4 ms·
But the nature of being an ISP is to run expensive (both in outlay and maintenance) cable/fiber to your house, and it doesn't make economic sense to run 2 or 3
by drewcrawford 13y ago
But the nature of being an ISP is to run expensive (both in outlay and maintenance) cable/fiber to your house, and it doesn't make economic sense to run 2 or 3 or 4 lines to a customer that only uses 1 ISP. It follows directly from your argument that a person should have a healthy market of water companies serving their house, or sewer companies, or electric companies.
> running big telco networks is hard, capital intensive, and requires a lot of expertise that municipalities don't really have
Capital-intensive, expertise-intensive things is exactly what municipalities do well. They do water, sewer, utilities, police, fire, EMS, etc. My city spent just over $3b last year, meanwhile Time Warner Cable spent roughly $3b on capital expenses in the entire United States.
- minimax 13y agoTo follow up with your water analogy: we don't expect the water company to deliver 400x the water to our homes than we did 20 years ago, but that's exactly what we expect of our broadband providers. It's hard to imagine a scenario where a municipal (or monopoly) broadband service will upgrade its network as fast as we expect. Competition is the answer. My city spent just over $3b last year, meanwhile Time Warner Cable spent roughly $3b on capital expenses in the entire United States. This is about as apples to oranges as it gets, but if your city spent $3B just maintaining water and sewer, and time warner spent $3B over the entire country then TW is winning in my book.
- Retric 13y agoWait, do you want them to upgrade there service or not? If the goal is to maintain service without upgrades then minimizing cost is great, if you want improvements then minimizing cost is a stupid way to evaluate things.
- welterde 13y agoOnce the fibre (if you don't buy junk) is in the ground upgrading becomes very very cheap. And given that the required access speeds aren't very high (100M-1G), they can be had rather cheaply second hand/new. So you can direct all the money from the cost savings to upgrading your backbone, where it does get rather expensive.
- drewcrawford 13y ago> we don't expect the water company to deliver 400x the water to our homes than we did 20 years ago, but that's exactly what we expect of our broadband providers. It is a 400x increase when measured by the yardstick of a software developer, but by the yardstick of "how much money does it cost to connect Joe Schmoe to a network that meets his expectations" the business has not appreciably changed in 20 years. What changes is the model number on the Cisco switch and the type of cable in the ground, but "what is the maintenance cycle on this electrical cable and power transformer" is a problem that municipalities have been successfully solving for well over a century. I think you are confusing the practice of actual dataspeed innovation (e.g. what Cisco, Juniper, Alcatel-Lucent, Qualcomm etc., do) with ISPs like AT&T or TWC who merely buy and install equipment and service cable runs. AT&T, Time Warner, Verizon, et al do not appreciably contribute to the number of packets it is possible to send in 10 seconds, except insofar as they buy products from those companies that actually do. That role (of buying Cisco products) can easily be shifted to a municipality with no downside to Cisco. In fact, from Cisco's perspective, large nationwide ISPs operate more or less as a consolidated union of individual markets and have a lot of bargaining power, so a more diversified customerbase that doesn't cooperate as tightly means Cisco can make more money, and since they are actually the ones who make the internet fast, that seems like a positive outcome to me. > but if your city spent $3B just maintaining water and sewer, and time warner spent $3B over the entire country then TW is winning in my book. Of course not. $3b is the total budget for the whole city, which is mostly public safety (police, fire, EMS, etc.) The capital spending on energy is only about $200m/year, which is about 50% production and 50% distribution. If you imagine that it is probably easier to run fiber than it is to run 110,000V powerlines everywhere, the ongoing capital expenditure to run the actual fiber lines themselves would be less than $100m.
- minimax 13y agoIf you want to make Internet service a regulated monopoly or a municipality provided service, there will be no incentive to upgrade to new technology when it becomes available. That is the virtue of competition. Competitors can differentiate their offerings based on price and features. Competitors will install faster networks in order to win market share. Competitors will find ways to reduce costs so they can offer service at lower prices to gain market share. You don't have any of that if the municipality owns the network.
- jessaustin 13y ago...time warner spent $3B over the entire country then TW is winning in my book. TW is winning, but their customers certainly aren't. If we actually had free markets in data connection, consumers would have other providers to choose, and TW would have to spend a great deal more on their aged creaky networks in order to stay in business. Why did we defer to the judgment of local government when they wanted to shackle residents to one provider in exchange for big bribes through franchise agreements, but now that they want to encourage that same provider to get off its ass and provide decent service they're enemies of capitalism?
- Pxtl 13y agoAdding one more detail to your post: roads. Roads are the huge one. The expense of roadways is positively staggering. If cities can do that, they can run some cable.