John Linton
.....but not quickly enough.
Actually I am only using the head line of this article rather than the following more detailed explanation because it would be nice if it were true:
http://newteevee.com/2009/10/17/is-p2p-dead-not-so-fast/
I usually manage to get my hands on the Arbor Networks report by now which I read because the trends in the US are reasonably indicative of what happens in Australia and it is useful from that viewpoint. It isn't really relevant to Exetel because of not only the differences in location but the difference in our tiny user demographic compared to an aggregation of a 100 or so ISPs in the USA. Nevertheless the overall analysis will be interesting to see when it becomes available.
The current AFACT/iinet law suit and its subsequent inevitable High Court follow up is currently focussing attention on the major uses of P2P which is the illegal transfer of copyrighted material between thieves. The cost of that ongoing theft to the various copyright holders is quantified in many media sources but I have no knowledge of what the actual figures might be. Whatever the loss is, it is sometimes characterised by being a 'profit' to ISPs who carry the thieves traffic on their networks and because of the volume (40% of all US traffic 2 years ago or whatever figure is suggested by some half way reputable source) which such commentators equate to profitable business for any ISP. I am supposing that such views are based on the simple arithmetic of "if the thieves didn't use 40% of an ISP's bandwidth then the ISP would carry 40% less traffic and therefore would be 40% smaller. I suppose, given the intellectual abilities displayed by the 'supporters of copyright theft' this makes perfect sense to them.
I have done work, in various capacities, for five ISPs over the past 14 years and have some knowledge of the costs incurred in providing an internet service to an end user. Doubtless my knowledge is incomplete and clearly the costs of providing internet services have changed radically over that time frame. However to make an assumption that ANY ISP's profits somehow can be equated to the volume of traffic it carries is not correct in any way. The reality is that if there was no P2P virtually every ISP would be more profitable is simply not the case. While bandwidth (IP and to a lesser extent customer connectivity) is a tiny fraction of the cost of what it was 15 years ago and IP is even now half the cost it was one year ago every other cost of providing an internet service to a P2P user has risen - and in some cases risen quite sharply.
The perfect customer of any ISP is one that never uses the service (and by definition therefore never needs any other aspect of the service which means they never incur any cost to the ISP other than the rental (or internal amortised cost) of the exchange port. Of course such a customer is unlikely to exist in the residential marketplaces though quite a few do in the business marketplaces. However I make the point as the starting place to illustrate that only usage costs the ISP money that isn't a fixed known cost to the business and the variations in usage caused by P2P increase the overall costs of running an ISP business disproportionately to all other types of users - though if the information in the leaked Arbor Network report is correct then youtube may have repaced P2P in that prime position.
The point is that steaming video and its equivalents (including P2P) demand more bandwidth at any point in time on any network than any other protocol - fair enough - that's what happens in any developing technology. However if there was no P2P then for the past 4 - 5 years no ISP would have required the amount of bandwidth they deployed over that time and their costs would have been less.
Now the arguers of more downloads equal more profit because the heavy down loader's plans cost more is a fallacy for at least three reasons but the first one will suffice to make the point. When you see an ISP provide a plan with major amounts of 'peak' downloads they do that by using the 'savings' associated with the plans used by low volume down loaders to 'give' those unused downloads to the heavy down loaders. The most profitable plans for ISPs are low usage plans - the least profitable (and most trouble causing plans) are high included download plans. The math should be obvious but if it isn't take a notional cost of 1 gb of Ip bandwidth and 1 gb of connectivity bandwidth and multiply the peak downloads in any set of ISP ADSL2 plans and make your own decision - the actual costs you select will make no difference.
Depending on the court's eventual rulings on the merits or otherwise of AFACT's case a win for AFACT will significantly benefit all ISPs.