On Jul 31, 2017, at 4:25 AM, Mark Delany <g...@juliet.emu.st> wrote: > > On 30Jul17, Matthew Moyle-Croft allegedly wrote: >> I???m fairly certain that billing systems of this scale are pretty doable >> and not a limiting factor and isn???t a reason to not do this. > > Right. > > What I was thinking is that NBN charges each RSP per byte - that's > billing/polling a couple of 100 RSPs at the NNI which strikes me as > very tractable.
Why are you proposing that a fixed-cost network needs to have a usage-based charge? The costs of the NBN come from the number of end users it connects, not how much they use it. The incremental cost imposed on NBNco for switching capacity when an end user moves from an average bitrate of 1 Mbps to 10 Mbps is insignificant, so that user’s service provider should not be penalized by a 10x usage charge. In a regulated monopoly, the price should reflect the underlying cost structure: Fixed price per port. Some ISPs have variable costs with usage. Fine. Let them set up whatever billing policies they like; some of them charge by the byte, some do fixed-cost all-you-can-eat with a traffic policy. Whatever. Their problem. But not NBNco’s problem. It seems strange to me that the industry would propose another broken model in response to the realization of the unsustainability of the current broken model, and I wonder about the extent to which Australian ISPs ability to reason about network economics has been wrecked by Stockholm Syndrome. - mark _______________________________________________ AusNOG mailing list AusNOG@lists.ausnog.net http://lists.ausnog.net/mailman/listinfo/ausnog
