> Why not data plane interface? (Control plane and data plane, and their > distinction, are not exactly new terms/concepts...)
Several thoughts... First, I see this as a scoping problem. Trying to create both a control plane and a data plane interface at the same time opens an entire can of worms that I don't know if we want to address. Chewable chunks. Second, there's a lot of other "stuff" going on in the data plane interface area, while there's almost none (or rather none) going on in the control plane interface area. Third, a control plane interface assumes an underlying hybrid model, which fits most of the unique use cases articulated to this point. So, I wouldn't say we should rule out a data plane interface forever, but rather that we should rule it out for right now --focus on that which is different, and then move towards the more common if the need arises or the work seems worth doing at some point in the future. At least that's my thinking. :-) Russ -- <>< [email protected] [email protected] _______________________________________________ i2rs mailing list [email protected] https://www.ietf.org/mailman/listinfo/i2rs
