> On Jul 20, 2015, at 3:58 PM, [email protected] wrote: > > Right, each protocol has its own constraint, but do you think creating an > additional generic marker will solve those constraints ? We would expect to > be able to have the generic marker to protocol tag and also two protocol tags > with different constraints to interact between each other (I mean for > example, learning a RIP tag and copying it to ISIS or OSPF).
My thought is that by not using an element that has protocol semantics, we can free the user from worrying about them when they don't care about whether the route will or will not get redistributed into a protocol that might use it. This is mostly to deal with your "local" property noted earlier. -- Jeff
_______________________________________________ rtgwg mailing list [email protected] https://www.ietf.org/mailman/listinfo/rtgwg
