View Issue Details

IDProjectCategoryView StatusLast Update
000450110000-014: PubSubSpecpublic2019-05-21 17:13
ReporterZbynek Zahradnik Assigned ToMatthias Damm  
PrioritynormalSeverityminorReproducibilityalways
Status closedResolutionfixed 
Summary0004501: PublisherId in UADP NetworkMessage - clarify equality comparison
Description

In Table 73, ExtendedFlags1, 5 different "PublisherId Types" are defined (Byte, UInt16, UInt32, UInt64, String).

The specification does not make it clear (for numerical PublisherIds) whether PublisherId-s with the same value but different type should be considered equal. For example, is NetworkMessage with PublisherId type "Byte" and value 123 considered to be from the same publisher as NetworkMessage with PublisherId type "UInt64" and value 123? I think the answer is Yes, but it would be better if clearly stated somewhere. (I know that PublisherId is defined elsewhere as being either String or UInteger, but that does not help in my view, unless we specify how different subtypes of UInteger should mutually compare for equality).

Or, reformulated: Do we, conceptually, have 2 types of PublisherIds, or 5 types? If we have 2 types, wouldn't it also be better to call the field in ExtendedFlags1 something like "PublisherId encoding", and not "PublisherId type"?

TagsNo tags attached.
Commit Version
Fix Due Date

Activities

Zbynek Zahradnik

2018-12-08 11:26

developer   ~0009679

Also see 0004505.

Zbynek Zahradnik

2019-01-11 11:14

developer   ~0009810

Also relates to semantics of 6.2.8.1 PublisherId (whether the concrete UInteger subtype used in PublisherId in the configuration needs to match the "encoding" of the UInteger on the wire - most likely not, but it'd better be specified).

Matthias Damm

2019-02-25 22:08

developer   ~0009933

Added following clarification to PublisherId field definition in NetworkMaessage:
PublisherIds are only equal if they have the same DataTypes and equal values

Added in OPC UA Part 14 - PubSub Draft 1.05.06 Specification.docx

Zbynek Zahradnik

2019-02-28 20:05

developer   ~0009990

The resolution is not my preferred one (and I think it has many disadvantages), but is acceptable. However, the reason I am reopening this issue is because the related issue, 0004505, has not been reviewed yet, and doing so may have influence on choosing the right solution for this issue. If after reviewing 0004505 we still think the proposed solution for 0004501 is the best one, I am OK with that then.

Matthias Damm

2019-03-05 21:27

developer   ~0010023

No additional change required. Related issue is resolved too.

Jim Luth

2019-05-21 17:13

administrator   ~0010251

Agreed to changes edited in telecon.

Issue History

Date Modified Username Field Change
2018-12-06 18:03 Zbynek Zahradnik New Issue
2018-12-08 11:26 Zbynek Zahradnik Note Added: 0009679
2019-01-11 11:14 Zbynek Zahradnik Note Added: 0009810
2019-02-25 22:08 Matthias Damm Assigned To => Matthias Damm
2019-02-25 22:08 Matthias Damm Status new => resolved
2019-02-25 22:08 Matthias Damm Resolution open => fixed
2019-02-25 22:08 Matthias Damm Note Added: 0009933
2019-02-28 20:05 Zbynek Zahradnik Status resolved => feedback
2019-02-28 20:05 Zbynek Zahradnik Resolution fixed => reopened
2019-02-28 20:05 Zbynek Zahradnik Note Added: 0009990
2019-03-05 21:27 Matthias Damm Status feedback => resolved
2019-03-05 21:27 Matthias Damm Resolution reopened => fixed
2019-03-05 21:27 Matthias Damm Note Added: 0010023
2019-05-21 17:13 Jim Luth Status resolved => closed
2019-05-21 17:13 Jim Luth Fixed in Version => 1.05
2019-05-21 17:13 Jim Luth Note Added: 0010251