From mboxrd@z Thu Jan 1 00:00:00 1970 From: ggm@algebras.org (George Michaelson) Date: Tue, 12 Dec 2017 11:42:58 +1000 Subject: [TUHS] V7 Addendem [ really lawyers and AT&T consent decree ] In-Reply-To: <20171212012835.B290518C08C@mercury.lcs.mit.edu> References: <20171212012835.B290518C08C@mercury.lcs.mit.edu> Message-ID: I don't think this list is the right place to conduct that particular debate. Its true RSVP didn't get traction, but the economics which underpin it are pretty bad, for the current Internet model of settlement and 'who pays, and when' -There was no point at which RSVP was going to deploy into the inter-carrier settlement regime we have, and have had for some time. It didn't actually mean buying 'more' of anything, it simply meant pushing people who wouldn't buy more, into smaller drop buckets. I'd counter (sort of) with a comment that I heard at NANOG San Jose from a US tier-1. There is more glass in the ground, than lit, by at least one order of magnitude. If you have congestion on any US domestic link, its not because you don't actually have clear channel, its because somebody is making money from artificial scarcity. I don't know for sure that the same is true trans-atlantic or trans-pacific, but it would not surprise me if there is a lot of unlit capacity, and more dropped packets than strictly speaking the glass expects. n-way conferencing is about as stressful as it gets for loss, and delay. I think its a minor miracle I can do 3 or 4 way, heads and voice at all. If I was paying, I'd expect better. Free QDU's are like greshams law: bad (cheap) comms drives out good (paid) comms. -G On Tue, Dec 12, 2017 at 11:28 AM, Noel Chiappa wrote: > > From: Steve Johnson > > > Recently I've been attempting to Skype on a group call with 5 people in > > Europe. I would LOVE to have a guaranteed bandwidth for my call. > > The Internet engineering community did quite a bit of work on resource > guarantees. (Google 'IntServ' and 'RSVP' - the latter is the control > protocol.) > > Unfortunately, there was never much interest in it. People started doing > voice with just plain 'best effort' service, and I guess it worked 'well > enough' that nobody was interested in IntServ/RSVP. > > Noel