Child pages
  • 2023-01-13 IISOMI Meeting notes
Skip to end of metadata
Go to start of metadata





  • Administrative
    • Call plan (schedule, leader)
  • Review IISOMI Action Items due/done (not discussed at this meeting)
  • YANG enhancement - Immutable flag
  • UML <target> to YANG augment statement mapping
  • Concrete vs abstract augmenting classes (not discussed at this meeting)
  • AOB

Discussion items





5 minAdmin: Call planAll

 Leader: Kam; (Not available: )

 Leader: Bernd; (Not available: )

 Leader: Scott; (Not available:)

 Leader: Kam; (Not available: )

 Leader: Cancelled

 Leader: Nigel; (Not available: )

 Leader: Scott; (Not available: Kam) (May adjust to 8:00 PST to accommodate Kam)

 Leader: ; (Not available: Kam)

 Leader: ; (Not available: Kam)

 Leader: ; (Not available: Kam, Scott)

Leader: ; (Not available: Scott)

0 minAction items due/doneAll

Not discussed at this meeting. Deferred to next meeting.

IISOMI Action Items done or past due

Action items done:

Due dates of the following action items updated:

25 min

YANG enhancement 


Scott brought up to our attention that in IETF, there is an individaul I-D "YANG Extension and Metadata annotation for Immutable Flag" 

  • Individual I-D:
  • There are usage examples in Appendix A of the I-D
  • Scott is asking for more usage examples from UML models
  • Kam mentioned the example of the granularityPeriod attribute of the CurrentData object class. 
    • See the extraction below from the G.8052.1 Amd.1 Data Dictionary
    • which is config-true at creation, and then it should not be changed after creation (i.e., immutable).
30 minUML <target> to YANG augment statement mappingAll

Question: To find out if the tool is inferring the augment path from the composition associations OR form the target property of the «Specify» association.

  • In the specify stereotype (target) of the uml model, the elements of the path of the target are separated by the delimitor "/".
  • Each element of the path has 3 components, namely ModelName:ClassName:MemberEndName.
  • The member end is an attribute of the class. ClassName is not necessary in YANG (according to Italo).

Answer: Scott found out that the tool infers from the target property of the «Specify» association.


So, not inferrring from the containment tree. Just by the target path, in each element it takes the first component (module name) and the third component (member end name) and then yangify. No intelligent at all.

However, the tool doesn't check whether the class (the middle component) is abstract or concrete.

1 min

Concrete vs abstract augmenting classes


Not discussed at this meeting. Deferred to next meeting.

Question statement: Should the UML Modelling Guidelines allow concrete specification classes like TAPI is using it?

Bernd offered to represent the 4 slides that he used on the last call (as several people on the call today were not present on that call).

Minutes added to Augmentation Issues in uml2yang discussion page.

Bernd noted that so far the guideline recommends only dealing with abstract augmenting class.

x minAOB


Action items