On Dec 18, 2009, at 12:55 PM, Robert Moskowitz wrote: > Marshall Eubanks wrote: >> >> On Dec 18, 2009, at 11:19 AM, Sam Hartman wrote: >> >>>>>>>> "Richard" == Richard L Barnes <rbarnes@xxxxxxx> writes: >>> >>> Richard> Here's (what the ITU claims is) the specific proposal that >>> Richard> has been made to the ITU: " An ITU spokesman said: "The ITU >>> Richard> has no plans to modify the BGP protocol, which is not an >>> Richard> ITU-T standard. "A proposal has been made, and is being >>> Richard> studied, to use BGP routers to collect traffic flow data, >>> Richard> which could be used, by bilateral agreement, by operators >>> Richard> for billing purposes." >>> >>> Richard> " >>> >>> Richard> Is this disingenuous or has the ITU really not heard of >>> Richard> netflow? >>> >>> What's so bogus about wanting to charge for traffic? >> >> Where I would raise a flag is, charge whom ? >> >> This sounds very much like the way that international long distance used to be done. That the Internet does not support that is to me, at least, not a bug but a very desirable feature. > > And you can get into transit charges. How much does Country B charge for data from Country A going to Country C? > > Oh such FUN!!!! Shades of X.25 gateways! Oh, oh, and Teletypes. Oh yes it was SOOO much fun working out the best way to route messages back in my Automotive days... We need X.400-style routing tables back! > I think this is at the heart of it, and why I'm taking it very seriously. This isn't about netflow or netflow++. As we all know, pairwise (peerwise?) settlement payments are quite common. But the only reason to put charging information into BGP is to enable -- and require -- much more complex payment regimens. From a government perspective, it might about about setting prices high for traffic from, say, an AS that hosts content they don't like. From some folks perspective, it might be "the" answer to the network neutrality debate: they don't discriminate, they just want to be paid by the sender. It would discourage folks who host or post too much content, or run services like Tor for free. I don't think this idea is coming from people who don't understand the Internet or its current economic models. I think they understand it all too well. I think this is a thoroughly bad idea that really should be stopped dead in its tracks, both for policy reasons and -- see rgm's comments -- technical ones. (Here's another amusing thought. Imagine that a charging rate announcement is denominated in Elbonian zorkmids. Do you want to download foreign exchange tables into your BGP preferences configuration?) --Steve Bellovin, http://www.cs.columbia.edu/~smb _______________________________________________ Ietf mailing list Ietf@xxxxxxxx https://www.ietf.org/mailman/listinfo/ietf