Assuming your talking about the US with your question (since the article is), than in the US it really is _that_ expensive.
The main reason it's "easy" in France (and UK, and other places) is that the big name's network are required by law to let you use the end point. That exists in the US on the phone lines (legacy telephone regulations). But not on anything else (no cable or fiber).
So unless you want to pour all of your time and energy into offering flaky, slow, unreliable internet over 100-year-old copper lines, with speeds below 6mb and minor outages daily, your out of luck re-using any lines in the US.
That difference alone suddenly makes it not remotely easy or cheap.
To deploy internet at a reasonable price, you have to start an entire fiber construction company, and local political lobbying firm, simultaneously, just to get a single customer.
We're talking a barrier to entry of at least 1 to 2 million dollars, to wire a neighborhood, just to get your first customer signed up. Customers who've been conditioned to pay no more than $80/month for internet access.
That cost is actually not expensive compared to typical municipal projects. But it's way more than you'll ever get funding for, to chase, unless you are already extremely wealthy from some other business / family history / etc.
Venture firms would rather fund one thousand new Instagrams than pour money into anything remotely resembling a utility business.
The main way around that cost barrier is to deploy "bad" technology, like wireless. That can be done on a shoestring budget. It works really well in places where there is zero density (rural, exurbs). But it's not the future, it's legacy deployment, with legacy problems and legacy speeds. The nicest microwave radios you can buy only barely match the speeds and reliability of the cheapest fiber deployment.
So, if your in any sort of reasonably-sized city, you'd be stupid to not roll out fiber from the start. And to do that, your looking at millions of dollars of cost.
No one will ever fund that. (I've tried -- countless times). Typically, only selfie-sharing iPhone apps can generate investment like that.
This is why most people in the US are pushing for "common carrier". Because no one here believes a honest business will ever deploy a modern network on any large scale, so the best we can hope for is an upgrade from 100-year-old phone lines, to 40-year-old cable lines.
The main reason it's "easy" in France (and UK, and other places) is that the big name's network are required by law to let you use the end point. That exists in the US on the phone lines (legacy telephone regulations). But not on anything else (no cable or fiber).
The regulations are on the books, and through the mid-2000s UNE-P was a somewhat-viable business model. However, the ILECs figured out that they can sabotage the fulfillment process with no repercussions. When your new customers can't get their service actually started, and the only thing you can tell them is "Verizon is being mean to us", you're not long for the business. That's why the only remaining CLECs are "facilities-based", i.e. they don't use the UNE-P program.
This is why I'm not so hopeful for a similar program on the data side. I've seen the kind of bullshit the incumbents pull, and I greatly doubt the FCC will ever prevent that. I don't think investors are dumb enough to fall for the 1996 bait-and-switch again, so any effective program like this would have to include strong structural protections for competitive ISPs. My understanding is that the most effective such programs in Europe have done that, e.g. forbidding the network owner from selling to retail customers.
The main way around that cost barrier is to deploy "bad" technology, like wireless. That can be done on a shoestring budget. It works really well in places where there is zero density (rural, exurbs). But it's not the future, it's legacy deployment, with legacy problems and legacy speeds.
That's the current state of the art, but maintaining that decrepitude has taken a decade of FCC foot-dragging on white spaces and other "cognitive" radio applications. We are nowhere near physical limits of spectrum bandwidth, and radio engineers of the future will laugh at our aversion to innovation in this space.
I think the best thing that could happen for competition and choice in data connection would be for radio regulation to be taken away from FCC and given to a less political organization like NIST.
Thanks for the insight, and thanks for the effort you put in trying to build a network !
I was aware that building your own fiber network can be extremely expensive (though there are tales of villages doing it, such as this one [0], because noone is willing to come to them. Yes, that case is special in that it's extremely rural), but I didn't know that you couldn't even use telephone lines because they're unreliable.
Comments
Assuming your talking about the US with your question (since the article is), than in the US it really is _that_ expensive.
The main reason it's "easy" in France (and UK, and other places) is that the big name's network are required by law to let you use the end point. That exists in the US on the phone lines (legacy telephone regulations). But not on anything else (no cable or fiber).
So unless you want to pour all of your time and energy into offering flaky, slow, unreliable internet over 100-year-old copper lines, with speeds below 6mb and minor outages daily, your out of luck re-using any lines in the US.
That difference alone suddenly makes it not remotely easy or cheap.
To deploy internet at a reasonable price, you have to start an entire fiber construction company, and local political lobbying firm, simultaneously, just to get a single customer.
We're talking a barrier to entry of at least 1 to 2 million dollars, to wire a neighborhood, just to get your first customer signed up. Customers who've been conditioned to pay no more than $80/month for internet access.
That cost is actually not expensive compared to typical municipal projects. But it's way more than you'll ever get funding for, to chase, unless you are already extremely wealthy from some other business / family history / etc.
Venture firms would rather fund one thousand new Instagrams than pour money into anything remotely resembling a utility business.
The main way around that cost barrier is to deploy "bad" technology, like wireless. That can be done on a shoestring budget. It works really well in places where there is zero density (rural, exurbs). But it's not the future, it's legacy deployment, with legacy problems and legacy speeds. The nicest microwave radios you can buy only barely match the speeds and reliability of the cheapest fiber deployment.
So, if your in any sort of reasonably-sized city, you'd be stupid to not roll out fiber from the start. And to do that, your looking at millions of dollars of cost.
No one will ever fund that. (I've tried -- countless times). Typically, only selfie-sharing iPhone apps can generate investment like that.
This is why most people in the US are pushing for "common carrier". Because no one here believes a honest business will ever deploy a modern network on any large scale, so the best we can hope for is an upgrade from 100-year-old phone lines, to 40-year-old cable lines.
The main reason it's "easy" in France (and UK, and other places) is that the big name's network are required by law to let you use the end point. That exists in the US on the phone lines (legacy telephone regulations). But not on anything else (no cable or fiber).
The regulations are on the books, and through the mid-2000s UNE-P was a somewhat-viable business model. However, the ILECs figured out that they can sabotage the fulfillment process with no repercussions. When your new customers can't get their service actually started, and the only thing you can tell them is "Verizon is being mean to us", you're not long for the business. That's why the only remaining CLECs are "facilities-based", i.e. they don't use the UNE-P program.
This is why I'm not so hopeful for a similar program on the data side. I've seen the kind of bullshit the incumbents pull, and I greatly doubt the FCC will ever prevent that. I don't think investors are dumb enough to fall for the 1996 bait-and-switch again, so any effective program like this would have to include strong structural protections for competitive ISPs. My understanding is that the most effective such programs in Europe have done that, e.g. forbidding the network owner from selling to retail customers.
Yes, that's very true.
Sabotaging fulfillment is yet another problem in re-using lines, which is another reason folks aren't in a rush to reuse lines in the US.
The main way around that cost barrier is to deploy "bad" technology, like wireless. That can be done on a shoestring budget. It works really well in places where there is zero density (rural, exurbs). But it's not the future, it's legacy deployment, with legacy problems and legacy speeds.
That's the current state of the art, but maintaining that decrepitude has taken a decade of FCC foot-dragging on white spaces and other "cognitive" radio applications. We are nowhere near physical limits of spectrum bandwidth, and radio engineers of the future will laugh at our aversion to innovation in this space.
I think the best thing that could happen for competition and choice in data connection would be for radio regulation to be taken away from FCC and given to a less political organization like NIST.
Thanks for the insight, and thanks for the effort you put in trying to build a network !
I was aware that building your own fiber network can be extremely expensive (though there are tales of villages doing it, such as this one [0], because noone is willing to come to them. Yes, that case is special in that it's extremely rural), but I didn't know that you couldn't even use telephone lines because they're unreliable.
[0] http://motherboard.vice.com/read/this-rural-community-is-bui...