[YANG] 2 more naming issues

[Andy Bierman <ietf at andybierman.com>]

Hi,

1) named augment

Is there any value in adding user-assigned identifiers to
augment-stmt instances?  SMIv2 has this feature.

The entire issue of augments, including augment of an augment,
should be reconsidered, and if so, augment-by-name instead
of augment-by-target may work better.

2) named uses

Is there any value in adding user-assigned identifiers to
uses-stmt instances?  This is relevant because a uses-stmt
can contain refinements, making it (in effect) a '1-off'
grouping definition.  It may be useful to refer to specific
uses-stmts in normative RFC text, outside the YANG module.
An identifier may be easier to read than an Xpath expression.

IMO, YANG naming would be a lot cleaner if:

   1) all groupings and objects shared the same naming scope
   2) all object nodes had a name

[object = data, rpc, or notification]


Andy




_______________________________________________
YANG mailing list
YANG at ietf.org
https://www.ietf.org/mailman/listinfo/yang