> On Aug 14, 2017, at 11:24 AM, Reshad Rahman (rrahman) <rrah...@cisco.com> > wrote: > > I am gradually catching up to emails so I may not have absorbed all the > emails I have gone through yetÅ . > > Regarding echo config, we agreed in Chicago to remove the echo config > based on the fact that implementations of echo are vendor specific. e.g. > An implementation which has echo as continuous would likely have echo > configuration whereas an implementation which has echo as on-demand may > use RPC. There is an example of a vendor augmentation for echo config in > Appendix A.1.
Right. The point was what the echo configuration should look like. Based on glancing at a couple of vendor configs that use echo, it's mostly "use echo" as a boolean behavior. This is in contrast with permitting the user to specify distinct echo vs. async timer values. I'm totally fine leaving the "on demand" case out of scope. -- Jeff