There is no trivially-accessible link on that site that backs up the lede claim on that page that I've been "sold out" to support throttling (of all things!). As a side note: if my traffic gets throttled (gasp), I'm probably just going to switch ISPs. I have options besides AT&T and Comcast here. Go, markets!
Here is what is great about HN: someone here is going to write one comment that is going to contain more signal than this page and anything it directly links to.
Unfortunately most places (AFAIK) don't actually have options between high-speed ISPs (I'm excluding DSL). The market doesn't work so well for utilities...I'd almost prefer government regulation, just not in this direction.
Perhaps the reason for the markets not performing well is the regulation in the first place?
IMO the idea of utilities being so different from everything else is only true because the government defines them to be different and then forces people to act in that (mistaken) framework.
Ideally, it's meant to be a recognition of and response to high barriers to entry: we can't have an arbitrarily large number of competing companies all rolling out their own overlapping* infrastructures, and the few we allow must put up a lot of capital up front. The applicability of this description depends on the particular service in question (e.g. it describes running water far better than it does cell phone service).
So the solution to the problem of high barriers of entry is to make the barriers even higher?
The issue I see with the idea of barriers to entry is that a monopoly (or oligopoly) rarely if ever actually increases the quality of the service provided.
Many rail against various companies for monopolistic practices, but it suddenly makes sense to have a monopoly water or cell phone service?
Perhaps this is a solution in search of a problem.
In some situations throttling is the best solution. It depends exactly what criteria is being used and what the motivations are. Outlawing any type of ISP throttling without making some exceptions is a bad idea. For example in the past I've seen throttling used to dampen the impact of ISP uplink outages. BitTorrent and Usenet get slower but most people browsing HTTP don't notice that the ISP has lost 50% of its uplink capacity for 4 hours. Is that a bad thing? What about smaller ISPs? This impacts rural users quite a lot. We focus on Verizon, Comcast, TWC, Cox, etc but there are scores of smaller cable ISPs in the US who simply don't have the money to do large scale network upgrades required to offset 2% of their users being bandwidth hogs. Should they go out of business? Raise prices? Implement strict bandwidth caps? or throttle the 2% of users at peak hours? For the bigger ISPs the turnaround times, planning and budgeting of upgrades is never perfect. Sometimes there are gaps due to mismanagement (solvable), vendor/partner delays (not so much so), weather delays (nope) or other things out of your control. Between upgrades do you let everyone suffer with slow speeds or just throttle the 5% of people using 50% of the bandwidth? It's really not a clear cut issue. We also have to consider that the alternative to intelligent throttling are strict bandwidth caps and metered billing. I would prefer the top 5% be throttled or pay more.
Sold me out? How? Speakeasy still gives me exactly the capacity they claim to be selling me, and they still have their US-based 24-hour support if there is a problem. Yeah, that costs more than AT&T, Verizon, and Comcast, but you get what you pay for. People that choose the "popular" ISPs do so because their Internet connection doesn't really matter. If it goes down, they can just do something else. If they don't get their advertised capacity during peak hours... who cares? They're only paying $8 per month!
I don't think this is a particularly bad thing, as long as the terms are made clear in the advertising. In my experience, it is; they say "up to" when they advertise their speeds.
Where do you get internet access for $8 a month? I'm paying $20 for 100KBps, and it's the only DSL in the area for a city of ~100,000 (with the area containing ~300,000).
Here is what is great about HN: someone here is going to write one comment that is going to contain more signal than this page and anything it directly links to.