This article contains some technical inaccuracies, some hand-waving to distract and make people look the wrong direction, and some outright falsehoods – in other words, the average net neutrality article from Wired.
I’m guessing that Wired is probably getting paid by someone to write these articles. Otherwise, there is no real explanation for it. Ignorance just isn’t enough to account for the frequency of their appearance and the inaccuracy of the “facts,” or the common irrelevant misdirections.
I don’t really have the patience to demolish the inanity point by point, but all the talk about content delivery networks really has nothing to do with net neutrality and the current debate over what ISPs are doing, which is attempting to extort money from large bandwidth users (and eventually, of course, others) by artificially bottlenecking traffic.
Technically speaking, there are two main ways to bottleneck traffic. First, there are various methods to actively do so. For instance, I could on a Cisco router use a committed access rate to rate-limit an interface(s) using a QOS group. (If that has no meaning to you, which it shouldn’t, don’t worry about it. Not really important.)
The second method is passive bottlenecking. For instance, I could put in a 100Mbs hub from 1996 where a 10GE switch should be. This is effectively what the ISPs are doing – they are refusing to provide the infrastructure on their side that would allow traffic to flow properly between a backbone provider (which Netflix et. al are on the other end of) and the ISP.
Content delivery networks – which have been around since Akamai was founded in 1998 (part of another inaccuracy in the article) – are a distraction and fairly irrelevant to this discussion.
Historically (as a commenter mentions), ISPs have welcomed caching servers in their data centers because it provided better service to their customers.
Now the ISPs have decided to charge large content providers for putting these servers in, or having a more direct connection to the ISP network, or both, while passively degrading the traffic that has not been paid for. In the past, they would have begged (or even occasionally paid for) these caching servers.
This will spread, of course, because the point is to turn the internet into cable television where you pay for packages, such as the “$89 a month package – includes Amazon, Ebay, and ESPN!” or the “$99 a month package – included Netflix and Hulu in 4K!”
This is the end goal.
The ISPs are using the leverage they have to achieve this, and Wired even while criticizing them a bit is enabling them by spreading inaccurate and/or irrelevant information.
I like this comment about Wired and its approach, which sums it up better than I could:
Does Wired not realize they have an educated tech savvy crowd that they are trying to trick? Do they not realize that it’s business suicide to become a clear enemy of your customers? Who will they advertise too if they have no readers?
As in many areas, Wired, being part of the establishment, only likes to examine options that are friendly to the established and along already-set paths. This article examining it in a larger context is but another example of that.
For instance, why is internet access not a public utility? Why are CDNs not available to all at decent prices? Why is our country the only one who can’t seem to do any of this correctly? Why is internet access in the US so terrible even compared to much poorer countries?
These are questions that do not defend the status quo, so they are ones that Wired will never ask seriously.