Hacker News new | comments | show | ask | jobs | submit login
Community-Owned Fiber Networks: Value Leaders in America (harvard.edu)
354 points by artsandsci 7 months ago | hide | past | web | favorite | 142 comments

Nice. Having read hacker news for years, I remember a lot of comments from angry Americans who were offended by Europeans mocking them about their broadband prices, and telling them that they were being screwed by big American telcos who formed an monopolising alliance artificially keeping prices up.

The comments used to range from "the country is just too big, the infrastructure work doesn't compare" to "European broadband infrastructure is paid by the states because of socialist-type taxation".

And finally, after years of municipalities fighting big telco, it is becoming ultra clear that yes Comcast, TWC and co have been screwing the American people for a looooong time.

And hopefully this is only the beginning, prices still have a long way to go!

2002 I'm in high school and my ISP is Adelphia they lay fiber to my parents door I'm thinking oh shit I'm about to get fiber over 56k modem. One month later Adelphia is in chaos mode over the internal corruption allegations, noone answers the phone there. I have to do a who is look up to find the server room phone number to fix and issue"how did you get this number!?". one year later big telecom is backing the government to prosecute Adelphia. 2006 TWC buys the Adelphia network and rips any working fiber parts off the darkfiber network and sells it. Still to this day my whole street and most the county has a new darkfiber network and noone has ever used it. I'm still stuck with max speed of 70/5 for $100 here. There might be some corruption.

>I remember a lot of comments from angry Americans who were offended by Europeans mocking them about their broadband prices, and telling them that they were being screwed by big American telcos who formed an monopolising alliance artificially keeping prices up.

This is from a small subset of Americans that are always looking for ways to defend large companies and argue against anything that may provide competition to the. It's certainly not the majority; I suspect that it's from employees of these companies that believe the standard company line.

This American, having worked for a small dialup ISP in high school and seeing the dirty tricks telecoms pull, was always with the other Americans that fight the propaganda from large incumbents that carved out geographic chunks of monopolies.

> And hopefully this is only the beginning, prices still have a long way to go!

Monthly non-promo rate for gigabit fiber:

Stockholm - Telia - 999 Swedish Krona (~$125)

London - HyperOptic - 60 Pounds (~$80-90)

Zurich - iWay - 79 Swiss Francs (~$80)

New York - Verizon - $80-90.

It's not clear to me that there is much room to go down in countries with similar cost of labor. (Keep in mind that median household income in France is about the same as Choctaw County, Alabama).

Small village in Serbia, 100 GB LTE for $10, or unlimited (genuinely unlimited, since it's "expensive" so nobody uses it) LTE for $30. Vip mobile, https://www.vipmobile.rs/privatni/internet/vikend_net This is even without a contract!

The speeds are about 100-120 Mbps download and 30-40 Mbps upload, ping is about 10 ms to google.rs.

Note: I don't work for Vip mobile and don't get paid for this, I just like their service, a lot.

If you adjust that $30 unlimited figure for Serbia's per capita income at the median, you get something closer to $250 to $300 in the US.

I paid 20$ for 100Mbps down in Taiwan, unlimited, in 2014.

I paid iirc 10$/month for unlimited LTE in Vietnam, can't remember the exact speeds but I remember being surprised at how quick it was.

The most population-dense of the 50 US states is New Jersey. Taiwan seems to have roughly double the density: https://www.wolframalpha.com/input/?i=population+density+new...

For that same twenty bucks you can get uncapped and unlimited 4G in Finland. Finland has 1/25 the population density of New Jersey. What was your point again?

My point was that Taiwan has about 20x the population density of the US. The higher a population density, the cheaper it is to deliver broadband per capita.

And what's your point? Are you saying that population density is not correlated with cost of broadband infrastructure per capita because there exists an outlier?

Unlimited LTE through T-Mo, including unlimited talk and text, is $70 including taxes. The closest DNA plan seems to be 35 euros ($43): https://www.dna.fi/documents/753910/853456/DNA_Liittymahinna.... But median household income in the U.S. is 50% higher than in Finland. Adjusted for that,[1] the U.S. price is about $47.

[1] Cell service isn't like an iPhone, where it's made in China wherever you buy it. U.S. cellular providers pay U.S. rents for tower sites and U.S. salaries for the skilled labor to install and maintain them.

You misread the price list, that's not the cheapest unlimited text, talk and data plan from DNA.

Anyway the cheapest list price for unlimited text, talk and data is from Globetel at 22,85€. https://globetel.fi/liittymät

However, nobody pays list prices, there are always specials as the cell phone companies battle each other for market share. The cheapest offer for unlimited and uncapped 4G is currently 2,90€ per month.

I didn't pick the cheapest plan, but the closest. There is a cheaper DNA plan but that's limited to 50 mbits, while T-Mo's service is not bandwidth limited. Also, Globetel seems to be what we'd call, in the U.S., an MVNO (it uses Telia's network). There are U.S. MVNOs with cheaper prices too.

Move the goalposts much?

What point are you trying to make, other than picking random plans?

The exercise here isn't finding a good cell plan, it's comparing service in different countries. Pointing to discounts or to MVNOs that may have service limitations makes it hard to draw a fair comparison.

If the point is comparing service in different countries then you can't just pick a single high-end plan and say these things are equal.

I will concede that including discounts make things tricky, even if they are an integral part of the market, but by what measure should MVNOs be overlooked? Unlike in the US, MVNOs in other countries generally have the same access to the network as the MVO service arm.

The fact remains that in Finland you can get a real uncapped and unlimited 4G plan for about twenty bucks. Everything unlimited for under thirty. And capped 4G plans for under ten bucks.

By what measure does the US come even close?

EDIT: for clarification the above prices are all list prices from MVOs, with the exception of the everything unlimited which is from an MVNO. The equivalent MVO plans are a few euro more.

And where i live in Texas (not a small city by far), I pay $90 for 200/20 (with my own modem) from spectrum (was TWC). The big change with spectrum has been removal of any tiers above 100/10. The 300/20 tier introduced about 5 years ago apparently still exists but you have to call and request it and there is something like a $200 connection charge, and it goes for about $110 a month.

About half way across the city there is some competition, and you can get Gigabit from a couple different providers (for about $100), but the service is really isolated to a few neighborhoods and even companies like AT&T won't even tell you which neighborhoods have the service if you happen to be looking for a house/apartment.

What is silly, is that there are a bunch of fiber providers in texas covering the little towns. For example hill country telephone coop https://www.hctc.net/resources/ (which actually has service maps) has a symetric gigabit plan for $200 in a bunch of towns with populations less than 10k..

We now have Spectrum too. It seems the trend is now to only super fast and expensive plans but the cheaper plans are being eliminated. The cheapest one is now $70 with nothing cheaper. I don't need 300/20 at home and really don't want to spend $70.

I had Everyday Low Price internet service (3Mbit down, 1Mbit up) with Time Warner, before the merger. I just did some reading online, and at least in some areas, Spectrum is still required to offer ELP service as a condition of the merger, though they do not advertise it; you have to specifically ask for it. One comment said they are only required to offer it until March 2019. Dunno if that's true or not.

Spectrum raised the price (with no notice) to $19.95 a few months ago - a 33% increase for this plan. In fairness, they also raised the speed prior to that around 10%, which of course was heavily advertised. The great thing about ELP, and the reason I switched to it, is I got sick and tired of calling TWC every year to renegotiate my rate. They bumped my 30/5 service from $55/mo to $65/mo, I called 3 months in a row, they said they took care of it, gave me a $10 credit a month or two, but kept billing the new rate. Their final offer was $59.95/mo. I just got sick of talking to them and went to $14.95/mo. The great thing about ELP is that it is a non-promotional rate, so I didn't have to call them anymore.

The hilarious thing is, after I switched to ELP, I got offers every week to upgrade to 30/5 for $49.95/mo, but I didn't care anymore at that point.

3/1 service still allows streaming Netflix and Amazon TV at 720p, though with the recent Net Neutrality neutering, I don't know how long it will be before Spectrum inserts artificial delays to make 3/1 unusable for TV streaming. If they do, I've had it and will switch to reading books for entertainment.

Spectrum's lowest-priced non-promotional service here (Louisville, KY area) is $65/mo. To get $40/mo, you have to bundle with phone or TV, getting you back up over $60/mo. To get $30/mo pricing, you have to bundle with both tv and phone, getting you up to $90/mo. Who wants a landline these days? It's just a pricing gimmick.

That was the game with TW. Cut your service to the bare minimum, then they would offer you some year long special at 1/2 the cost of the regular rates.

You could even do this over the phone most of the time. I would call and tell them the $80 they were charging me was to much that I wanted Internet Basics (or whatever they were calling the deal they had made with the city to offer an extremely inexpensive tier). Frequently they would initially deny such a thing existed, then they would transfer me to someone who would be able to setup the $10 a month plan.

From that point they would make all kinds of incredible offers. They would even send me mailers for basic cable for free if I added home phone service for $12 a month (or some such). 12 Months later they would start to slowly raise the rates (I guess trying to apply the boiling frog theory) and the offers would cease.

Apparently spectrum doesn't want to play these games.

I also have the 3/1 plan for $15. It's great value but I wouldn't mind a little more speed. But the next level they have now is $70 which is too much for what I need. I bet a lot people who would be fine with around 20MB have a more expensive plan because that's all that's offered.

> ...and even companies like AT&T won't even tell you which neighborhoods have the service...

Slightly off topic, but I always work around this by picking an address in the neighborhood and trying to sign up for service.

Gigabit fiber in Stockholm is more like $70-90 depending on provider/connection. Telia being the former monopoly with a strong brand and a tier 1 provider is more or less the most expensive broadband you can buy. Symmetrical gigabit is of course also more expensive than 1000/100, 100/100 or 100/10. Prices are still a problem, but not quite that bad.

Stockholm is a great point of reference because it's one of the least "distorted." The municipal dark fiber provider doesn't receive any public subsidies, but also doesn't have public obligations (universal service, cross-subsidizing service for low-income people, building out to all neighborhoods, etc.). It charges what it needs to in order to have a sustainable network, and passes those costs along to ISPs such that the ultimate price of access is highly reflective of the "real costs" of building and operating a fiber network.

Ah so cherrypicking it only goes to the upper and middle class areas all ISP's should be paying into the USO.

Stokab ultimately achieved 90% coverage by building out in a demand-driven way (though it took 16 years, compared to the 3-5 year timeline typically imposed in the U.S.). Which is how everything else works: at first only rich people had smart phones, now everyone does.

Universal service is an admirable goal, but the implementation is economically nonsensical. It is, in essence, a tax on telecom service. In the case of the universal service fund, it's an explicit tax. In the case of build-out requirements, it's an implicit tax: instead of taxing an ISP in cash, you're compelling it to provide a service it would not otherwise provide.

But it makes no sense to impose industry-specific taxes on telecom service. It's not like gasoline or alcohol, which warrant taxation to discourage the negative externalities they create. You tax industries that you want to discourage, not ones you want to encourage.

The downside of this economic distortion is that it suppresses competition and entrenches incumbents. There is no "minimal viable product" if you're starting an ISP in Baltimore. There is no sniping away at an incumbents' highest-margin markets. You end up not being allowed to use the tactics that startups in other industries use to take on incumbents.

Contrast how Sweden approaches broadband deployment in rural areas. It simply gives rural residents a tax credit to subsidize construction of a fiber line. It's the same way we approach other forms of welfare. We don't require Whole Foods to build stores in poor neighborhoods. We give direct aid in the form of SNAP and WIC benefits.

Where did you get that from? Stokab covers 90% of Stockholm with fiber. 4G and fixed broadband coverage is 100%.

I haven't checked the rest, but my fibre in France (provide d by Free.fr) is worth 30 euros/months with no subscription required (as in I can cancel anytime for 0 fee).

Minneapols/St. Paul

US Internet - 1 Gbit for $70/mo or 10 Gbit for $298.00/mo. Coverage is limited to south Minneapolis and a little bit of Edina.

Northstar Fiber - $70/mo. Covers only apartment and condo buildings

Centrylink - $120/mo. Has service around the entire metro, but only in random pockets.

Washington, DC: $50/mo + tax, gigabit. Promo of 2 mos free brings it to $41/mo. http://www.rcn.com I've been very happy with it, aside from upgrading my modem from docsis 3 to 3.1.

Can you get decent upstream with that? Upstream always seemed to be where cable internet fell over for me. Hard to run backups over a 10mbps link.

But 99% of home bb users don't do that you want professional level service on a consumer service that is not and never will be designed (cost reasons) to replace business service.

I have gig symmetric for $79.99. not a promotional rate, no caps.

Edit: the reason it costs more for symmetric with some ISPs has very little to do with "cost reasons" and very much to do with market segmentation. It's the same reason math coprocessors used to cost extra, even though it was on every package already. If a small group will be the only users and will pay extra, then make a separate SKU which includes it, call it business/pro/enterprise and charge a premium. It's profit maximizing pure and simple. The super obvious check is that there are ISPs which compete on friendlyness and don't do this. They're not special, they just make less money.

I do, and I’m stuck at 100/5. Knock it to 98%.

Nope. I get about 1gb down, but like 50mbps up, max.

Looks like completion at work Verizon Fios charges far more than that for lower speeds in MD and VA. However, NYC's is still twice what it costs in many cities around the world.

I'm not sure whether those other "cities around the world" are a fair comparison. Install and operating expenses for fiber are dominated by labor costs, and skilled labor in the U.S. is expensive. The median adjusted household income in the U.S. is 90% higher than in Spain, and 40% higher than in France. That's why I used London and Zurich as points of reference, rather than say Barcelona (where, as you point out, you can get gigabit for under $45/month).

Copenhagen - Hyper - 299 DKK; Gigabit - 349 DKK ($50-57).

That should be a fair comparison with London and Zurich.

I think there's some reason why it's so cheap, I remember someone explaining, but I don't know what the reason was.

https://www.hiper.dk/ - https://gigabit.dk/

Where did you get the numbers for the household income from?

FiOS in VA is $79.99 for two years (not a promotional price) for gig. Source: it's awesome.


The plan is technically $195 monthly with a ton of “special bonus discounts.” Existing customers aren’t allowed to sign up at all yet, and the price will be higher when they can.

That article from almost a year ago (which was just mindless regurgitation of the ars Technica piece which was also wrong) was not reflective of reality. It actually wasn't even correct when it was written, but you had to know to ask for the lower rate.

Don't ask PR people sales questions. Call sales, be friendly, you'll get the real price. Call PR or sound like an I-Hate-ISPs jerk, pay the jerk tax.


I'm getting 1000/1000 for $80 here in Portland, OR.

Is that the connection speed or the actual achieved through put?

1000/1000 for $60 in Chattanooga, TN. Love muni internet

How?!? With who? Centurylink?

I'm in PDX with the same speed and rate. I have Wave Broadband. I don't think it is available everywhere - possibly only in condo and apartment buildings.


Yep, CenturyLink.

Paris - 1GB - 230 TV channels- Unlimited calls to landlines in 110 countries and mobile for 10 of those countries - High-end multimedia modem with 4k capabilities - 40euros (~$50)

If you don't want all that there are fiber offers for about $25

Sydney - 90 - $60AUD (~$50) in a country as big as America with 10% of it's population

I'm really curious what the oversubscription rates are, that to me was always the big question around cheap internet access. I have 100/100 FiOS, and I get that all day every day for less than $80.

San Francisco - Sonic - (~$40/mo) for symmetric 1Gbps up/down

$75 for 25/5, or $105 for gigabit, with Comcast in Milford, Delaware.

For reference Google Fiber is $70 for 1000/1000 or $50 for 100/100.

Note (might not be obvious to all): Google Fiber is available in only a couple of places and expansion seems to have stopped. In the USA, symmetric gigabit speed is very rare and typically very expensive.

AT&T offered me symmetrical gigabit for $80 /month and 100/100 for $40 /month. This is in Fishers, IN and only available in a small slice of the city.

I stuck with comcast because uverse tv sucks and they took over 3 weeks to get me hooked up when I moved. Comcast was same day hookup. I get 100/10 for about $70.

In Chattanooga, TN, municipal gigabit will set you back $70/mo.

Montreal: $150 CAD for 1000/100 with Bell.

$90 for 250/20 southern Ontario with start.ca (backs onto Rogers pipes).

No data usage caps / overrages though, which I don't think you can get with Bell directly.

Ottawa: barely 10 minutes from downtown is $90 (non-promo) for 50/10 with Bell.

It’s effectively like a compromise between big telco and the government: “we’ll open access to broadband, if you allow us to consolidate at the top of the market.”

Yet the cost is simply shifting to cellular data plans, a market with equally high barriers to entry dominated by even fewer players. Incidentally, the cellular company and broadband company are often one and the same, thanks especially to recent consolidations.

Most of these telecoms realize broadband is a higher cost investment for lower return than cellular. So with their left hand, they’re opening up broadband. But with their right, they’re strengthening their hold on cellular. Really they’re unloading a cost center onto the public, as they attempt to move beyond the past and optimize for the future.

>"Having read hacker news for years, I remember a lot of comments from angry Americans who were offended by Europeans mocking them about their broadband prices, and telling them that they were being screwed by big American telcos who formed an monopolising alliance artificially keeping prices up"

Firstly in general people don't like to be mocked.

"A lot" of comments? Really? By a lot you mean a small handful?

Lastly the idea that Americans especially those who read HN would be even remotely unaware that they were being screwed by their well-known telecom duopoly is patently absurd. I'm pretty sure no HN reader ever needed someone from Europe to tell them they were being "screwed" by Big Telecom.

Independent and best at everything- even self-critique. Bravo!

Your comment is not only incoherent but also not even grammatically correct. You contributed exactly zero to the discussion, bravo.

You're correct on price, you're very wrong on the actual infrastructure and its performance.

The US has caught up and surpassed Europe. It is in fact questionable if the US was ever behind, it has had faster speeds than Germany, Italy, Russia, Poland, Spain and France for 20 years. Have you seen how slow average speeds are in countries like France and Italy? They're far behind.

The US average broadband speed is faster than all but five or six small European nations currently. US broadband distribution above 15mbps is also greater than all but a few small European nations.

You're also generalizing to about as great an extreme as you possibly can (Scandinavia is not all of Europe), given how far behind broadband speeds are in the majority of Europe compared to the US.

If the US approach failed so miserably, how come the vast majority of Europe is so slow? Once you get outside of Denmark, Sweden, Norway, Switzerland, and Finland, speeds begin to collapse.

There are ~750 million people in the European Union, about ~710 million of those live in nations with slower average speeds and lower broadband distribution than the US.

If we're comparing something more like US states versus small European countries, Massachusetts, New York, New Jersey, Maryland, Delaware, and Rhode Island have avg broadband speeds on par with the fastest small countries in Europe.

Countries the US beats on average speeds and 15mbps distribution: France, Germany, Italy, Belgium, Netherlands, Britain, Ireland, Spain, Austria, Poland, Portugal, Russia, Greece, Romania, Estonia, Slovakia, Slovenia, Latvia, Hungary, Bulgaria, Lithuania

Your own links confirm OP’s point. In the first link (Akamai) the US is listed above Romania, and in the second (Speedtest.net) its below. But in both lists, the US is above the big EU countries (UK, France, Germany, Spain, Italy, which comprise most of the EU population).

I didn't say anything about speed though... I'm just saying prices have been artificially high, and the fact that municipalities have needed to take the matter in their own hands is a proof that the whole thing was pretty bad.

20 years ago, Spokane Washington had to redo its streets, so they laid fiber downtown before repaving. This allowed ISP's and telecoms to plug in and run very cheap ISP's and hosting services. Downtown telecom/isp exploded with connectivity.

We found an older building had highspeed access due to fiber. We rented a small office, (closet really), and paid for 100meg connection to an ISP on a lower floor. Ran an entire ISP with dialup/dsl and hosting that was cheap, and even offered credit card processing. Years before comcast or large telecom offered DSL.

Ive seen over the years, Spokane have large backup hub for PNW banks and other large bandwidth hungry companies move in. Hands on hosting services, Internet turn key web hosting, etc. All in a midwest type town due to the availability.

Even across the border in Idaho, some smaller towns that now have fiber have startups popping up due to cheap land, cheap bandwidth, and low cost of living. I'm always seeing startups hiring big data dbas, network engineers and programmers.

I just wish they would expand further into the very small rural towns. Wifi at my folks small town community center is packed with people due to dialup being the only thing available to the community. Either drive into Spokane, hope you have Verizon LTE coverage, or pay for limitted usage sat.

When did governments stop caring about infrastructure as a direct investment in local commerce?

With a large contingent of the population believing that government is the problem rather than the solution, infrastructure investment by the government becomes a very contentious proposition even if it makes financial sense.

So you think government infallible? I guess you look at Comcast and go "what a bunch of assholes" am I right? The problem is that the same people who work at Comcast also work in government. There's a really good reason why Americans are fed up with government, and it's not because of talk radio--it's because incompetence is very well identified at all levels.

Yeah, I hate Comcast. Maybe enough to give municipal a go. But I doubt very much that in the long run municipal Internet will end up anything but a cluster from top to bottom.

Infrastructure is simply hard--it has to pretty much be a monopoly with all the problems that monopolies create. It won't matter if it is a government monopoly or a private one at that point. At least with Comcast I can tell them to screw off and buy shitty DSL (true story, I did!). With government my only choice is to move.

By the way, all the cities around mine have muni liquor stores. Only my city has privately run stores. The privately runs stores are way better. My city is also the only one that has private trash collection. The prices aren't a hair's difference, though. It doesn't matter except that if I don't want my trash picked up I am free to take it to the dump myself and pay to drop it off there instead. It would save maybe $10 a month, e.g. not worth it.

In the end, I just want ethical companies selling me fast broadband. I don't really trust our city or county to do Internet when they screw up so many simpler things all the time. They have enough to do and enough people on the payroll.

The US will have to raise taxes dramatically at all levels, to do everything that people want it to do (or figure out how to grow the economy far faster). 40-50 years ago, entitlements for example were a small cost for states and the Federal Govt (a huge slush fund for the Federal Govt in fact, they wasted trillions of dollars in entitlement money they were supposed to save). Costs have climbed a lot in all regards, while revenue as a share of GDP has largely not gone up with it.

If the US had fiscally responsible politicians, it would have a modest public debt, and trillions of dollars in a sovereign wealth fund derived from decades of surplus revenue for eg Social Security. Infrastructure would be a non-issue.

Simply put, priorities. You can do this (spending up), but you have to do that (taxes up); or you can choose not to have that (spending down), and so you can do this (taxes down).

Taxes are going up a lot either way, given the massive explosion in entitlement costs inbound, or QE is going to handle it with inflation. The US will hit a trillion per year in deficits in a few years, primarily due to entitlement costs, and partially due to excesss military spending.

This is all accurate, with the caveat that we could still have univeral broadband and other first world services provided to all if we cut back on unnecessary military spending (a conversation for another thread!).

Quite a bit of the problem isn’t spending per se, but waste, inefficiencies, and corruption (not to mention us dragging forward GDP, impoverishing the future).

> with the caveat that we could still have univeral broadband and other first world services provided to all if we cut back on unnecessary military spending

Without question. The US could shave $200+ billion off of military spending fairly easily, with no increased threat to its well-being.

However the US has superior broadband to Europe already. I don't see how the US is failing at delivering first world service in that regard. It's expensive, too expensive perhaps. If you adjust for wages, it's still too expensive but not nearly so (the median EU wages are far below the US median for example).

Japan is generally renowned for having very first rate infrastructure. They've spent incredible sums of money in relation to their GDP on that over the last ~30 years. The US has nearly caught up to Japan on average broadband speeds and distribution. That's a fact that seems to shock people when they hear it, because for so many years the US broadband situation was ugly by comparison.

Can you share more information (or a link) on Japan's broadband penetration and speed demographics?

The taxes are going up because the whole economic sector is abandoning the republic, leaving the citizens to stay and pay.

My personal opinion about this is that privatization replaced investment in infrastructure. When the companies and congresscritters realized they had all this post-New Deal infrastructure being owned run and operated by the government, they started doing their normal shadyness and tossing kickbacks to their good ol boys, and voila, horribly crippled infrastructure because some rich people wanted even more money.

Of course the excuse is always some variation of "government can't run anything right", "this service costs too much to operate", and "if we sell these assets thats revenue for us!", while ignoring all the nuance and interdependency that factors into infrastructure.

The end result for citizens is lesser quality service at a higher price, almost every time.

This is where corruption leads. Both parties are owned by K-street.

There are a few other places in WA with fiber, too -- looks like Chelan County is one. They also have super-cheap electricity and are somewhat lower crime than Spokane, from what I can tell (which is the main drawback of Spokane; it's pretty nice in a lot of other ways.)

The study's methodology of comparing only the lowest-priced tier service isn't that great, because it may be more indicative of the degree to which a provider is willing to cross-subsidize service tiers rather than real cost efficiencies. To take the top 3 providers from the chart on page 8: FiberNET in Morristown charges $200 for gigabit, Sebewaing Light & Water charges $160 for 1000/100, and LUS charges $109 for gigabit. The "market rate" for AT&T/VZN/CenturyLink is $70-90 (excluding promos).

There is also the question of what happens going forward. Quite some time ago, Easton MD (on Maryland's eastern shore) put in a cable internet service through its municipal utility. Today, you can get 200/10 for $100/month. Not bad, but head a bit east to Salisbury, MD and you can get 1000/35 through Comcast for $105 ($80 for the first year). That's the big fear with municipal services. Upgrading head ends won't get you press like rolling out a broadband service in the first instance will.

That being said, municipal broadband certainly has a place. Maryland is spending a bunch of money building a fiber backbone in the rural parts of the state, and municipalities are hooking into it: http://www.westminstermd.gov/419/Westminster-Fiber-Network. These are places that don't have the population and density to support potentially even a single private provider, much less multiple competing ones.

> Today, you can get 200/10 for $100/month. Not bad, but head a bit east to Salisbury, MD and you can get 1000/35 through Comcast for $105 ($80 for the first year). That's the big fear with municipal services. Upgrading head ends won't get you press like rolling out a broadband service in the first instance will.

A counterpoint to your argument is Time Warner didn't improve speeds or customer service until Google Fiber started building in Kansas City. Only a few ISP's serve large parts of the country. If they can ignore you while getting monopolistic profit, they will. Building a locally owned competitor might be the only way for lower tier cities and communities to get the gigabit fiber and proper customer service that is so attractive to techies like me.

Edit: Now that I think about it. I have only seen your 'fear' come up once. With Provo's locally owned fiber, which yes, they seemed to have bungled. They did get Google to come and take it over. Many other places have done so much better. Noah from AskNoah/Jupiter Broadcasting lives in Grand Forks, ND. When they had a flood that required them to redo the sewer system a decade ago they put in fiber and now have gigabit internet to residents.

Purely personal, but if my only options in Easton, MD were the municipal and Comcast, I'd probably choose municipal. The 800Mb speed doesn't make up for the fact it's Comcast and their horrible service.

I pay $85.99/mo for Fios for 25/25. I started out at about $54, and they've been bumping it up. I called and complained, and they said that's the best rate.

As an existing customer, I cannot get the "gigabit internet for $79" rate that's advertised on the website. That $70-90 rate for Verizon is the promo rate, because it will go up over time after you join.

Ok, so I can switch to Comcast, and get the $44 rate for a year, and then switch back, but that sounds...dumb.

On this specific Verizon issue - you can set up a "move" with Verizon on line even though you're moving to the same address and you can get the nice gigabit deal. Search for the details. I did it. I was surprised it worked. Nice reduction in cost and better speed.

I've had issues with Verizon letting me get the "New Customer" specials after cancelling then re joining. They wouldn't give me the deals (even though they mailed me them) because I had already had service through them.

Make an account in your wife's maiden name.

Always the dumbest part of their business model - reward switchers and screw your customer base unless they threaten to leave, then give them the same deal (maybe).

Do those municipalities have data caps?

AT&T fiber has a 1 TB data cap. That's effectively a 3 megabit average connection if you use it 24/7 with security cameras, cloud backups, etc.

Judged accurately like this, AT&T fiber is effectively a 0.003 gigabit connection.

> The study's methodology of comparing only the lowest-priced tier service isn't that great, because it may be more indicative of the degree to which a provider is willing to cross-subsidize service tiers rather than real cost efficiencies. To take the top 3 providers from the chart on page 8 ... LUS charges $109 for gigabit. The "market rate" for AT&T/VZN/CenturyLink is $70-90 (excluding promos).

AT&T's 1000/1000 in Lafayette is $80 with promo rates/$90 without, plus a $180 ETF and data caps (though removing the cap for free is often part of a promo deal). However, it doesn't match LUS's coverage area yet.

The next-fastest active residential service in Lafayette today is Cox 300/30, for $80 with/$105 without promo rates.

Closer to even offers, LUS's 100/100 is $50 with/$63 without promo rates, AT&T's 100/20 for $60/$70 + $10 per 50 GB over 1TB cap ($90/$100 without data caps), and Cox's 100/10 is $60/$85.

Cox doesn't compete with LUS's 60/60 plan; their next plan down is 30/3 for $40/$64. AT&T's 75/10 DSL is $70/$80 with data caps.

Comparisons on the lowest price makes a certain amount of sense if your main goal is to make it affordable.

Also, readers should note that they didn’t include AT&T or Verizon in their analysis because the website’s T&Cs prevented it.

Some of these community fiber networks also do financing with a balloon payment at the end of 20 years. That makes them look cash flow neutral or positive on an annual basis while they are actually losing significant money.

[Offish-topic] Lately I've been thinking about solving the last-mile problem through a free-space optics approach for suburban areas. Stick a mast on your roof that can illuminate your neighbor's masts; relay to a FTTC node for backhaul.

Sort of like an extended range LiFi / mesh setup. It's attractive because it could avoid RF licensing entirely, but rain/dust/smoke is an issue.

Does anyone know if any company is doing anything like this right now?

I've done this before. The problem with the free-space optics is they're incredibly sensitive to anything that might come in the way of the link - like birds, rain, snow, drones, anything you can imagine! Also the range is very limited. A free space optics link with a failover to unlicensed RF link works pretty well.

Although tbh it's not that difficult to deal with RF licensing, depending on where you are exactly.

This looks very interesting. How real is this? Looks like they had targeted Nov 17 to ship. Anyone know if there are units out in the wild?

What is the range on that?

"Up to 150m"

Oh I'm blind. I didn't see that, thanks.

Heard abaout a University which used this to connect the Network in two Tours in the eighties. Place had a lot of fog and birds- so there where problems, but connection was fast.

Highrises would be generally better for this approach- or some long atenna/tethered drone. At this stage though, wouldnt it be rather cheaper, to build a little softrobot and have it drag fibre along preexisting cable channels/stormdrains/sewers - basically just doing the infrastructure decent for cheap yourself.

There are existing solutions that use robots to install fiber optic cables in storm drains, sewers and other ducts. Not that cheap, but the biggest problem is that you need permission of the infrastructure owner to do it.

Terrabeam tried during the .com boom, but couldn’t get it cheap enough. Ronja is a open source 10mbit optical link. I ‘m not aware of any attempts in the past few years to do this though.

Like others have said, optical works best in areas with clear weather for most of the year. In addition, tall buildings move alot, to it works better on shorter structures.

Could any radio engineers weigh in on how practical ultra-directional radio antennas could be for this? I'm imagining collimated radio beams (like lasers) which could bypass the need (not necessarily the existing legislation) for spectrum regulation. What if every house had a few microwave drums on the roof?

Very practical, that's how some backhaul works(and why you see giant microwave dishes on mountain towers).

This Ars[1] article does a great job covering what goes into it. I know of more than a few WISP out there in WA state. They can't do the same data rate as fiber but getting a connection is pretty straightforward if there's any large mountains nearby.

[1] https://arstechnica.com/information-technology/2015/11/how-a...

This is kind of starting to happen with 60Ghz radios becoming more available. I think the problem with 'ultra-directional' radio antennas is that they're actually pretty difficult to build and even when they can be built they're very large. But with 60Ghz there is a lot of oxygen absorption, meaning that the signal dies off really quickly in all directions meaning that it won't interfere as much with other nearby installations. The equipment is just barely coming on to the market, though.

60 GHz PTP bridges have been around for 8 years or so, at 1GbE line rate. They're not rocket science, exactly, but knowing how to properly engineer a network with them is something that not every ISP can figure out.

Are you thinking of the siklu ones or someone else? I'm not aware of any that have been around that long (not arguing just curious) I was thinking of the IgniteNet and MiktoTik versions which I think are both more recent as well as PtMP applications which (I think?) Aren't really generally available yet.

True on knowing how to properly use them! Although to be fair I think most of the ISPs using them don't have a lot of use for really short, high capacity backhauls they're often looking for distance in rural areas.

Bridgewave shipped 60 GHz products 6 years before Siklu... Ignore the Ignitenet, it's a USB 802.11ad dongle glued into a radome. It's a toy. The Mikrotik doesn't have the reach to be a serious product. You can use "real" 60 GHz PTP stuff at five nines at 500-600 meters distance in a moderate rain zone, the mikrotik is good for maybe 90 meters.

Ah yes I forgot about Bridgewave. Agreed on all points for infrastructure needs. I would only add that when I read the OP (of this thread)'s comment I imagined using very very short/inexpensive 60Ghz links from neighbor to neighbor to create a kind of mesh. Possibly not what OP had in mind but it's the way I interpreted it. For that use the MikroTiks might actually make a lot of sense. (Very cheap and low range.)

Edit: Also since you're obviously knowledgeable about this stuff and I don't see a way to contact you on your profile - I've been running a matrix chat room around some of this stuff and would love to have you join! https://riot.im/app/#/room/#startyourownisp:matrix.org

I am a radio and network engineer, we use 71-86 GHz band FDD point to point 1 Gbps to 10 Gbps radio products extensively. One good example of a commercial ISP user that has publicized their use of such is Webpass (since acquired by Google) with their rooftop sites in the SF bay area.

Also potentially https://common.net/ though I’m not sure which specific transmission mechanism they use.

Oh hey I work there. We're using mostly 5ghz wireless gear.

I saw them mentioned in another HN thread recently. From the photos on their web site, they use Ubiquiti gear (probably the Ubiquiti airMax LiteBeam) which runs on 5ghz.


Just wanted to pipe in as I live in Chattanooga, TN one of the cities mentioned in the report. $58 a month for 100/100 ($70 for 1000/1000 and $300 for 10000/10000. that's right anywhere in the county you can get 10G/10G) and a bill that is always correct.

I am here in part because of the fiber network. That said it is not just the fast internet, but the quality of the service people. They have always shown up on time and gone above and beyond each time. They know how to run a cable properly (drip curve for example).

Plus they have a no data cap and net neutrality pledge.

Feels like most of the rest of of the country is in the dark ages on most or all of those points.

I live in a rural town in New Mexico of about 10,000 people. We get 50 megabit symmetrical fiber from the local co-op for about $60 a month. I think the fiber was funded by a USDA grant.

One of the nice things is not having to deal with big-co customer service like Verizon or Comcast. You call up and you get a real person who lives in your town.

Data point of one: I'm on one of these networks, here in Concord, MA, and it's amazing. We already had municipal power, so it's the same organization and the same bill. I've been running for ~4 years with no blips at all.

It's the first thing I recommend to people moving into town.


So, are there anyone already doing this? From the attitude towards Comcast, it looks like there will be lots of funding from people dying to get a better alternative.

Only slightly related, but is there a way to undo the range reduce in wifi routers. Basically- my old WiFi Router would cover the whole building- and after lightning strike - the newer model only covers a quarter of it. Support told us to buy repeaters. Wich i guess, is what they want people to do.

Are there ways to restore the transmissionrange of the old models?

Don't use repeaters each hop cuts the bandwidth in half

Use power line to connect to the router (for wired hosts) or have multiple AP's plus power line as the DS for your mobile devices.

Which band are you using now, 2.4 is slower but has better range 5Ghz is faster but shorter range also look at congestion are you using the same channel as the next house

Why can't you use your own router/ap? get one with external arenas and firmware (eg ddrwt) that can adjust tx power

Would love a how-to of community organized fiber. Seems like a simple concept with a ton of paperwork and hoops to jump.

Not exactly what you're looking for, but I've been working on http://startyourownisp.com. It's focused on wireless rather than fiber but could still be applicable.

Great resource, thank you.

...but it’s illegal in North Carolina.


this was on HN not long ago (legends of the ancient web) http://archive.is/9evG0

told how early radio was even more community than business, time to loop the loop

Lincoln Nebraska, Allo rolling out FTH throughout the city. 1Gb/sec for $90. Spectrum is trying to compete, but failing dramatically. No sympathy for the devil.

So...this is deregulation working?

Only in that alternative reality where "preventing municipal fibre networks" is the sort of regulation that the Republican Party wants to de-regulate.

In reality, it's among the only regulation they like. Plus preventing Tesla from selling directly to consumers, and making government just small enough to fit in your bedroom.

Which deregulation? The data in this study is from 2015-2016.

Semantics is important here. Most of these projects aren't community owned, they are socialised: taxpayers are compelled to fund them.

It may seem like a small difference, but it's not - there are moral, economic and engineering consequences to forcing people to subsidise your broadband use.

Imagine if people referred to State-owned factories in Soviet Russia as "community owned". That would immediately, and correctly, be criticised as propaganda.

Most of these systems are not socialized. They're from municipally-owned utilities that recover their expenses from ratepayers, not tax dollars.

And if they don't recover their expenses?

It's like transit systems. Ostensibly the user (fare-payers) are funding the system. But the true costs far exceed the amount collected, so in reality they are subsidized by the taxpayers.

In other words, the worst of both worlds. Taxpayer funded - so people are forced to pay for systems - and then 'user pays', so you pay a _second_ time to use a system that you've already paid once for. Neat :(

Why do you consider rates as distinct from other forms of compulsory taxation? It may be that rates where you are are a different thing to rates in Victoria (where I live).

Here, rates are:

* compulsory

* "progressive", based on land value, not value of services provided

* go towards services that may not be used by the ratepayer specifically, for example municipal swimming pools, not just say rubbish collection or road maintenance

“Rate payers” is U.S. regulatory jargon for “user fees charges for use of regulated services.” Typically, electric, water, or sewer. In the US, the cost of these services is not paid for with tax dollars, but recovered from fees charged to those who use those services.

Ah, right! So actually there is a large difference - rates in the US are opt-in? So if I chose not to participate in a municipal broadband system, I wouldn't have to pay for it?


Thank you for the clarification! I've been consistently downvoted on this issue in the past, and always assumed it was an expression of disagreement, rather than being due to an error on my part.

Semantics is important here. Which is why things like your completely incorrect statement are even worse.

No need for the aggressive tone. It turns out that "rates" in the US are very different to "rates" in Australia, where I live. See the conversation above.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact