On Tue, Jul 10, 2018 at 4:30 PM, Jeffrey Haas <jh...@pfrc.org> wrote:
> [Note that I'm making this comment for posterity so that the IESG doesn't > ask the same thing of other draft authors.] > > On Thu, Jul 05, 2018 at 06:18:42AM -0700, Eric Rescorla wrote: > > On Thu, Jul 5, 2018 at 6:10 AM, Reshad Rahman (rrahman) < > rrah...@cisco.com> > > wrote: > > > <RR2> The pyang tool does this for real references (leafref) but not in > > > the grouping case (which is reuse and not reference). Even though > there is > > > replication in the tree diagrams, I believe there is benefit in seeing > the > > > complete tree for each module. > > > > > > > Well, this is a comment so you're free to ignore it, but IMO it makes it > > very hard to read these. > > The fundamental issue here is one of clear representation of a node being > instantiated vs. the semantic idea that we're seeing a template (grouping) > instantiated. > > At the moment, pyang output is showing that when you use a grouping, you > get > an actual node in the tree. Given that things like augmentations cannot be > done to groupings as a logical entity and have impact everywhere, this is > perhaps an important detail, if annoying. > > The proper place to pursue a fix to this is asking for a bis on RFC 8340. > As I already said, it's a comment, so feel free to ignore it. -Ekr > -- Jeff >
