On Wed, Jul 1, 2015 at 7:41 AM, Benoit Claise <[email protected]> wrote:
> Hi Lada, > >> >> >>>>> - In the introduction, you mention: >>>>> Typical use cases are: >>>>> >>>>> o Deactivating a subtree in a configuration datastore while >>>>> keeping >>>>> the data in place. >>>>> >>>>> o Complementing data model information with instance-specific >>>>> data. >>>>> >>>>> o RPC operations may use metadata annotations for various >>>>> purposes >>>>> in both requests and responses. For example, the <edit-config> >>>>> operation in the NETCONF protocol (seesection 7.2 of [RFC6241] >>>>> <https://tools.ietf.org/html/rfc6241#section-7.2>) >>>>> uses annotations in the form of XML attributes for identifying >>>>> the >>>>> point in the configuration and type of the operation. >>>>> >>>>> >>>>> Don't you have any other examples than those 3? >>>>> What about showing these examples with the spec. in this document? >>>>> Note: I see that the first one is documented with module >>>>> example-inactive >>>>> >>>> Well, the backlash I received with >>>> draft-lhotka-netmod-yang-annotations-00 (expired) show that whilst there >>>> is consensus about general utility of annotations, practical definitions >>>> may be controversial. >>>> >>> Even as example? >>> >> Yes, I am afraid. >> > I would be in favor to have at the very minimum one example. > Otherwise, it's difficult to visualize how these metadata should be used. > > We support an attribute called "owner" which is basically the I2RS client-id. At ;least one solution proposal for I2RS is going to allow retrieval of the metadata from the ephemeral datastore (e.g., client-id, secondary-id). Perhaps an example like this will not be controversial. > Regards, Benoit (as a contributor) > Andy > > _______________________________________________ > netmod mailing list > [email protected] > https://www.ietf.org/mailman/listinfo/netmod >
_______________________________________________ netmod mailing list [email protected] https://www.ietf.org/mailman/listinfo/netmod
