Internet providers have left rural Americans behind. One county is fighting back | Hacker News Reader

Internet providers have left rural Americans behind. One county is fighting back

theguardian.com

42 pointsrntn68 comments

Building your own optical fiber ISP, even rural, is not hard if you use the right tech [1], but most use the wrong technologies [2] and are more expensive to get of the ground.

I would be thrilled to set one up for you (remotely).

You need 1000 customers per employee and starting capital of $100K to be profitable. You use Starlink as backup to your main backbone.

[1] https://www.researchgate.net/profile/Merik-Voswinkel/publica...

[2] https://www.youtube.com/watch?v=DXYaAd5ubok&t=173s

Seems like you missed the part where large ISPs have successfully forced states to ban municipalities from building their own networks.

It doesn't matter how hard or easy something is if it's illegal, even if the laws that make it illegal are themselves extremely obvious examples of regulatory corruption.

Not all states. About 20 do, the rest do not.

States and municipalities don’t need to build it; small community-run organizations (nonprofits or not) can do so in Oregon.

The world is a lot larger than the US. Why do you assume only municipalities build their own networks?

Because the article is specifically about the insanity that is "internet in rural US"?

Can this be done for smaller areas by sharing a full time employee? For example, our community has a shared well and reservoir with a part time water manager that administers/maintains our system along with many others in the area.

Yes, I run multiple fiber ISPs.

I'd love to get in touch and see if we could collaborate.

Sent you an email.

Yes, you can run a small ISP part time, even remotely.

The process of signing up enough members of the community is a bit more work, but can still be done part time.

An optical fiber ISP will cost between $200 and $1000 per house to build, $10 per month fees are possible.

The concept described in [1] is in no meaningful way different from a traditional fiber build and does not offer any cost advantage.

You are also totally overselling the capabilities of your Fiberhood router.

P.S. Your website in the paper is defunct and the whitepaper on the site in your profile is non-existent.

The rural county I grew up in got tired of empty promises and with cash from the Obama administration has started rolling out fiber county wide. You can get gigabit fiber for a reasonable cost 10 miles from the nearest town and down semi maintained dirt roads. Any place the electrical lines go.

It works (and is the best option bar-none) until the big few lobby the state to outright ban municipal ISPs as happened in my state (NC). So frustrating.

While simultaneously saying broadband is available in your area because one house in the entire census track has it. We challenged the agreement with the town and they told us that we techmically had fiber service but it would cost 50k to activate it since they had to run fiber from their pop. Coupled with informing us that once we did so, our neighbors could then hook up for 49 bucks.

> We challenged the agreement with the town and they told us that we techmically had fiber service but it would cost 50k to activate it since they had to run fiber from their pop. Coupled with informing us that once we did so, our neighbors could then hook up for 49 bucks.

If only there were a way to spread that cost equally among the people who would benefit from it, and a legal structure for collecting that payment, representing the interests of the constituent people, and ensuring that the telco held up its end of the arrangement!

It baffles me that people are in support of that kind of thing. I doubt the majority of voters are, but the fact that anybody would say "yeah we should have our utilities privately managed for profit" is absurd

The question that would be asked, is "do you want the government subsidising and disorting the internet market"

Unlike: Do you want the government paving roads for you to drive on?

> The question that would be asked, is "do you want the government subsidising and disorting the internet market"

The other question that would be asked is, perhaps less explictly, is "do you want to drive [employer of X,000 jobs in the region] out of the state?". It doesn't matter if those jobs would be replaced by superior jobs at a publicly-operated telco, or if some of those jobs are complete bloat (and therefore paid for by inflated utility bills among all subscribers). Nobody wants to engage in that argument; it's a losing battle.

Yes, there's an easy answer to that, but when faced with the option of a short-term sacrifice in exchange for a long-term collective gain, people invariably opt not to take it.

Was your county able to get muni broadband? How did you fight the big companies?

Any electricity customer in EPB's jurisdiction can get 20gbps (synchronus) fiber to their door — this includes dead-end dirt roads beyond multiple one-lane bridges (i.e. "out in the sticks"). 300mbps is only $59 (including all fees).

More publicly-owned utilities should exist, similar to Chattanooga/Hamilton County, Tennessee.

Massachusetts created a program to provide middle-mile connectivity in the rural western part of the state and then towns hooked up the last mile. A lot of them already had municipal electric companies who did it or they partnered with a small company for the last mile stuff. It's not cheap, but it's definitely reasonably priced. It's $75/mo for gigabit with a 2Gbps backhaul to the wider internet serving around 500-600 homes in my friend's town. And that's not some intro rate where it skyrockets once the promo expires so it's actually reasonable.

The town has a much higher broadband adoption rate than the US as a whole despite being sparsely populated (under 100 people per square mile). It's also not a rich area with per-capita income under $32k. In the US, 71% of households have wired broadband, 89% in Massachusetts, and 93% in his town. So I'd say it's a pretty great success.

Since the article talks about South Carolina, I'll note that SC is the 7th lowest for households with broadband at 83% (that includes wired, wireless, and satellite). AL, WV, LA, AR, NM, MS round out the bottom 7 (MS being bottom). Massachusetts is the 9th highest with the top 8 being WA, CO, UT, CA, NJ, NH, OR, MD (WA highest).

>gigabit with a 2Gbps backhaul to the wider internet serving around 500-600 homes in my friend's town

That sounds rather oversubscribed.

Yeah 3mbit/household doesn’t really sound great here unless your other options are even more limited. I don’t know what typical over subscription rates look like, but 20% of the homes watching Netflix at once saturating the entire town’s link sounds… high.

I’m in the boonies and the DSL connection I had as a redundancy for a time was 7mbit down, and that’s by far the slowest of the options available to me.

3 Mbps/household is right about the lower bound for average Internet usage. It's not optimal, especially going forward, but it's fine for now.

Upping those 2G backhauls to 10G is simple and cheap, however.

I suspect that's a bit low to cover evening streaming usage, but my main issue isn't so much the available bandwidth as selling it as 1 gbps. There should be some statutory minimum sla or something to limit the ISP's bullshit of selling you an ULTRASPEED™ GIGABIT* 1† Gbps‡ plan that reliably drops to about 0.5% of the advertised performance every night.

3 Mbps/household is enough for evening streaming, statistical multiplexing kicks in as the number of subscribers grow. It's a bit naff, but not overly so.

Averaging 3 Mbps of usage does not mean the gigabit connection drops down to 0.5% nightly. The key here is for the ISP to have enough excess backhaul to account for the occasional burst above the 3 Mbps average. As long as there is sufficient headroom, nobody will notice the oversubscription.

This is 500 houses so oversubscribed that 1 user using their nominal bandwidth would consume 50% of the total backhaul. There isn't sufficient headroom for no one to notice.

Yes, the 2G backhaul is underdimensioned. However, that's an easy fix with a 10G upgrade. Just need to swap the optics.

In my experience, 10G wavelengths are cheaper than 2Gbps EVCs.

An interesting quote from the article I found personally relevant "adding broadband increases the value of a property by about the same as adding a bedroom".

From my personal experience living in a rural community, I would say that is a very conservative estimate.

Is that still the case even with Starlink?

Not really. Starlink wiped out the value of shitty DSL lines and microwave links to rural spots. It’s not a substitute for fiber though.

When my friend was looking at places in the middle of nowhere along the coast of Oregon to move to now that his kids are off to college, Starlink completely changed the calculus so he didn’t have to worry about connectivity at all.

My brother recently got decent wired or fiber broadband at his house in rural Maine. But he had Starlink for a while before that and it was a game-changer. You could actually do all the normal Internet things. The prior wired connections got 1Mbps down with a tailwind and the subsequent mobile hotspots had very restrictive data caps and also were still pretty slow.

Can you share the story about the provider wiring up that rural Maine connectivity?

I don't know the details. It's not that rural. It's a few miles from Ellsworth (small city) but it's down a private road. They had sorta broadband part-way down the road already. I think it's Verizon they have but am not sure. They were actually happy enough with Starlink but they switched when this became available.

Thanks! Always curious to hear more about this topic, trying to get fiber pulled to a property in Old Forge NY that currently is using StarLink (which is fine for now).

As far as I know it "just happened." Presumably whatever provider decided that it would be profitable to extend Internet to a group of homes that are decidedly not low income. Maybe they had some requests from homeowners.

I'm not even sure why my brother switched. It was presumably cost as they were perfectly happy with Starlink. (Which, as I wrote, I was perfectly happy with as well the times I was up there.)

Pre-Starlink, lack of decent Internet options were increasingly an issue with the people along the road. Not that I go up there a lot but I couldn't have worked remotely from up there pre-Starlink.

Love Old Forge! Used to drive through it all the time going to Lake Placid when I was young.

In Oregon we have “community” telcos that have been installing high speed internet in rural areas for decades, and continue to do so[1]. The service/tech provided by these companies is far better and cheaper than the national/large ones. This works well enough that I’ve seriously considered starting one to serve the ~100 homes in my areas that are an ignored island in a sea of fiber around us.

[1] https://www.telecompetitor.com/blue-mountain-lights-up-fiber...

I think WISP/point-to-point wireless over the last mile, with fiber only in the core town areas, is much more promising than pulling fiber long distances to every rural residence. This technology has come a long way and is super affordable now. That's what the government should be subsidizing. But more importantly, the feds should be preventing local governments from outlawing municipal ISPs (https://broadbandnow.com/report/municipal-broadband-roadbloc...).

Wireless tech is a flash in the pan. The gear has to be replaced every few years as it ages out.

Contrast that to fiber, you install it once and use it for decades.

Government should only be subsidizing permanent infrastructure.

Not my experience, and even if the depreciation is faster, the capex is so much lower that point-to-point wireless still comes out ahead.

It's not just the capex, but the opex also. Forklift upgrades aren't free. Furthermore new wireless kit like Tarana is expensive.

Why are we pretending starlink doesn't exist?

Are you pretending that Starlinks upfront cost of $599 and $120/mo is affordable for the community that this article is about?

$120 per month is well within the range of most monthly internet plans. The article doesn't say how much monthly service for this county-driven initiative will cost. Yes, there's an upfront cost (though cable providers typically have a hookup too) but it's not huge in comparison to monthly fees.

I'd consider paying that much for 10gpbs, but nothing less.

That's nice for you.

We got starlink for our rural property. We were previously paying over $200 a month for vastly inferior service (long range Wi-Fi), much slower and dropped out all the time.

If you can get a wired connection, get it. Starlink is for people who can't, and it's far and away the best option for those people.

If you have better/cheaper options, by all means take them. Many people do not so it's pay that or do without.

Some people, but I very much doubt most.

I wrote "many." And, yes, many people do not have good wired Internet access and paying a relatively small premium to get the equivalent of wired broadband access at all is worth it for a great many of those people. No one is asking you to pay if you have better options.

I don't think we are. Starlink is better than pretty much all of the alternatives if you can't get good wired broadband. (And can be a real game-changer in rural areas.) But it's mostly worse than good wired alternatives.

It doesn't really change the point of the article.

There are ways to make fibre broadband available, even in rural communities.

Fibre is always going to beat Starlink.

The ways to make it available involve stringing hundreds of miles of line. Of course it's possible, it's just prohibitively expensive because of the extremely low density.

Starlink costs $15k over 10 years and at the end of that you have nothing to show for the outlay.

Fiber costs far less than $15k on average and it is as good as new after 10 years.

But most rural consumers aren't ready/able to pay the outlay needed for the build. Normal suburban/urban residential broadband costs >$1000 per household to build, but rural broadband where you need many poles or lots of trenching is many times more expensive.

In some places the cost could easily be over $15k per home passed.

It’s not the cost per passing that matters, it’s the average cost per passing that matters.

Most consumers can’t or won’t pay the outlay up front, but that’s why you use financing.

Why are we pretending that rural residents are not deserving of something better than Starlink?

It is better than DSL, but it sure as hell isn’t better than fiber — by a long shot.

It's not even better than cable. There was someone on HN maybe 2 months ago who shared their speed tests and every evening it dropped to 6mbps.

Neighborhood congestion is the bane of every ISP, and starlink is no exception here. It definitely gets slower during Netflix hours, but it never becomes unusable.

> Neighborhood congestion is the bane of every ISP

Not really, congestion is only an issue for poorly designed or poorly run ISPs.

It doesn't have to be better than fiber. 100 megabits is more than enough bandwidth for almost anyone, especially people accustomed to having no bandwidth.

Talking about what people deserve is a fun way to avoid the actual question of who will pay for it. You can buy a starlink satellite for $600 today, then get broadband service for $120 per month.

People here are obsessed with fiber but I don't think the bandwidth I get from Comcast in an exurban Boston community is a lot better than Starlink.

Bandwidth needs could certainly increase in some manner in the future that makes a gigabit connection to every home in the US that gets electricity a practical necessity. But, for now, Starlink or the equivalent is a perfectly reasonable alternative at a cost that is on par with various other utilities.

is it really the same as broadband? with their latency? 5G might be more realistic

Same as fibre? No. Same as broadband generically? Pretty much. I've used it and it pretty much lets most people do what they consider normal internet stuff (including streaming video, video calls, etc.) I'm not a gamer so can't speak to stringent latency demands.

Note that Starlink is LEO satellites so different from geosynchronous satellites in terms of latency.

Latency is surprisingly good, usually under 20 milliseconds. Not the same as a wired connection obviously, but plenty good enough for almost every application.

Reply on news.ycombinator.com