Hacker News Reader

Comment by mikekij | Hacker News Reader

As impressive as the number of Starlink satellites is, the service has become unusable for me in Southern California, particularly for video calls. I’ve been told that the bottleneck is land stations, not satellites. Either way, I’ve had to cancel my Starlink service both at home and work.

View on HN

It's not really a service meant for SoCal, but I'm sure they see it as a valuable stress test.

Every time Starlink comes up it’s the same argument. “It’s not meant for high population density areas”

Without the “high population density areas” it is simply not economically viable

Based on what? What user count do you think Starlink needs to be economically viable?

Napkin math time because the economics for satellites are not intuitive compared to stuff where density is better.

10% of the surface of the earth is inhabited by people. 510 million square kilometers * .1 is 51 million square kilometers. At a service density of only one user per square kilometer, that’s 51 million users for regular service.

Setting that aside, do you care to guess how much an oil rig in the North Sea will pay for Internet? And that user will put roughly the same load as Ted in SoCal will on the network.

Satellite Internet is not for people in cities, period. Starlink, viasat, whatever.

I have no idea, but this company has subsea cables across many North Sea oil rigs. Maybe they are laid as part of the pipelines.

https://www.tampnet.com/oil-gas

I agree with you, and your 2nd point about the b2b areas being important.

However I think the general math you do needs a bit more nuance. You'd want to exclude all people living in cities so you'd need to exclude the land mass taken up by cities to get a more accurate estimate.

Also I'm not sure where you get the 10% number from for people inhabiting the world. From this report they talk about humans impacting 15% of the landmass https://www.visualcapitalist.com/mapped-human-impact-on-the-...

Again, I agree with your point but I'm skeptical about your back of the napkin math.

> You'd want to exclude all people living in cities so you'd need to exclude the land mass taken up by cities to get a more accurate estimate.

No you wouldn’t. The entire point is that Starlink works fine in cities for a small count of users.

it's probably economically viable without selling to consumers at all. if they get planes boats and military, that on it's own would be plenty to sustain it

I doubt it. The ongoing operational costs of a mega constellation like this is huge and I doubt they will be able to sell their services to maritime, aviation and military sectors at a much higher price than existing operators.

They will also meet hefty competition the next few years from Amazon Kuiper, Oneweb and I assume existing operators are scrambling to meet the competition. Interesting times.

Most operating costs estimates assume that sending up more satellites is an expensive proposition.

For starlink, its not. Maybe for the first time in history.

They already sell it to maritime and aviation sectors at significantly less than existing operators, though significantly more than the regular service.

But at this point in their plan it doesn't matter if they undercut existing providers. If the marginal cost of the extra client is small enough then it is just extra revenue. And you get the clients operations dependant on that extra bandwidth that they won't want to give up.

> Without the “high population density areas” it is simply not economically viable

That's their headache, and does not negate the previous point.

But it doesn't necessarily need high subscriber rates in any one place to make lots of money. If you just focus on those high population areas, 1.7 billion people live in cities. If you can sell to a few in a thousand then you will have several million subscribers across the world and they will likely be wealthier people. And then there are cruise ships, airliners, oil rigs, ski chalets, military compounds, backhauling WiFi for musical festivals, and millions of people living in the woods with rubbish DSL.

Furthermore, a lot of people seem to think the world is divided into Manhattans and log cabins in the woods someplace. In fact, a huge number in the US and presumably elsewhere live nearby small cities etc., have electricity and maybe other utilities, but don't have what most people would consider decent Internet access.

Yea, there are hundreds of thousands of tiny tiny towns all across the US with terrible internet service. You could run a perfectly good ISP for a village of a few hundred people off of one Starlink dish.

Yeah, I guess "it's not meant for high population density areas" argument would work if rurals didn't also have garbage service on Starlink.

I have used it regularly in a rural area, it works amazingly better than what was available prior, and keeps improving as more satellites are launched.

If nothing other than 1 Gbit fiber will satisfy you I guess Starlink isn't that great. But, honestly, my wired broadband 50 miles outside of one of the largest cities in the Northeast isn't 100% reliable and near-infinitely fast either.

As in your case, my brother's house basically didn't really have Internet before though was finally get able to get a Verizon hotspot which "worked" but you couldn't really stream video, for example, because of data costs. Now, it's completely practical to work from there as well as have access to Internet entertainment etc. options which simply wasn't the case before.

compared to what?

pigeons

SpaceX have sought approval for 12,000 satellites, with a possible extension to 42,000. Right now they have less than 4,000. I think they have substantially fewer satellites than they believe they require, and the service is probably doomed in the long run if they can't get Starship flying.

Reply on news.ycombinator.com