Hacker News Reader

Comment by dougmwne | Hacker News Reader

This is already quite an achievement! In just 10 launches, SpaceX is operating the world's largest satellite constellation and can already provide broadband of reasonable quality to some customers. In rural areas, I have struggled to catch weak LTE signals with fixed antennas that often didn't hit these speed and latency numbers. This absolutely seems on track to be an excellent alternative to rural LTE or existing satellite internet providers. If it gets better from here with more launches, it's all gravy as far as I'm concerned.

View on HN

I would like to see a kit from SpaceX which provides a solar powered Starlink to LTE bridge. Such kits could be installed anywhere, and provide service to places such as farm communities (install them on top of grain silos, for example). That way Starlink could be available to more customers without them having to have their own premise equipment (beyond an LTE hotspot).

I'd sort of expect LTE operators to pay SpaceX for kit to do literally this.

I worked for the Telco that services all of Northern Canada (world's largest operating area for a telco).

We installed one of the world's first 3G towers that uses only satellite as backbone. Think fly-in only communities that are thousands of kilometers away from anything in the Arctic. Many of our sites are served by radio backhaul for that reason.

The sat backhaul tower is extremely expensive and temperamental. It does work, but not very well. 3G (and LTE) are not very fault tolerant, and phones drop calls when the quality of the call drops below thresholds that can't be set. AFAIK they have not installed another one because of all the problems.

For data the cost is astronomical - many thousands of dollars per GB. Hopefully starlink can at least fix that.

This is what turns me away from the astronomers complaining about the constellations. They’re not interested at all in the community whose problems satnet helps resolve. Society has had plenty of time to run data lines out to the hardest to reach areas and clearly cannot and will not do it any other way.

I wouldn’t blame the astronomers for that. People who lead different lives get affected by technology in different ways. It’s good for everyone to tell the world how they are being affected. Ultimately people in positions of power are going to make decisions and it’s good for them (and activists, lobbyists, and advisors who they talk to) to have all the relevant information. It’s not necessary or possible for everyone to know how everyone else is affected before they speak about their own concerns.

And in 10 years, Starship should make it possible to put more space observatories in orbit, so it should get better for the professional astronomers eventually. It'll still suck for amateur astronomers and photographers, though.

Although, with the way 2020 is going, we'll probably miss observing the asteroid that kills us all because a Starlink got in the way of the observation.

The reason we don't use space observatories as much anymore is because we have adaptive optics now and don't need them. The concerns of astronomers are real. It's not for Starlink to ruin our ability to observe the sky for all of humanity. Incredible entitlement.

You can easily turn that last part around.

It's not for astronomers to ruin our ability to provide essential (and yes, internet is essential at this point) services for billions of people around the world that otherwise can't get it. Incredible entitlement.

It just depends on what you value more, observing the sky, or providing internet.

What right do you have to turn that statement around? Incredible entitlement. /s

But really adaptive optics is nice, and lets us compensate for atmospheric distortions. But not light filtered by the atmosphere. Or clouds. Or daylight. And we still have to compensate for quakes and vibrations, and temperature change. It's simply just cheaper to do on land than in space.

Imagine if it wasn't an order of magnitude more expensive in space. Or imagine a large telescope array on the Moon, where you get stability, ability to repair, and no atmosphere to get in the way. With SLS and Starship, all of that starts to look possible in the next 20 years.

Seems like dealing with satellites is a truly trivial application of "adaptive" optics.

You can't possibly expect me to believe that you have the processing power and technical know-how to remove the influence of atmospheric distortion, but that you're helpless when confronted with discrete fast-moving objects for which you have ephemerides. Sorry, that dog just don't hunt.

Initial versions of Starlink would reflect so much light that they would overwhelm significant areas of the sensors. If the reflection issue is resolved with the new sun shades, then the satellite trails can be removed during post-processing.

They can also use a shutter, for the cases where postprocessing really isn't an adequate answer.

The more light-sensitive the observation is, the longer they're integrating, ergo the less harm that a fast transient satellite pass will do.

I thought the main reasons were cost and maintainability? Both of which should be improved with the ability to launch larger, heavier, and more numerous payloads.

> It'll still suck for amateur astronomers and photographers

spaceX could just sell time on satellites as a service, and anyone (amateur or professional) can just buy and use that time to photograph anything in super high fidelity and with no atmospheric interference.

Funny that astronomers and space types were all about setting aside other people's problems when it came to doing cool stuff in space, e.g. the famous NASA reply to Sister Mary Jucunda[0]. But now, it's "won't somebody please think of the astronomers"

[0] https://lettersofnote.com/2012/08/06/why-explore-space/

I'd like to hear some of your install stories.

> many thousands of dollars per GB

Which year was that? Depending on airline, I was able to use free to $20 per long haul leg sat net and stream 4K YouTube...

I imagine they mean GB/s of capacity, not per GB.

This was back in ~2014, though last I looked into it significant amounts of data from a sat is still many thousands of dollars per GB. (NOT GB/s)

Home based satellite internet wouldnt be viable if it cost thousands of dollars per GB. There are low-volume satellite services that do have a pretty high charge per megabyte, but thats because they are intended for applications that only need a few dozen KB per day or so.

Fwiw I don't buy this. A satellite specced for a gig and being used at say 1/16 capacity (to account for orbit) would make say 100 dollars a second. That's nearly 9 million a day, no way do satellite providers make that much, or there would be a lot more supply.

How could you possibly tell that they're able to provide reasonable quality to customers by looking at a single speed test? There is a whole bunch of other stuff that goes into running a network other than doing a single speed test. A lot of the things that people complain about other satellite providers fall into that category.

Australian (Sydney, 30 minutes train from CBD) here, I pay $60/m for 45mbps but my line potential is 35mbps. My average speed is around 20mbps (just ran a speedtest and it was 22mbps).

Some places have 100mbps, but most don't live in those areas. I am lucky to get the speeds that I do, honestly.

I'm watching both 5g and starlink with great interest as they will be up and running before the next infrastructure initiative by the government.

Maybe Sydney is actually rural and we don't know it yet

Unfortunately, I doubt Starlink will be generally available that close to Sydney CBD. The population density would be too high.

Maybe Sydney’s rural fringe? (Dural? Glenorie? Maroota? Wiseman’s Ferry? Camden? Rural parts of Central Coast like Holgate, Lisarow, Somersby?)

Parts of australia are too rural for starlink... Without a ground station within ~500km you won't be getting service. That'll be the case right in the middle of australia I'd guess.

Yea, I’m currently on a mix of cellular and traditional satellite. The combo is fairly expensive but necessary because one or the other goes out on occasion. It’s fine, but ideally you don’t want to ever think about your internet. You just want it to work. No data cap, no noticeable latency.

Reply on news.ycombinator.com