3 ms·

I'd like to see this as well. I suspect the money would be better spent subsidizing/establishing community-owned ISPs in underserved areas. There are still ol

by hedora 3y ago

I'd like to see this as well. I suspect the money would be better spent subsidizing/establishing community-owned ISPs in underserved areas. There are still old telephone co-ops from the early days of the telephone in the US. The one I'm thinking of is extremely rural (population density is < 2 people per square mile), with a median county household income under $25K. They put in fiber to the home years ago, and 1GBit symmetric is well under $100 a month. (They have cheaper plans too.) Similarly, some places have started using companies like this: http://nextlevelnetworks.ca http://nextlevelnetworks.ca You pay them to build out a local fiber network. You just pay the line installation, and then the community owns the network. The minimum project size is about 100 houses because that leads to a $100/month per house bill from the incumbent network's $10,000/month fiber uplink. Concretely, imagine there was a subsidy for any area that doesn't have 1GBit symmetric for under $100 a month. Starting in the poorest communities, use the ACP money to establish a local co-op (per community, creating jobs) that charges 1.1x the cost of the fiber uplink per month, and saves 10% for network expansion / maintenance. The ACP burned through $14.2B since 2021. It costs about $10K to run fiber to a house (it's almost all labor, so in urban areas you have short runs with high cost of living for the workers, and in rural areas you have long unobstructed runs with low cost of living). They could have wired 1.4M low income houses for fiber with that money, and the co-ops could then charge the houses well under $100 per month for connectivity.

cowpig 3y ago
This thought process makes some sense to me, though I have doubts about the implementation. Is fiber really the best fit for rural areas? What about some kind of wireless solution? Surely that would be more cost-effective?
- hedora 3y ago
  Wireless could also work, but it requires line of sight, and you have to put it on a roof / poll, which isn't always feasible. Another subtle problem with rural wireless is that you have to point to a well-connected area, so the cheapest way to service houses that are on the side of the hill facing away from town is often either fiber or putting the tower in orbit. Otherwise, you'd need to run a fiber trunk line to the middle of nowhere to support the ISP side of the wireless link. Fiber has the property that you can run it in parallel to existing power lines, so any house with power already has an existing right of way to piggyback on (owned by the wrong person, but imminent domain exists). Also, you can choose to trench the fiber or run it on telephone polls. If I were running things, I'd look into strategically deploying fiber-to-the-home to areas with good line of sight to poorly connected areas, and then use a hub and spoke network topology where the spokes are directional wireless and some of the homes are hubs. That creates a second order problem though: You need backup power at the hubs or parts of the network will go down when the electricity goes out. Long term, it's probably cheaper to maintain a network of buried fiber than it is to maintain the electricity infrastructure needed to keep the towers online.
- tallowen 3y ago
  My experience living in Montana has been that wireless internet providers can be great but they don’t provide great low latency service. It creates a perception that one can’t live in certain areas and do remote work since it’s too hard to do video. When it comes to broadband there is a question of “how good is good enough” but for the people living in Hamilton Montana (population 5k) I hope there can be a path to infrastructure investment that leads to high quality internet in town.
  - hedora 3y ago
    Wireless internet should be low latency (unless you're talking about geosynchronous satellite). During high demand, our local WISP has 1 second latency to the internet because their wired backhaul link is terrible. It's less than 1ms to the tower. Starlink to the internet is in the 20-30ms range, and it's bouncing into orbit and back.
- hn_acker 3y ago
  In the long run, fiber is the cheapest [1] and the best technological fit in rural and non-rural areas. Satellite internet is unreliable in storms due to attenuation, has decreasing bandwidth as more users sign up (which is one reason Starlink is expensive), and creates space junk down the line. Wireless solutions on land via 5G and the like require wires (will become mostly fiber) to carry the internet traffic to intermediate transmitters. 5G also needs more cells [2] per unit area compared to 4G to compensate for attenuation. [1] https://www.eff.org/deeplinks/2020/06/why-slow-networks-really-cost-more-fiber https://www.eff.org/deeplinks/2020/06/why-slow-networks-real... [2] https://www.lifewire.com/5g-cell-towers-4584192 https://www.lifewire.com/5g-cell-towers-4584192
hedora 3y ago
I guess I spent a lot of words saying: The ACP tries to make an expensive thing less expensive by artificially increasing demand. What if we tried increasing supply instead?
- rtkwe 3y ago
  Is it really something that has the same supply and demand curves though? Once the wires are run and especially with fiber the ability to add another customer is pretty cheap. The main goal from what I've read is to encourage building out into poorer areas by boosting the customer base in areas where it might otherwise be uneconomical to build into. Most things diverge from the econ 101 model significantly.
  - _heimdall 3y ago
    I expect other parts of the business are more susceptible to the usually supply constraints, like customer service and maintenance.
    - smegger001 3y ago
      What customer service? this is Comcast we are talking about. Consistantly ranked in the top ten worst companies in custom service nation wide. And maintaince already showed up only if the felt like when they felt and blamed your equipment when they do.
      - _heimdall 3y ago
        Quality of service and their budget for service aren't quite the same. I worked at a major US telco for a while, their service wasn't anything to write home about either but you better believe they had very specific quotas of how many people they had to hire or contract relative to their customer count. At least in the US, union contracts often get extremely specific on that kind of thing.
mcmcmc 3y ago
Co-ops are fantastic for utilities, over half the landmass of the US is serviced by rural electric co-ops under the NRECA established by FDR during the Great Depression. Internet connectivity should absolutely be treated as a public utility similar to electricity.
- Triphibian 3y ago
  I have outstanding fiber-to-the-door internet provided through my rural electricity co-op. I have no doubt that we'd still be waiting for internet if it was up to Verizon to supply it.
genewitch 3y ago
It's the USDA that provides grants for small community ISPs but their rules are clownshoes. your service area can't overlap any other service area, regardless of how awful it is. Well, technically, that's not true. If there's 1.5mbit in the area, they won't fund you - and it's not 1.5mbit in practice, it's 1.5mbit advertised. I went around and around with the government saying that at&t obviously doesn't want to maintain this network so why not get something up before they let the copper rot in the ground. Nothing doing.
- Sabinus 3y ago
  >your service area can't overlap any other service area, regardless of how awful it is That's regulatory capture by the majors. They don't want actual competition to reduce the value of their 'investments' so they avoid competing with each other, and smother all other parties in legislation and lobbying.

by hedora 3y ago

cowpig 3y ago

This thought process makes some sense to me, though I have doubts about the implementation. Is fiber really the best fit for rural areas? What about some kind of wireless solution? Surely that would be more cost-effective?

hedora 3y ago
Wireless could also work, but it requires line of sight, and you have to put it on a roof / poll, which isn't always feasible. Another subtle problem with rural wireless is that you have to point to a well-connected area, so the cheapest way to service houses that are on the side of the hill facing away from town is often either fiber or putting the tower in orbit. Otherwise, you'd need to run a fiber trunk line to the middle of nowhere to support the ISP side of the wireless link. Fiber has the property that you can run it in parallel to existing power lines, so any house with power already has an existing right of way to piggyback on (owned by the wrong person, but imminent domain exists). Also, you can choose to trench the fiber or run it on telephone polls. If I were running things, I'd look into strategically deploying fiber-to-the-home to areas with good line of sight to poorly connected areas, and then use a hub and spoke network topology where the spokes are directional wireless and some of the homes are hubs. That creates a second order problem though: You need backup power at the hubs or parts of the network will go down when the electricity goes out. Long term, it's probably cheaper to maintain a network of buried fiber than it is to maintain the electricity infrastructure needed to keep the towers online.
tallowen 3y ago
My experience living in Montana has been that wireless internet providers can be great but they don’t provide great low latency service. It creates a perception that one can’t live in certain areas and do remote work since it’s too hard to do video. When it comes to broadband there is a question of “how good is good enough” but for the people living in Hamilton Montana (population 5k) I hope there can be a path to infrastructure investment that leads to high quality internet in town.
- hedora 3y ago
  Wireless internet should be low latency (unless you're talking about geosynchronous satellite). During high demand, our local WISP has 1 second latency to the internet because their wired backhaul link is terrible. It's less than 1ms to the tower. Starlink to the internet is in the 20-30ms range, and it's bouncing into orbit and back.
hn_acker 3y ago
In the long run, fiber is the cheapest [1] and the best technological fit in rural and non-rural areas. Satellite internet is unreliable in storms due to attenuation, has decreasing bandwidth as more users sign up (which is one reason Starlink is expensive), and creates space junk down the line. Wireless solutions on land via 5G and the like require wires (will become mostly fiber) to carry the internet traffic to intermediate transmitters. 5G also needs more cells [2] per unit area compared to 4G to compensate for attenuation. [1] https://www.eff.org/deeplinks/2020/06/why-slow-networks-really-cost-more-fiber https://www.eff.org/deeplinks/2020/06/why-slow-networks-real... [2] https://www.lifewire.com/5g-cell-towers-4584192 https://www.lifewire.com/5g-cell-towers-4584192

hedora 3y ago

I guess I spent a lot of words saying: The ACP tries to make an expensive thing less expensive by artificially increasing demand. What if we tried increasing supply instead?

rtkwe 3y ago
Is it really something that has the same supply and demand curves though? Once the wires are run and especially with fiber the ability to add another customer is pretty cheap. The main goal from what I've read is to encourage building out into poorer areas by boosting the customer base in areas where it might otherwise be uneconomical to build into. Most things diverge from the econ 101 model significantly.
- _heimdall 3y ago
  I expect other parts of the business are more susceptible to the usually supply constraints, like customer service and maintenance.
  - smegger001 3y ago
    What customer service? this is Comcast we are talking about. Consistantly ranked in the top ten worst companies in custom service nation wide. And maintaince already showed up only if the felt like when they felt and blamed your equipment when they do.
    - _heimdall 3y ago
      Quality of service and their budget for service aren't quite the same. I worked at a major US telco for a while, their service wasn't anything to write home about either but you better believe they had very specific quotas of how many people they had to hire or contract relative to their customer count. At least in the US, union contracts often get extremely specific on that kind of thing.

mcmcmc 3y ago

Co-ops are fantastic for utilities, over half the landmass of the US is serviced by rural electric co-ops under the NRECA established by FDR during the Great Depression. Internet connectivity should absolutely be treated as a public utility similar to electricity.

Triphibian 3y ago
I have outstanding fiber-to-the-door internet provided through my rural electricity co-op. I have no doubt that we'd still be waiting for internet if it was up to Verizon to supply it.

genewitch 3y ago

It's the USDA that provides grants for small community ISPs but their rules are clownshoes. your service area can't overlap any other service area, regardless of how awful it is. Well, technically, that's not true. If there's 1.5mbit in the area, they won't fund you - and it's not 1.5mbit in practice, it's 1.5mbit advertised. I went around and around with the government saying that at&t obviously doesn't want to maintain this network so why not get something up before they let the copper rot in the ground. Nothing doing.

Sabinus 3y ago
>your service area can't overlap any other service area, regardless of how awful it is That's regulatory capture by the majors. They don't want actual competition to reduce the value of their 'investments' so they avoid competing with each other, and smother all other parties in legislation and lobbying.