2] subsetof [1,2]?

[18:15] Ralf Handl (SAP SE): [3,1,4] subsetof [4,3,1]?

[18:16] Michael Pizzo: Does it depend on whether the collection is ordered?

[18:18] Michael Pizzo: I would think that, for ordered or non-ordered, the element of the left operand must occur at least as many times in the right operand.

[18:24] Michael Pizzo: For non-ordered collections, [3,1] subsetof [4,1,3] should be true.

[18:24] Michael Pizzo: Question is for ordered collections: is [3,1] subsetof [4,1,3]

[18:32] Michael Pizzo: Reopened ODATA-1002, assigned to Hubert with the following comment:

[18:32] Michael Pizzo: In reviewing the application, two questions came up:

1) is [1,1,2] subsetof [1,2] ?  (do elements of the left operand have to occur at least as many times in the right operand)

2) is [3,1] subsetof [4,1,3]? (does order matter)

For #2, if the collections are unordered, then [3,1] subsetof [4,1,3] should be true, but what about for ordered collections?

[18:38] Ralf Handl (SAP SE)5: 5.1.1.10 Lambda Operators

[18:53] Michael Pizzo: Created new issue  ODATA-1078 - What happens if lambda variable name matches a (complex) property name?

[18:53] Ralf Handl (SAP SE)5: ODATA-1078 is OPEN

[18:54] Michael Pizzo: https://issues.oasis-open.org/browse/ODATA-1078

[18:55] Michael Pizzo: Proposal:

1) It is not required

2) In case of conflict, lambda variable name wins (you shouldn't pick a lambda variable name that conflicts...)

[18:56] Michael Pizzo: Also; client can always use $it to "back out of" shadowing...

[18:56] Mark Biamonte (Progress): I move to resolve OData-1078 as proposed

[18:57] Michael Pizzo: I second

[18:57] Ralf Handl (SAP SE)5: ODATA-1078 is RESOLVED as proposed

[19:55] Ralf Handl (SAP SE)5: 6.Next meeting [9:50 am PT]

[19:57] Ralf Handl (SAP SE)5: Mike: skip next meeting, use time to pre-read CSDL documents (both of them)

[19:57] Ralf Handl (SAP SE)5: Mark agrees

[19:59] Ralf Handl (SAP SE)5: a.Thursday May 25, 2017: no meeting

b.Thursday June 01, 2017 during 7-10 am PDT (16:00-19:00 CEST)

c.      Thursday June 06, 2017 during 8-11 am PDT (17:00-20:00 CEST)

[19:59] Ralf Handl (SAP SE)5: c: June 08, not 06

[19:59] Ralf Handl (SAP SE)5: Mark on vacation on June 08

[19:59] Ralf Handl (SAP SE)5: 7.AOB and wrap up [9:55 am PT]

[20:00] Ralf Handl (SAP SE)5: Meeting is adjourned

From: odata@lists.oasis-open.org [mailto:odata@lists.oasis-open.org] On Behalf Of Handl, Ralf
Sent: Montag, 15. Mai 2017 09:59
To: odata@lists.oasis-open.org
Subject: [odata] Agenda for OData TC meeting on 2017-05-18

Here [1] is a draft agenda for the OData TC (Technical Committee) meeting scheduled on Thursday May 18, 2017 during 8-11 am PDT (17:00-20:00 CEST). For additional information, such as dial-in details and chat room, refer to [2]. For TC timeline, see [3]. Feel free to suggest additions or modifications.

Thanks.

[1] Agenda

1. Roll call [8:00 am PT]

a. Self-registration link: https://www.oasis-open.org/apps/org/workgroup/odata/event.php?event_id=43978

2. Approve agenda [8:05 am PT]

3. Approve minutes from previous meeting(s) [8:10 am PT]

a. Minutes from May 11, 2017 TC meeting: https://www.oasis-open.org/committees/download.php/60691/odata-meeting-173_on-20170504-minutes.html

4. Review action items [Action item list: https://www.oasis-open.org/apps/org/workgroup/odata/members/action_items.php] [8:15am PT]

a. Action items due

i. None

5. V4.01 [8:20 am PT]

a. Document walkthrough