[greesil]

Yeah but you don't need 10000 satellites to provide cell coverage when 100 would do. Hybrid ground station networks have been a thing for a long time now... if your goal is to provide mobile networks for the lowest possible cost to the end user. That's not Starlink's goal -- these "uninternetnetted" folks don't have money to pay, they typically don't even have electricity. Rural broadband in developed countries do. Telematics does. So don't give me this bs of efficiency for the whole world.

Edit: to be more explicit about the architecture, you can run 100 satellites with better/ larger / more antennas that'll do the work of 1000. SpaceX has a different model, which is more low quality, disposable satellites because they have an inexpensive launch vehicle. Ultimately their system of 1000 might cost the same or less than the 100 high performance satellites. But I can't figure out why the heck they want 10000 satellites unless they want to just have phones communicate directly to orbit, no starlink base station as middlemen. So now they've clogged up LEO so they can replicate what terrestrial networks provide, AND there's a perfectly viable alternative architecture that didn't require nearly as many satellites. That's my gripe.

[inemesitaffia]

Someone doesn't know how orbits work, how big space is, or how expensive other satellite services used to be pre starlink. Or how bad it can get with latency.

And yes the unreachable people have family in reachable areas that can pay, sometimes the Internet is what keeps them away, since solar and batteries exist

See the Marubo for example

[greesil]

You mean me, the aerospace engineer, who has published in this field, doesn't know how satellite constellations, link budgets, astrodynamics, the orbital environment, the systems engineering, and the other minutiae of this, doesn't understand the tradeoffs in building these things? What do you know, in a professional capacity? Tell me you're in this field, or piss off.

Tell me how big is space. Like, is it as big as the back of a giant tortoise upon which the flat earth moves? Is it bigger than a bread box, a telephone box, or a TARDIS? Wat is it???

[inemesitaffia]

I'll ask you an easy question.

There's another LEO provider with around the number of sats you recommend. What's the provider and how much do they charge for fixed use in the USA?

I worked at a cellular Telco serving customers with an ARPU under $20. Also worked at a satellite tech reseller.

These unreachable poor people were my bread and butter.

Your statements show ignorance about the markets, space and other minutae. Piss off.

How much would these better antennas end up costing? Is Rajeev Badyal an idiot?

Frankly unless you're thinking B2B then cell rather than B2C your choices are worse, not better. We have o3b and it's way better than starlink for stability. But there's several $$$$$$ big reasons I can't pop out to a shop and buy a receiver.

[greesil]

It's not about existing competitors, it's about the optimal constellation design that (a) serves the customers needs and (b) doesn't create a hazardous LEO environment after one accident or ASAT event happens. I don't think anyone's seriously tried to serve the market that Starlink does at LEO. Iridium and Globalstar targeted low bit rate low power terminals, they never tried for home internet. This market could be served by a constellation of similar size as these legacy providers if it was designed to do such. But nobody likes taking risks other than Elon I guess, so here we are.

[inemesitaffia]

I don't know about constellation size issues wrt to customer needs. But I presume one important factor was making sure the terminal costs were lower than a mid-range phone. Especially since Kuiper claims to be focused on that. It was an issue with Teledesic. Cheap sats also mean treating them like cattle rather than pets and that the on orbit spares are one orbit away from replacement

Lasers mean low bandwidth usage areas don't need ground stations.

Being in VLEO means most talk about Kessler is just regurgitated claims by people who don't haven't bothered to read his papers and don't know the time taken to clear orbits by debris. Some of them don't even realize that if you change the apogee via collision the perigee changes too.

I don't know if it's possible to serve with the same number at this altitude, but I can see even Kuiper is going for over 5k sats and OneWeb is selling antennas for 9k (down from 25k) and monthly service for fixed users in the US is 2500/m. I remember congestion issues with less sats (I wonder why they use nco=1 when Kuiper is 4)

As for the risk, people are already claiming the (dubious) success was already self evidently inevitable even when we can't see evidence of profits and that Elon's done nothing special.

Not even the struggles of Iris², Rivada, OneWeb, Kuiper, and pretty much every launch provider will convince them though.

I'm looking forward to what competitors bring and I'm willing to bet I won't be able to buy a terminal for less than $0.5k (unsubsidized) and service for less than $50/m in Africa from another provider before 2028 (Artemis). But I'm hopeful I'm wrong. I was paying ~$250/m for 4G. Now lower via devaluation but Starlink is $23/m here

[greesil]

I'm haven't gone deep into Kessler syndrome but the guy himself thought we were already at a tipping point more than a decade ago. In hindsight I think the model needed updating, but adding tens of thousands of new satellites seems like pushing our luck. Maybe SpaceX sees this as an advantage. When you've got the least expensive launch system, you can afford to launch armored satellites and launch vehicles.

Geez I forgot about Kuiper. Good god what are they doing.

If you're concerned about terminal cost, the whole point of ATC is that someone puts up a cell tower with a LEO satellite backbone, and people can use cheapo $20 mobile phones.

Anyway, developed countries' rural broadband and military contracts are where it's at. I'm sure Starlink -could- make a profit if it isn't doing so already.