Hacker News Reader

Comment by TDL | Hacker News Reader

Laying cable isn't the only way to deliver internet service. Wireless ISPs are a thing.

Wireless spectrum is limited and either exclusively owned or subject to congestion. Dedicated ine-of-sight direct links are possible and avoid some broadcast problems, but still have property rights issues (they manifest a little differently, since you can smoke a connection without establishing property rights to the whole route, but then you are at risk of someone on the route interposing a temporary or permanent obstruction at any time.)

The existence of wireless connections does not change the fact that broadband last-mile service is a natural monopoly.

That isn’t realistic in this decade for rural/remote areas. Satellite internet also failed because of the unusable latency incurred. The only way to guarantee bandwidth is a physical connection because spectrum bandwidth is finite, whereas point-to-point bandwidth scales nearly linearly with more physical channels.

> That isn’t realistic in this decade for rural/remote areas.

Not OP, but I recently lived in a rural area and had 3 options for Internet:

  * Mediacom cable with permanent 20-50% (or more) packet loss

  * CenturyLink DSL at 1mbps (not a typo)

  * Local wireless company, 7mbps, no packet loss

About 2 decades ago I worked for a different wireless ISP (same rural area) that delivered the best local service by FAR. (Cable wasn't yet an option, and satellite had crazy latency.) So wireless was superior ~2 decades ago in that rural area as well as recently, for many people. So I don't know what you mean by wireless being "[unrealistic] in this decade for rural/remote areas."

My position is that municipal fiber is the way to go, but give wireless ISPs credit because they're helping a lot of people get on the Internet who would barely have connectivity otherwise.

They still need access to a (relatively) local backbone. If they started collecting any significant market share from the local mono- and duopolist providers, I suspect that they would simply be banned from those backbones, which are (mostly) maintained by those same providers.

Yes, they're a thing. A shitty thing.

Wireless ISPs would immediately collapse if they had to serve the amount of Netflix traffic wired internet connections use in an evening. They simply aren't a viable alternative for the population at large.

You might want to check out Webpass, which is now owned by Google Fiber. They can and do deliver high speed internet. Webpass specifically targets dense population centers.

Dense population centers are the easiest to cover and already have the most choices. Furthermore, Google Fiber has a history, like most Google municipal projects, of failing to finish and then languishing. They never finished deploying to Starbucks, and what was deployed was still slow. Also, Mountain View Google WiFi, a wireless repeater mesh, was horrible, useless and poorly maintained.

>Webpass specifically targets dense population centers. That, in itself, is the problem. Nobody competes in rural areas.

> Webpass specifically targets dense population centers.

The solution nobody needs, effectively. It's already economically sound to run cable/fiber in dense population centers, because the number of customers you reach with every mile of new cabling is huge.

Don't get me wrong, Webpass is very cool. But it's point-to-point and requires you to build a lot of infrastructure on the top of buildings, which requires a lot of permissions, which isn't always easy to get. The wireless ISPs that do exist outside of urban areas have to use cellphone networks, and they are not ready for the amount of data transfer required. Yet.

And most of them suck.

Reply on news.ycombinator.com