On Wed, Aug 23, 2017 at 12:46:37PM -0700, Andy Bierman wrote: > On Sun, Aug 20, 2017 at 1:03 AM, Juergen Schoenwaelder < > [email protected]> wrote: > > > On Sat, Aug 19, 2017 at 07:02:04PM +0000, Acee Lindem (acee) wrote: > > > All, > > > In the context iana-routing-types.yang, we’ve been having a discussion > > of the merits of identities vs enums. We’ve followed the lead of RFC 7224 > > and used identities which allow augmentation. However, for IANA code > > points, there could be merit in having the type represent the actual > > numeric value. Any thoughts on this? > > > > > > In the next version of YANG, it would be useful for a base identity to > > allow it to have a "base-type" (mutually exclusive of "identity-ref"). For > > all identities a "base-value" would be allowed as long as it conformed to > > the constraints of the actual or inherited (via “identity-ref”) “base-type”. > > > > > > > The question here really is who is charge of controlling assignments > > of a name space. > > > > - If there is a single authority controlling the assignments, an enum > > works fine. > > > > - If the assignments are not controlled by a single authority, an > > identity works fine. > > > > We sometimes have situations that are somewhere in between, i.e., a > > central authority controlling assignments but delegating parts of the > > name space to other authoritities. I agree, we do not have good > > support to model this explicitly in YANG today. > > > > > I think Acee is asking about a new feature that does not exist in YANG, > which is to add > an enum value-stmt to an identity-stmt somehow. Seems like a valid > use-case.
Identities can be without a central authority controlling assignments - so there is little hope that values assigned to identities carry a useful meaning unless there is a mechanism to partition and delegate number spaces along the identity hierarchy. /js -- Juergen Schoenwaelder Jacobs University Bremen gGmbH Phone: +49 421 200 3587 Campus Ring 1 | 28759 Bremen | Germany Fax: +49 421 200 3103 <http://www.jacobs-university.de/> _______________________________________________ netmod mailing list [email protected] https://www.ietf.org/mailman/listinfo/netmod
