View Issue Details

Jump to NotesJump to History
IDProjectCategoryView StatusDate SubmittedLast Update
000800710000-005: Information ModelSpecpublic2022-05-24 14:452023-04-11 16:33
ReporterJouni Aro Assigned ToJeff Harding  
PrioritynormalSeveritymajorReproducibilityalways
Status closedResolutionfixed 
Product Version1.04 
Target Version1.05.03 RC1Fixed in Version1.05.03 RC1 
Summary0008007: Ambiguity in the definition of 3DOrientationType
Description

The RpyOrientationType was added to the RSL (Relative Spatial Location) spec to express that the Roll, Pitch and Yaw defined in the type shall correspond to the mathematical interpretation detailed in the Annex section of the spec. The parent type i.e. the 3DOrientationType (https://reference.opcfoundation.org/v104/Core/docs/Amendment11/7.26/) introduced in the base specification is more generic with respect to the mathematical interpretation. Although it defines A,B and C to correspond to Roll, Pitch and Yaw, it does not define the necessary mathematical details. The thinking has been that subclasses can define the precise meaning of these angles, but that would lead to an unsustainable situation.

We should not expect that various OPC UA Client applications can deal with any sensible way with different variations of 3DOrientationType without knowing the exact meaning of the angles.

Therefore, instead of enabling an ambiguous meaning of the angles in the first place, the definition of 3DOrientationType should be fixed to the preferred mathematical interpretation, as is currently defined in the RC version of the RSL Specification.

Whenever there are already existing types that define alternative definitions for the angles, new types should be derived from OrientationType with new angles A,B and C. This will enable the client applications to differentiate between the "standard" interpretation of Roll, Pitch and Yaw and the alternative interpretation. Since these A.B and C will be defined in another namespace, they will be different properties and no ambiguity will rise.

TagsNo tags attached.
Commit Version
Fix Due Date

Activities

Jim Luth

2022-06-14 15:48

administrator   ~0016840

Last edited: 2022-06-14 18:44

Agreed to:
Add a note to 3DOrientationType that it is flawed and will be deprecated soon.
Add text to OrientationType that this is the type that should be derived from and cite the example of RpyOrientationType in the RSL spec.

Jeff Harding

2022-06-20 14:21

developer   ~0016900

Added description to 7.26 that the 3DOrientationType will be deprecated and the OrientationType should be used to create types.

Jim Luth

2022-06-22 12:41

administrator   ~0016982

Agreed to changes edited in Munich F2F.

Jouni Aro

2022-06-28 15:44

reporter   ~0017066

I would like to revisit this, since 3DOrientation is declared with angles A, B and C, as rotations around X (roll), Y (pitch) and Z (yaw); 3DFrameType is defining Orientation with 3DOrientation and 3DOrientationType and therefore it would make sense, if we could just clarify that 3DOrientationType matches with 3DOrientation, which obviously has been the original idea, and is the only reasonable way to use them.

We need to additionally clarify if the mathematical transformations that are planned for the Relative Spatial Location spec could be added to the Part 5 to clarify that part as well, since fixing the rotations should also fix the transformation in practice.

Jim Luth

2022-08-02 15:32

administrator   ~0017208

Last edited: 2022-08-02 20:30

Discussed in UA Meeting:

Add deprecation warning to 3dFrameType.
Add deprecation warning to 3dFrame.
Remove the description for A,B and C in 3DOrientation.

Suprateek Banerjee

2022-09-26 05:11

reporter   ~0017843

3DFrame shouldn't be deprecated. The removal of the descriptions for A, B and C in 3DOrientation make the 3DFrame structure usable again.
RSL Specification is using the 3DFrame in its model.

Jeff Harding

2023-03-14 16:03

developer   ~0018864

Propose we do remove the descriptions for the 3 dimensions of the 3DOrientation DataType.
As suggested with the removal of the specific definitions of the 3 dimensions of the 3DOrientation DataType there is no need to deprecate 3DFrameType and 3DFrame.

Jeff Harding

2023-03-14 16:07

developer   ~0018865

Removed dimension descriptions of the 3DOrientation DataType.

Jim Luth

2023-04-11 16:33

administrator   ~0019161

Agreed to changes edited in Web meeting.

Issue History

Date Modified Username Field Change
2022-05-24 14:45 Jouni Aro New Issue
2022-06-14 14:58 Jouni Aro Description Updated
2022-06-14 15:48 Jim Luth Note Added: 0016840
2022-06-14 15:49 Jim Luth Target Version => 1.05.02 RC1
2022-06-14 15:50 Jim Luth Assigned To => Jeff Harding
2022-06-14 15:50 Jim Luth Status new => assigned
2022-06-14 18:44 Jim Luth Note Edited: 0016840
2022-06-20 14:21 Jeff Harding Status assigned => resolved
2022-06-20 14:21 Jeff Harding Resolution open => fixed
2022-06-20 14:21 Jeff Harding Fixed in Version => 1.05.02
2022-06-20 14:21 Jeff Harding Note Added: 0016900
2022-06-22 12:41 Jim Luth Status resolved => closed
2022-06-22 12:41 Jim Luth Note Added: 0016982
2022-06-28 15:44 Jouni Aro Status closed => feedback
2022-06-28 15:44 Jouni Aro Resolution fixed => reopened
2022-06-28 15:44 Jouni Aro Note Added: 0017066
2022-07-05 14:15 Jim Luth Target Version 1.05.02 RC1 => 1.05.03 RC1
2022-08-02 15:32 Jim Luth Note Added: 0017208
2022-08-02 15:32 Jim Luth Note Edited: 0017208
2022-08-02 15:33 Jim Luth Status feedback => assigned
2022-08-02 20:30 Jim Luth Note Edited: 0017208
2022-09-26 05:11 Suprateek Banerjee Note Added: 0017843
2023-03-14 16:03 Jeff Harding Note Added: 0018864
2023-03-14 16:07 Jeff Harding Status assigned => resolved
2023-03-14 16:07 Jeff Harding Resolution reopened => fixed
2023-03-14 16:07 Jeff Harding Fixed in Version 1.05.02 => 1.05.03
2023-03-14 16:07 Jeff Harding Note Added: 0018865
2023-04-11 16:33 Jim Luth Status resolved => closed
2023-04-11 16:33 Jim Luth Fixed in Version 1.05.03 => 1.05.03 RC1
2023-04-11 16:33 Jim Luth Note Added: 0019161