WTF is NET NEUTRALITY, anyway? And how can we make everything better?
Ignore those beardies - here are the facts
Going to war for um, capitalism
Let's return to you as video provider. Netflix - as you too would hope to - grew and grew. It bought access to a private network, Cogent, to deliver its streams to its customers' ISPs quickly, and it also used caching inside ISPs to store its most popular shows. That latter relieved Cogent of some of the traffic load, and saved Netflix a bit of dosh.
Others have taken different routes. Google operates the world's biggest private network, and it had to build it so that it could deliver YouTube's video. Nobody else can use Google's road.
Netflix found as it grew that congestion at Comcast - the massive US ISP - was causing it problems. So it did a fairly sensible thing. It dropped Cogent and peered with Comcast directly. This actually saved it money. It's a popular myth - propagated by staggeringly bad reporting - that Netflix now "pays twice". Er, no: it pays Comcast directly, and it pays less than it used to. Net neutrality campaigners portray this as Comcast "demanding a ransom", or a kind of extortion. That isn't actually the case.
The Neutralists fear that every "over the top" content provider will have to pay not just to get the video to the ISP, but will also have to pay the ISP to carry the packets the "last mile" into people's homes. A popular graphic campaigners like to exchange shows an ISP demanding fees from Twitter and eBay, and passing them on to the end user. The sky is falling!
Well, let's see.
For a start, eBay and Twitter don't need a 'fast lane', they don't need to hire a CDN. They use very little bandwidth compared to even a modest streaming media provider. They don't generate asymmetrical data flows between networks. And even if they did need a fast lane, their choice of providers is quite broad. There are many networks to choose from.
"Ideally, a smart content company would have several different competing CDN type operations to choose from – and they'd choose the lowest cost routing," explains Sanders. "But what they should be worrying about long-term is the fact that almost none of the traffic these days is going over what you might call 'the internet'."
And here the real competitive concerns about the direct deals are plausible. Sanders sums it up:
"The buyer (a Netflix or Google) achieves a far greater efficiency of content delivery, and is able to compete better for the contended, congested bandwith on the consumer network, which effectively squeezes out competitive video from the internet."
"The dark side of this is that by choosing which route to direct traffic into a network, a large content provider can coerce consumer ISPs into co-locational private peering, by deciding which route their content takes. This only happens when the flow is hugely asymmetric – when there's no peering of equals."
Have you got one of these, Mr. Beard? Thought not
And the result of that is that a Netflix or Google can squeeze you, dear video startup, out of the game.
This is the real competitive issue raised in the USA by the latest Netflix agreements, and it also explains why Google is such a spectral and ambivalent presence in neutrality debates. Google sponsors 150 front groups, many of whom campaign "for" regulation, to "Save The Internet" - but it's also keenly aware of how its own fast lanes give it a competitive advantage. It's got tens of thousands of servers at ISPs around the world. Not many plucky startups have that.
In other words, Netflix/Comcast is not "Big Old Media Video" bustling out "Plucky Internet Startup Video" - it's "Big New Over-the-Top Internet Video" ensuring it can muscle out "Small New Startup Internet Video". It's Netflix buying a competitive advantage.
Should gay packets be allowed to marry?
By now I hope you'll see that the map you have been given does not describe the territory you're living in. On the map - the map the Neutrality campaign presents to us - are things like highways, lit by things like "packet equality", and featuring "threats" (here be demons!) like "pay to play". But we can't use a faulty map to develop wise policy, particularly if this entails technical regulation.
It's hard to resist quoting Tom Lehrer when earnest bearded young men campaign to "Save The Internet" - particularly when they give us a misleading picture of how it actually works:
(Lehrer sang: "We are the Folk Song Army / Everyone of us cares / We all hate poverty, war, and injustice / Unlike the rest of you squares.")
It's certainly been true that scaling an internet business (like a Twitter or a Facebook) is cheaper and faster than scaling a physical business. Growing your corner shop to compete with Tesco and Asda nationally takes a huge amount of capital investment. But guess what? Scaling Twitter of Facebook costs money too. Twitter only got to an IPO after burning through hundreds of millions in VC capital. The utopian premise that you can rule the world with just a good idea always was a bit of a fib. But we want (I hope) a better internet - one where new ideas can flourish, rather than be crushed by incumbents (like Netflix) with deep pockets. What shall we do?
That's really scope for another article, but I'd suggest one or two things right away.
Many readers complain that the US access market is not competitive. Yet more people prefer to complain than switch providers. Some 89 per cent of Americans have a choice between at least two broadband companies - so why not switch? Why not throw a switching party for everyone on your street? Start acting like customers - it's the only thing that scares the hell out of an ISP.
Another useful ploy is to ensure that existing US anti-competition laws are enforced. In an area with a duopoly, collusion between a DSL and cable provider should be illegal. Are you sure they're not colluding? Really sure? There are competition authorities that love to examine this sort of thing - so make their day.
Further out, identify and support competitive services and technologies. The FCC drove local ISPs out of business by allowing the local Bells to retail at wholesale rates. Get this repealed. If you're a gamer, seek out an ISP that promises low latencies, for example. Encourage others to do so. Wireless technologies now look a better bet for rural coverage. Get agitating.
And finally - remember that large corporations will always want a free ride, or a leg up - and yes, sometimes internet companies do too. They're not afraid of telling lies or manufacturing myths to gain an advantage, particularly when there's a lot of public good will toward the internet "as it is today".
So don't be taken for a mug. ®
Sponsored: Benefits from the lessons learned in HPC