On 7/18/07 2:13 PM, "David Barrett" <[EMAIL PROTECTED]> wrote: > For example, regarding the "hole punching" approach (by which I assume you > mean "opening a hole and trying to keep it open" versus ICE's "open a hole > 'on demand' for each connection"), it's been demonstrated to work at large > scale with at least 90% direct peer connectivity across the internet, today, > with a much simpler message flow and implementation.
ICE makes me want to pound red-hot nails into my eye sockets. That said, 90% really isn't nearly good enough for residential deployments. For cheap-o residential equipment, the margins are so thin that a support call or two is sufficient to wipe out the profit on that sale. Part of the reason that ICE is the mess it is is that the group has been iteratively fixing the various failures by adding more tests and probes and messages and whatnot. There are certainly a lot of other ways to achieve reliability but regardless of which one you choose, reliability really does have to be a priority. Ancillary to that is that when you fail you should fail in a way that can be diagnosed with reasonable ease and transparency (auditability is good!), and I think we all know what happens when traditional NAT hole-punching fails - stuff just doesn't work. Melinda _______________________________________________ Sip mailing list https://www1.ietf.org/mailman/listinfo/sip This list is for NEW development of the core SIP Protocol Use [EMAIL PROTECTED] for questions on current sip Use [EMAIL PROTECTED] for new developments on the application of sip
