Comcast vs. Net Neutrality
For anyone who missed it, Comcast laid out a new plan for managing the volume of its network traffic on Friday. (Ed note: Ben normally rants about this stuff, but he’s been swamped, so I’m crimping off an email from him.)
At first glance, it seems as if the new plan clearly violates the principles of net neutrality.
Comcast said on Friday that under the plan designed to give all users their “fair share” of bandwidth it would focus on managing the traffic of customers who are using most bandwidth when the network is congested.
It said it will use software on its network to determine if particular subscribers have been the source of high volume of traffic and will temporarily give traffic from those subscribers a lower priority status.
It said that when a subscriber’s traffic is assigned a lower priority status its traffic could be delayed if the network is congested but would not be delayed if there is no congestion.
Seems like they are punishing their users based on usage. The issues of usage-based pricing models has also been up for debate and tested in other markets by other providers, but will probably turn off heavy users. And what defines a “heavy user” anyway, especially in these days of P2P delivered legal content, Hulu, and streaming from Netflix?
However, upon further inspection and reading, the plan is much closer to what Vint Cerf has proposed - that usage only matters at peak times. If a network is getting close to capacity, and the top 5% are using more than 80% of the bandwidth, thus slowing down the entire network for everyone, the need to manage bandwidth outweighs the adherence to strict net neutrality.
As Ben said, “There needs to be balance between net neutrality and the need for ISPs to manage their networks to provide equal access. If it violates net neutrality then net neutrality is an impediment to quality of experience for internet users and needs to be adjusted accordingly.”
Well said, Ben.
September 23rd, 2008 at 9:36 am
Thanks. Just for clarification purposes that 5% of users / 80% of bandwidth example is just an off the cuff example - I did a little research and it looks like 5%/50% is closer to the reality.