The following errata report has been rejected for RFC7950,
"The YANG 1.1 Data Modeling Language"

--------------------------------------
You may review the report below and at:
https://errata.rfc-editor.org/eid8784/

--------------------------------------

Status: Rejected
Type: Technical
Reported by: Maria Matejka <[email protected]>
Date Reported: Feb. 25, 2026, 8 p.m.
Rejected by: Mahesh Jethanandani (IESG)

Section 9.12.4 says:

Original Text
-------------
If the second member type in the union had been of type "string"
   instead of an enumeration, the current value would have matched, and
   the resulting configuration would have been valid.

Corrected Text
--------------
If the second member type in the union had been of type "string"
   instead of an enumeration, the current value would have matched, and
   the resulting configuration would have been valid. However, the leaf's
   value must then be reinterpreted if a matching instance is added later.

Notes
-----
>From the example wording, it looks like that the event-to-be-watched is only 
>target instance removal for a successfully resolved leafref. Yet from the 
>specification, one can infer that one indeed needs to watch also for creation 
>of a target instance for unsuccessfully resolved possible leafref, and 
>therefore the example is misleading.

Verifier's Note: 

The concern expressed as part of this Errata does not survive scrutiny of 
YANG's validation model. Per RFC 7950, validation is holistic: it is triggered 
by any change to the intended configuration and evaluates the entire datastore 
without reference to its prior state. Consequently, when a leafref target is 
created, the same validation pass that validates the new instance also 
re-evaluates all outstanding leafref constraints, including those nested within 
union types. No implementation-level asymmetry exists; no special monitoring of 
target-creation events is required beyond normal change-triggered validation.

The existing text in Section 9.12.4 correctly describes the behavior 
illustrated by its specific example. It does not purport to enumerate all 
events that trigger re-evaluation — that is a function of the general 
validation model defined elsewhere in the RFC.

Additionally, the sentence at issue sits within a paragraph addressing 
schema-node concerns while itself speaking to data-node behavior. This is a 
presentation issue in an example section, not a normative deficiency. If the 
example text causes implementor confusion, that is appropriate material for a 
future revision (e.g., yang-next) rather than an erratum to the current 
specification.

--------------------------------------
RFC7950 (draft-ietf-netmod-rfc6020bis)
--------------------------------------
Title               : The YANG 1.1 Data Modeling Language
Publication Date    : August 2016
Author(s)           : M. Bjorklund, Ed.
Category            : Proposed Standard
Source              : netmod (ops)
Stream              : IETF

_______________________________________________
netmod mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to