Re: [netmod] Does the YANG "status" statement inherit from its parent node?

[Martin Bjorklund <[e-mail]>]

Andy Bierman <[e-mail]> wrote:
> On Wed, Dec 21, 2016 at 1:32 AM, Martin Bjorklund <[e-mail]> wrote:
> 
> > Ladislav Lhotka <[e-mail]> wrote:
> > > Martin Bjorklund <[e-mail]> writes:
> > >
> > > > Robert Wilton <[e-mail]> wrote:
> > > >> Hi,
> > > >>
> > > >> The definition of "status" in RFC 7950 in section 7.21.2 (full text
> > > >> below), states:
> > > >>
> > > >> If no status is specified, the default is "current".
> > > >>
> > > >> From my interpretation of the text in the draft, this implies that the
> > > >> status of the "new" child leaf in the following example is "current",
> > > >> and that this example is allowed!
> > > >>
> > > >> My questions are:
> > > >>  - Is my interpretation of the current text correct?
> > > >
> > > > Yes.
> > > >
> > > >>  - Is this actually the best behaviour, or should it inherit like the
> > > >>    config statement?
> > > >
> > > > I think the idea was that if the status != current, it is better for
> > > > the reader if it is explicitly stated.
> > > >
> > > >>  Should I raise an errata for this?
> > > >
> > > > No.
> > > >
> > > > However, we could have said that a current node under a deprecated
> > > > node (etc) in the same module is an error, in order to force people
> > > > (through the useage of YANG validators) to detect and fix this.
> > >
> > > Since "current" is the default, correctly deprecating a subtree would
> > > mean to explicitly add the "status" statement to every single node in
> > > the subtree.
> >
> > Yes.
> >
> 
> Please explain what it means for YANG to say
> "The parent node is deprecated and going away but the child nodes are not.
> They are current and are staying around."  This does not seem to make any
> sense.

Agreed.  But this should be invalid also if the status statements are
given explicitly:

  container a {
    status deprecated;
    container b {
      status current;
    }
  }

> Clearly an obsolete node removes all access of its descendant nodes.
> There is no way to access /foo/child if /foo has been removed from the
> server.

Yes.

> So how do I access a deprecated /foo/child node inside an obsolete /foo
> container?

You can't.


/martin