some things are simple
Well, I'm at least glad to see that network neutrality is beginning to garner blogospheric attention. Kevin Drum posted an honest appeal for clarifying commentary on the issue, to which Atrios responded with, um, characteristic pithiness. This is all good; some folks aren't yet clear about what's going on. They want either additional nuance or for someone to explain a few more times that the situation really is as black and white as it seems.
The problem is that the nuance-providers are ready to serve, whether their interjections are justified or not. In the process, they make the issue seem much more complicated than it needs to be. Witness Crooked Timber's fretting that imbalanced service will lead people to dial down their packet timeouts, flooding the net with junk in order to get their message through faster. Abandoning network neutrality would break the internet's tenuous social contract! Gasp!
This is an extremely silly idea. Application developers generally don't handle this stuff — it happens at the level of network libraries, or even lower, at the interface driver or TCP stack. There are fewer different implementations of these pieces of software than you'd think, and their authors are not going to break the internet just because Johnny's upset that YouTube keeps stuttering. These people regularly get into epic, months-long flame wars over differences of opinion about algorithmic implementation that are much smaller than this. Also, they use SpeakEasy. They're not going to break the internet without a reason so incredibly good that it only exists as a Platonic ideal.
Could there be rogues? Yes, of course. And they'd be caught and blocked, the same way that someone running a DoS attack, or voluminous ping scans, or an open SMTP relay on a consumer connection would get caught. Enforcement right now could use some beefing up, but in the unlikely event of abandoned-NN actually endangering the system's network infrastructure (rather that just its societal and economic infrastructure), you can bet there would be remedies. Contrary to Henry's assumed social contract, there are already a lot of jerks on the internet. We have ways of dealing with them.
The other side of the fake-shades-of-gray crowd is nearer and dearer to my heart: check out this lengthy series of posts over at Freedom to Tinker, which inspired the CT post. Call it the Garrulous Geek approach — to understand any technical issue, you have to start by talking about the different energy levels an electron can occupy. By part 36 of the series, we'll have gotten to principles of mass-producing crystalline silicon — almost there!
But I'm being unkind (particularly given my own guilt on this score). Ed Felten's discussion of how traffic shaping policies, inequitably applied, could degrade internet service is interesting and thoughtful. It's also beside the point: we're not talking about different ways of marshalling a limited resource. Is it important and worthwhile to think about how to prioritize traffic when a consumer's data connection is fully utilized? Yes, of course.
But dropped packets are, by and large, not what's at issue. Most of the time, most consumers are only using a fraction of their bandwidth. The average person simply doesn't have a gigantic Bittorrent download going in the background. If they do, then yes, it's good to ensure that VoIP traffic gets priority. But again, that's not what we're facing. Network neutrality is about cases like this one or this one, where Vonage customers lost service or were forced to upgrade their accounts by their predatory ISP, because the ISP didn't want competition to intrude upon its own plans for VoIP domination. They're getting more artful at these shady tactics all the time, too — rumor has it that Vonage customers using Comcast's network experience significantly degraded service shortly before Comcast deploys their own VoIP offering in an area. Ed's attribution of this problem to an architecture that innocently produces jitter problems is, I think, extremely generous — particularly given customers' earlier ability to use Vonage without problems.
We're not talking about preferentially scheduling cable company VoIP packets over Vonage ones when we face a bottleneck (although we should talk about that when we get a chance) — this is about a private firm intentionally crippling the services of another in order to provide an advantage to their own competing product, regardless of whether bandwidth is scarce or not.
So please: stop looking for nuance. It's simpler than you're making it out to be. Here, let's let AT&T chairman Ed Whitacre explain:
"They don't have any fiber out there. They don't have any wires... They use my lines for free — and that's bull... For a Google or a Yahoo or a Vonage or anybody to expect to use these pipes for free is nuts!"
But of course, they don't use them for free. They pay a broadband wholesaler. You pay the Ed Whitacres of the world for your home connection. And Ed and the broadband wholesalers (he's one, too) have complicated agreements governing how they exchange traffic equitably. Everything's paid for; nobody's getting away with anything.
It's as simple as this: Mr. Whitacre and the other ISP stakeholders have convinced themselves that when someone isn't paying them money, it constitutes an injustice. They're wrong — really wrong. Don't give an inch, don't give equal time, don't pretend there's more to it than this. There isn't.

Comments
Huzzah!
Exactly. Whiteacre's statement is Grade A baloney. No one's getting their Internet service for free.
This is nothing more than classic rent-seeking behavior by the telcos. They want to capture a piece of Google's market, and Amazon's market, and the market of everyone else that's made money over the Internet. And they want to do so using their monopolistic (or at least oligopolistic) power. Under the Robinson-Patman Price Discrimination Act, there's one word for this: illegal.
That point is key: the telecoms are playing upon the wide misconception that the intermediate hops of traffic are carried gratis by the backbone providers. This misconception is understandable because home users get a rather distorted view of the internet: most users have capped-rate, unlimited access, download-biased payment plans; if that's all you know, it's hard to imagine alternatvies.
Consider this a question of innocent ignorance: What happens if major institutions are unwilling to pay for preferential bandwidth? In particualr, I am thinking of Google, but also major not-for-profit institutions like universities or the federal government. Is the slow down going to be enough for "ordinary" personal users? Could this drive users back to dial-up service, hence rendering the fiber-optics worthless?
Post A Comment