4 ms·

Jared Mauch didn’t have good broadband–so he built his own fiber ISP

31 points by metabagel 6y ago

xenadu02 6y ago
What amazes me is how cheaply he built this. Less than $200k with an expected break-even of 42 months. Think about that for a minute... ATT, Comcast, et al can't be bothered to invest in infrastructure that would pay itself back in 3-4 years. Remind me why we need these companies again? They seem to be failing at capitalism.
- spideymans 6y ago
  >What amazes me is how cheaply he built this. Less than $200k with an expected break-even of 42 months. Small town governments take note.
- fat_pikachu 6y ago
  Judging from the article, he built the network over the course of four years, presumably without paying himself. Even if you think a major ISP could shave a year off this timeline, it would be a much larger investment than 200K.
dogline 6y ago
This is _exactly_ was I was thinking of organizing for a small rural forest town in N. California. I'd become a small ISP for the town. Maybe I could get a grant. I wasn't looking forward to customer complaints, but I think I could have done good for the town. However, with Starlink almost imminent, this sort of capital outlay and risk doesn't seem like a good idea. Starlink, ideally, will give reasonable bandwidth with little physical infrastructure, to places which have no other service or cell other than Hugesnet.
syntaxing 6y ago
Wouldn’t a WISP (wireless ISP) be better than a fiber ISP? Or are WISP not ideal for places with colder weather like Michigan as this article?
- bcrl 6y ago
  WISPs are usually bottlenecked by the amount of unlicensed spectrum available. Geography can also makes operating a WISP difficult. Hilly and heavily treed terrain is poorly served by wireless, as are high density residential neighbourhoods. As one academic put it: investment in wireless merely delays investment in fibre.
notyourday 6y ago
Jared Mauch is not a rando. He has gobs of experience running ops for actual ISPs including backbone engineering for NTT. Is it something that those without such experience can do? Certainly. Would it be relatively "easy" for them? Not anywhere close.

31 points by metabagel 6y ago

What amazes me is how cheaply he built this. Less than $200k with an expected break-even of 42 months. Think about that for a minute... ATT, Comcast, et al can't be bothered to invest in infrastructure that would pay itself back in 3-4 years. Remind me why we need these companies again? They seem to be failing at capitalism.

spideymans 6y ago
>What amazes me is how cheaply he built this. Less than $200k with an expected break-even of 42 months. Small town governments take note.
fat_pikachu 6y ago
Judging from the article, he built the network over the course of four years, presumably without paying himself. Even if you think a major ISP could shave a year off this timeline, it would be a much larger investment than 200K.

dogline 6y ago

This is _exactly_ was I was thinking of organizing for a small rural forest town in N. California. I'd become a small ISP for the town. Maybe I could get a grant. I wasn't looking forward to customer complaints, but I think I could have done good for the town. However, with Starlink almost imminent, this sort of capital outlay and risk doesn't seem like a good idea. Starlink, ideally, will give reasonable bandwidth with little physical infrastructure, to places which have no other service or cell other than Hugesnet.

syntaxing 6y ago

Wouldn’t a WISP (wireless ISP) be better than a fiber ISP? Or are WISP not ideal for places with colder weather like Michigan as this article?

bcrl 6y ago
WISPs are usually bottlenecked by the amount of unlicensed spectrum available. Geography can also makes operating a WISP difficult. Hilly and heavily treed terrain is poorly served by wireless, as are high density residential neighbourhoods. As one academic put it: investment in wireless merely delays investment in fibre.

notyourday 6y ago

Jared Mauch is not a rando. He has gobs of experience running ops for actual ISPs including backbone engineering for NTT. Is it something that those without such experience can do? Certainly. Would it be relatively "easy" for them? Not anywhere close.