On Jan 10, 2008 6:15 PM, Henning Schulzrinne <[EMAIL PROTECTED]> wrote: > If it's optional, would it mean that the upper layer has to be aware > of its existence, possibly re-running ICE on ORCHIDs? >
I think it may mean that the upper layer (other applications) need to be aware they don't have to run ICE, at least for efficiency (they could do it). > Can we envision hybrid HIP-non HIP overlays? > > Again, I'm rather worried about the complexity of this whole effort. I > think one reasonable criterion of going forward is that those who are > *not* interested in using HIP shouldn't have to pay for it, in terms > of implementation costs, interoperability or performance. > I fully agree. It only makes sense if it's actually a separable module. Bruce > > On Jan 10, 2008, at 6:05 PM, Bruce Lowekamp wrote: > > > Henning, > > > > I think it can done in a modular way that will make it possible for it > > to be optional to implement and run. Then vendors and deployments can > > decide whether they want to use it or not. > > > > I'm not entirely convinced that HIP is the right solution, but I'm > > interested enough in it that I think it will be cool to play with and > > see how well it works. That to me says it should be an optional > > component. If it's successful, obviously it will be more widely > > adopted. > > > > Bruce > > > >> > >> > >> > > _______________________________________________ P2PSIP mailing list [email protected] https://www1.ietf.org/mailman/listinfo/p2psip
