Hi, for identifying and referring to mount points, two options are proposed:
1. Use a YANG extension, such as yangmnt:mount-point, for giving a name to a schema location. The specification of subschemas (mounted data models) then uses these named mount points. 2. The specification of subschemas uses schema node identifiers for addressing mount points. The rules can be same/similar to those for a target node of augment. My opinion is that #2 is a better choice because - schema node identifier is a known koncept, and it can refer to any schema location (of course, restrictions may be imposed, as they are for augment's target node); - an extension is not appropriate for such a fundamental data modelling mechanism; - schema node identifiers can be directly used even if the mount point is inside a grouping, the mount-point extension requires some CLRs. It is questionable whether specifying mount point in a parent model is good or bad. Its author must be able to know all potential mount points in advance, which may not always be possible. On the other hand, explicit mount points do not guarantee that new stuff can only appear there - it may be added to any location through an augment. Lada -- Ladislav Lhotka, CZ.NIC Labs PGP Key ID: E74E8C0C _______________________________________________ netmod mailing list [email protected] https://www.ietf.org/mailman/listinfo/netmod
