Ed Felten has posted a really great technical overview on the Nuts and Bolts of Network Neutrality. The paper isn’t long, so hopefully lots of non-technical people will read it. The connection between technical router policy and larger network neutrality policy is very valuable.
I’m unclear on one point, though. Prof. Felten argues that the importance of quality of service guarantees, often a retort to proposals for network neutrality mandates, can be overemphasized. He suggests that QoS is inapplicable when the speed needed by the application is either greater than, or significantly less than, the speed of the network. Speed, in this context, seems to be both latency and bandwidth. Prof. Felten suggests that voice applications don’t really need QoS, because the speed of the network is great enough in comparison that the “valleys” of speed are never too severe.
The reality, though, is that we’ve never had an Internet with end-to-end QoS guarantees. While voice might not need QoS, other high-bandwidth interactive services might benefit highly from regularized speed. This kind intelligence just seems to make sense to further additional innovation.
Prof. Felten concludes that we need more time to determine what regulation would be appropriate. I think that’s right. The bills we’ve seen ignore the subtle technical points that he covers in today’s paper.
Posted by M as Net Neutrality, Interesting Link at 2:34 PM EDT
4 Comments »
I thought I’d talk a bit more about
Tim Schneider’s recent
post at
Public Knowledge.
Tim links to an incident where Verizon terminated an EVDO customer’s account. Verizon claimed that the “extraordinarily high usage” of the account was proof positive that the customer had violated the terms of service by streaming or downloading music or video.
But what if we had some form of net neutrality? It would seem that Verizon could still be free to terminate accounts at their discretion. It might make good business sense to do so. If bandwidth costs Verizon money, even over DSL, then excluding the tiny fraction of their customer base who consume a vast majority of their bandwidth would make sense. They might not even need to raise the bogeyman of “terms of service” and face the non-neutral net problem that Tim describes. They might simply have the right to terminate for unusually high usage, or for nothing at all. Or, less controversially, to move you from a flat-rate plan to a metered plan.
The problem with flat-fee Internet might be that less skilled, less demanding users (mom and pop checking email) may be subsidizing highly skill, high consumption users. If $29.95 has a reasonable profit margin for the 300Mb a month downloader, can it still cover the cost of the 30Gb a month downloader?
There’s a very excellent article (via Techdirt) by Tom Evslin, founder of AT&T Worldnet, that questions whether bandwidth usage like this actually costs the ISP at all. It certainly has questioned some of my assumptions. But for services like cable Internet, the local loop can be saturated by heavy use, and certainly EVDO and wireless will have similar issues.
Now, granted, the Snowe/Dorgan amendment might have limited this with Section 12(a)(1): “not block, interfere with, discriminate against, impair, or degrade the ability of any person to use a broadband service to access, use, send, post, receive, or offer any lawful content, application, or service made available via the Internet;”
But this goes well beyond, I think, most peoples’ concept of net neutrality. Here, in an effort to block preferential and discriminatory treatment of content over the Internet, we suggest that there’s a limitation on when and how an ISP can deny service to users.
Net neutrality might be sound, but doesn’t this kind of regulation seem a bit too much?
Posted by M as Net Neutrality, Internet Policy, Interesting Link, Recommended Media at 11:10 AM EDT
No Comments »
Tim Schneider at Public Knowledge has written a very excellent post on the usage restrictions that Verizon places on the their EVDO usage. Tim suggests, I think rightly, that this is a future we might see without network neutrality guarantees.
Some commentators have raised the specter of cable companies turning the Internet into cable TV, with top-down bundled content relegating individual and non-corporate content to an Internet version of the fuzzy local access channel. It seems far more plausible to me, though, that phone companies will seek to transform the Internet into the cell phone / cell network model.
If technology progresses to the point that computers can be effectively, if not completely, locked down, we might have Bells distribute free computers that could only connect to their network. Their network might have a complicated per-email free structure, with premium zones and bundled content.
Granted, the tide seems to be turning against this. AOL apparently plans to open up their walled garden. If AOL, with the content pockets of Time Warner, doesn’t see a future in this type of access/content bundling, maybe we should be less concerned and more optimistic.
Posted by M as Net Neutrality, Interesting Link, Recommended Media at 10:42 AM EDT
1 Comment »
Nicholas Carr has a good review of Jonathan Zittrain’s article, The Generative Internet. Zittrain’s article is a must read, though I feel bad suggesting that, having still not gotten to Yochai Benkler new book
.
I wonder who bears the costs of accidents or negligence at the edges of the neutral network? Even with overwhelming benefits, there’s clearly a cost to having new innovation at the edges of the network. I think distribution of that cost is a very appropriate question, though I’m not at all sure what the answer should be.
Read the rest of this entry »
Posted by M as Net Neutrality, Internet Policy, Interesting Link at 12:31 PM EDT
No Comments »
