Hacker News Reader

Sonic.net is, for my money (and they get my money, rather than Comcast or someone else who would sell me a fatter, but less capable or otherwise somehow restricted, pipe for less money), a perfect example of a "good ISP".

What makes them good? They act like a dumb pipe. They don't block things. They don't log things. They don't throttle things. They don't care what I use my service for. They fix things when things break. They stay out of my way.

And, somehow, they manage to bring in enough money that they're deploying FTTP trials in multiple cities.

> there's no money in building "dumb" anything

Historically, there have been many successful utility providers that earned a lot of money producing simple products. Telephone companies, power companies, and oil refineries, neither won a stronghold by producing "smart" products. Most however did produce illegal monopolies.

1) Utility providers have historicaly been legal monopolies. I.e. in return for rate regulation, they get protection from competition.

2) The telephone network was a marvel of engineering at the time. Same for the power grid, power plants, oil refineries, etc. That infrastructure is the product of huge amounts of "hard engineering."

1) ISPs are also essentially monopolies. 2) It is also incredibly expensive to build ISP infrastructure.

1) Not really. Most developed areas have at least two ISP providers, the local cable co and the local telephone company.

30% of Americans have 1 or fewer wireline broadband connections. Satellite and 4g cover a lot of them and provide at least some competition.

That 30% is mostly rural or small urban pockets that are just outside of DSL's effective range.

Does 4G really offer much pure competition when the two biggest (by number of subscribers) wireless companies, Verizon and AT&T, also do business in FTTP/DSL services in many of those very same markets?

Verizon: http://www.broadbandmap.gov/about-provider/verizon-communica...

AT&T: http://www.broadbandmap.gov/about-provider/at&t-inc/nationwi...

Satellite, 4G and microwave connections are poor substitutes for a wired connection. The sheer amount of latency and packet loss makes them not even an option for my connection. I would hardly call them competitive. For a wired connection, I have precisely one option: Comcast.

Even if latency and packet loss weren't an issue for my use, there are additional restrictions that come in to play. For instance, my HOA doesn't allow any satellite dishes or microwave antennae to be mounted in the neighborhood. While petitioning my HOA for change would be much easier than trying to petition for national legislation, why should those companies be allowed to divvy service areas like gangs agreeing to stay out of each others' "turf"?

A duopoly is no better than a monopoly for all intents and purposes. It creates the illusion of choice. ISP A sucks at this, this and this, whereas ISP B sucks at that, that and that. There's a voting analogy to be made there, but let's try to stick to the topic at hand....

Point-to-point microwave latency should be the same as fiber latency at the same distance.

Multipoint systems used by most WISPs are TDMA-based and should have a fixed latency well below 10ms depending on how many active users are on a cell.

HOAs aren't allowed to prohibit antennas under 1 meter in diameter per OTARD [ http://www.fcc.gov/guides/over-air-reception-devices-rule] You should educate your board and contact the FCC if they balk.

> Point-to-point microwave latency should be the same as fiber latency at the same distance.

Actually, it's often less because the microwave travels in a straight line while the fiber has to follow the utility poles or trenches (which often aren't in a straight line due to right of way issues). High frequency traders performing latency arbitrage use microwave dishes instead of fiber because it's a fraction of a second faster than using fiber connections on a cross-country connection.

Microwave dish systems are, however, a nightmare to support because transmission can be affected by the weather, birds, and a whole lot of other things. They also aren't very high bandwidth, and if you have a lot of dishes that means a lot of interference. I'd say they're perfectly fine for things like extending your LAN to your cabin up on a mountaintop, but not great tech to run an ISP off of (though people still try).

I'd be psyched to get a broadbrand provider that works like my electric company. The rates are reasonable, the power sufficient, the reliability is very high.

The electric utility in Chicago just decommissioned a pair of century-old coal plants that were costing the city tens of millions of dollars a year in negative externalities. Yeah, the electric utilities are exactly who I want investing in infrastructure that has becomes obsolete in 5-10 years.

Well my utility is a distribution company, not generation, so their infrastructure is mostly wires and substations. The equivalent for broadband would be fiber cables and switching offices, which seem like they actually could last for a very long time.

Once fiber is run to every house, I can't imagine why it would ever need to be replaced. I think it's actually physically impossible to have a faster data transport medium.

Didn't we just go through the BPON -> GPON upgrade? If you're just talking about the fiber, you're not even talking about the last mile. It's the last 100 feet. What about all the equipment cabinets in every neighborhood?

What about them? Once they're sited and hooked into the lines, replacing the gear inside is not difficult.

The question is whether a regulated monopoly could provide good broadband connectivity at a good price. Equipment upgrade scheduling is an implementation detail.

And it's a detail that is already priced into the service. It should be possible for such a monopoly to provide the same level of connectivity as Verizon or Comcast at lower prices, because for their current prices Verizon and Comcast already provide that level of connectivity plus content and software and marketing.

Big ISPs do everything necessary for connectivity, but they also develop and sell products that run on that connectivity. For example Comcast's on-demand and DVR everywhere solutions compete directly with services like Netflix, HBO Go, NBC.com, ESPN 3, etc. Developing, maintaining, and marketing that costs Comcast a lot of money--money that a connectivity-only monopoly would not need to spend.

It would be like if your electric company owned the lines, owned the power station, and sold you appliances like TVs and washing machines. Instead, electric companies just sell electricity, and everyone else sells you things that use electricity. I can't think of a reason that a similar architecture would not work for the Internet.

Wont decoupling stuff help? Make the last mile maintainer a separate entity.

We need last mile liberation and we need to prevent ISP into entering content business.

When it comes to oil producers and Telephone companies, both has been fined heavily for doing illegal monopolies. If there have been legal monopolies in trade for regulation, those companies would had not been found guilty in both US and EU courts.

As for your second point, my comment is only about dum products, not competition in production of said products. The cheaper you can produce a product, the more reliable the product chain is, the better a given company will do in profits.

>At the end of the day, users are willing to pay, e.g., $15/month for Netflix, and don't care about who gets what part of that money. Fights over net neutrality and copyright are about how this $15 gets split up between: content creators, infrastructure providers, and user-facing websites.

This is incorrect, the actual fight is about killing Netflix as a business. The idea is to gradually force Netflix to raise their prices to a point where it no longer seems to be a cheap alternative to cable/satellite television.