Beta users of Starlink get downloads of 11 to 60 Mbps | Hacker News Reader

Beta users of Starlink get downloads of 11 to 60 Mbps

arstechnica.com

461 pointstrulyrandom590 comments

Anyone complaining about this has never experienced the frustration of rural living when it comes to internet. This is absolutely amazing, and it will just get better.

Even a reliable 10mbps would be a godsend in many places.

Honestly, one of the barriers to my moving to a more rural area and working remotely was broadband access, but Starlink could very well solve that problem...

You're not alone there. I wonder if starlink will increase property values in rural areas.

Especially with COVID showing the downsides to living in urban areas.

I'd rather live in an urban area during a pandemic. It's counter-intuitive but the access to testing and hospitals is very important.

https://theconversation.com/rural-america-is-more-vulnerable...

Larger amounts of people mean more chances of exposure.

Being in urban environment means higher dependence on fragile supply chains- food security is totally dependent on JIT grocery store inventories.

Rural homes, by virtue of being isolated, are easier to self isolate.

However, if you’re able to work from home, you can self-isolate pretty much anywhere, and drive wherever you need to get food.

If urban environments start seriously breaking down due to a food supply chain problem, rural areas won’t be safe from desperate people.

Anyway, both approaches have their pros and cons, but as someone who has a fairly severe back problem and not much inclination to try to live off the grid, I’d rather be in an urban area.

A friend of mine who served in Afghanistan said that the Afghanis living in the mountains had no idea America had invaded and no idea about a war or 9/11. At first they had thought the Americans were Russians. I think you over estimate how big America is and how remote some places can be.

I've traveled through some pretty remote places in both hemispheres, and the US has nothing on isolation compared to the mountains in Central Asia. Remoteness won't really protect people here, not like it would there.

> rural areas won't be safe from desperate people

At least for America, most city folk don't seem to understand just how armed their rural counterparts are. They'd be pretty darn safe.

I don't know if rural folks may realize how well-armed we are in the cities. Nearly everyone I know has a gun, and none of us is the crowd you'd normally associate with gun owners in cities.

Well the the stats don’t lie. Gun ownership is significantly lower in cities than in rural areas/small towns.

Per capital is higher in rural, but there are going to be more guns in a N mile radius of you in the city, and a lot more armed individuals (ten people with one gun each are more dangerous than one person with ten guns).

Is it possible that the stats are of the flavor (total guns)/(total people) and inflate rural ownership because of a gun for small game, a gun for large game, a gun for self defense, and in certain parts of the country a gun for extremely big fish? Or is gun ownership when computed as (count of people with guns)/(total people) still inflated in rural areas?

As kortilla mentioned, while you may have a different experience, the statistics are pretty clear. Also, not all firearms are equal, and while I can't back this part up with stats, my general experience has been that rural folk are much more likely to own more effective firearms: many own a semi-automatic rifles chambered in an intermediate cartridge and some own semi-autos chambered for a full rifle cartridge. Both are a lot more deadly than grandad's pump-action shotgun or a 22 lr plinking rifle, which is closer to what most city folk seem to own. All that aside, most rural people use firearms more often rather than keeping them in closets "just in case" and therefore will be much more effective.

That's actually a pretty good pro to stay in cities.

Mass hysteria + lots of guns = not a place I'd like to be.

> rural areas won’t be safe from desperate people.

That's why you need to buy yourself a private, remote island, or at least a farm on a remote island. Also, if you don't live on it alone (regardless of ownership), it may not import more food than it exports, otherwise the other inhabitants will want to get your food.

Of course this is hard to find such an island as there aren't many such places and unless you're very rich it'll be beyond your budget anyways.

Also want to stress again that it won't be enough to just have an island, it has to be remote as well otherwise people just swim over. The number of boats is limited usually so likely few people will get to you.

> unless you're very rich it'll be beyond your budget anyways

I mean, there's a decent supply of islands (many with houses) that go for barely more than a Bay Area single-family home.

Not even "barely more". With 30 seconds on Google I found a 6 acre private island in Northern Ontario with a fully furnished 2200 sq ft house connected to marine electric and phone service for 360k USD.

Something more remote is even cheaper, because there's a ton of supply of islands that are only accessible by float plane, and not much demand.

This may be a pandemic, but it’s not a zombie apocalypse.

Yes, this is so true. My spouse has a rare chronic condition and moving anywhere rural is a non-starter. Even with great doctors in a big city, it's still a struggle getting the care she needs. I can't imagine if we lived somewhere with less than stellar healthcare.

I guess it depends on whether you are vulnerable health-wise. If you're fit and healthy and not of retirement age I really can't understand this kind of thinking.

I'd rather live in an urban area during a pandemic.

Consider yourself an outlier. I subscribe to several regional real estate newsletters, and for the last five months they've been all about people moving out of the cities and into small towns.

It's counter-intuitive but the access to testing and hospitals is very important.

As with any disease, the best treatment is prevention. Which is what being rural gives you.

I'd argue that there's less need for testing in the first place when your nearest neighbors live a mile away.

> I'd rather live in an urban area during a pandemic.

Depends what kind of pandemic. If being in a city increased your exposure and infection rates 10 fold, and the disease is a killer without treatment, it's a lot worse to live in a city.

Rural areas have often had higher rates of covid 19 infections than cities in the area. Look here for some examples: https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/coun...

Some of this is simple variance, but rural areas are surprisingly dangerous over time where cities simply show up sooner.

> have often had higher rates of covid 19 infections than cities in the area

Several factors at play:

- people thought COVID19 was not a thing outside of cities so people were more careless in rural areas

- wearing of masks is much less in rural areas than cities

- more local communities in rural areas (religious services for example)

But if you NEED to isolate yourself it's much easier to do it in the countryside than in a city.

> But if you NEED to isolate yourself it’s much easier to do in the countryside than in a city.

In practice there are several flaws in that assumption. The lack of grocery deliveries is one major issue with rural isolation. Keep digging and you find several short vs long term issues. Consider, if you need a car you then need to eventually take it in for repairs etc.

In rural areas you have more space to stock food to live completely off the usual distribution networks for a longer time. In cities if you live in a small apartment you are tributary to shopping very often or having deliveries frequently, which increases the chance of contamination.

You can keep a years supply of food in a small closet. Everyone can’t suddenly buy a years supply of food in a crisis. Which is how this really plays out, if everyone tries to suddenly stock up on supplies that fails. So space is not a limiting factor in a pandemic.

Further, going to work is a larger concern. A higher percentage of people in cities can WFH indefinitely.

> You can keep a years supply of food in a small closet.

Yeah, I worked out the space requirements once, and it's actually feasible to stock a lifetime's worth of Calories in a fairly small apartment.

Or yacht prices, once it works on them. Right now tho, with not much to go I'd say they should drop. I have a scraper, but analysing data is pain.

As long as you are willing to be somewhat discriminating, you can find rural areas with great internet service. My farm has a fibre connection and the ISP offers 100/100 service. If you travel a few miles down the road, the ISP there will provide gigabit service.

In Australia I moved to a very rural place, buying a house based on where I could get wireless broadband as part of our nationwide broadband rollout - I now live in a beautiful remote place overlooking the ocean with 50/20mbps Internet. Before here, I lived in another rural place near the ocean, however there was no broadband and 4G was too expensive. I ended up making friends with someone who had a 1gbps fibrelink 10km away - we spent weekends building a point to point network, and I ended up with an extremely reliable 20mbps connection for about $300 in parts over Mikrotik routers.

As other people have said, you can make it happen if you really want!

Ok you got me with "overlooking the ocean". I will admit, I was a little bit envious. But then I recalled the great Terry Pratchett quote that you are living on a continent where pretty much every animal, except for maybe some sheep, is equipped with some venom and actively trying to kill you. Plus some of the plants. So I guess it's fair play.

Move one more island over to New Zealand. The animal kingdom is much more friendly.

Rubbish. I went to New Zealand once and the fucking mozzies were the size of a golf ball.

I'll take the one redback encounter every 5 years in Australia thanks.

Haha, an Aussie complaining about the insects from other countries. That's rich!! When the mosquitoes are that size, they are easy to see and catch. Here in the southern part of the US, we have to keep our small kids on leashes lest the "mozzies" carry them away.

Well as a kiwi I remember at the tender age of 11 going to Australia for the first time and finding a stick insect on the footpath about as long as my forearm the first time we left the hotel. That was fun.

Actually I lived there a couple years in my late teens and you look up at the power lines and see big spiders in them but that's about it.

I've been far more terrified of my encounters with Suzumebachi and Mukade in Japan.

And the one time I went to the USA I nearly had a panic attack as soon as I left the hotel and didn't spend more than 2 hours outside the whole 6 days I was there cos the idea of a lot of poor, disenfranchised, mentally unstable people with guns just freaks me out.

The seriously poor won't have guns, they're too expensive. Generally the (...mentally present...) homeless are terrified of getting injured, because of course no health insurance. They're not doing muggings.

Knives and sticks are a different matter.

Los Angelino here I gotta say the homeless are absolutely 100% not afraid of being injured and are only on this plane of reality in a physical sense. Take a drive down Venice boulevard to Skid Row and you will see encampments of a symbiotic nature - literal fused together tents with working lights on the ground level of skyscrapers with garbage fires to keep warm. Australia has the reputation for killer bugs, a walk into south side Chicago is exponentially more terrifying than any wildlife encounter as at least animals follow patterns and will typically leave you alone if vice versa. We are an unhinged and well armed people with lack of affordable healthcare. But I do have 90 mbps download and upload speed so I kinda just keep my head in the clouds.

They’re only actively trying to kill us if we’re actively trying to kill them. So long as you don’t wander into the bush like an idiot, your encounters with dangerous critters are limited to whatever the cat (literally) drags in.

This is part of a global 5G/Soros/Gates conspiracy to keep Australia uninhabited - I've never been bitten by anything - maybe a bee when I was 12? And I spend the majority of my time walking, hiking, surfing, sailing, camping, etc.

Minor nitpick, bees don’t bite, they sting. Biting happens with the mouth (ants, beetles, etc) not a stinger.

It is a very much overdone reputation. I've lived in Aus all my life, much of it rural and I don't know a single person who has been bitten by a snake or shark let alone died. Spider - maybe but most a very shy. But again some bitten, not one died that I know of.

Compare to the USA. Cougars and bears should scare you MUCH more than anything in Australia. And rabies. We have no rabies.

Some rabies-like diseases but you would have to be extremely unlucky to get those (lyssavirus from bats for instance)

In the USA you might also just you know, get shot.

It took all of two minutes to find a shooting incident in Australia from last year, showing not only that youre wrong, but that yet again gun control stops law abiding citizens but not criminals.

https://www.npr.org/2019/06/05/729901315/deadly-shooting-sho...

I'm wrong about what? The grandparent made a joke about dangerous animals killing people in Australia, I made a joke about guns killing people in America. Also picking out a single data point is a strange way to prove your point and try and start an argument about gun control in the US vs Australia, a debate I don't care to have since I don't live in either country.

Ok so you have no care about the subject, no care about either country, you saw it fit to respond.....

And yes one datapoint is all it takes to show you can die in Australia from a gun just like in the US and nearly, if not every, country on earth. What a wonderful thing to joke about.

Well tone is hard to convey over this medium but you saw fit to hijack this thread and push whatever adgeda you have about gun control.

That you don’t care about.

Honey bees are the number one venomous killer in Australia.

From an Australian perspective, the USA can look like a place where the earth might suddenly turn to jelly and knock your city down, or giant wind vortexes could descend from the sky and carry you away - things which seem more terrifying to me than a few animals I can easily avoid.

I kind of assume those things are blown out of proportion in media much the same way our wildlife is.

We should take into account the fact that Terry Pratchett lived in England where there is very little in the way of either deadly wildlife or significant natural disasters.

Those are generally problems in the warm parts of the US. If you can survive snowy winters, you won't see much of the more exciting natural disasters. There are no hurricanes or tornadoes in Maine.

Heh Heh Heh. Well, Isaac Butterfields response applies:

https://youtu.be/t1uZZ_xRYvQ?t=153

Snakes and spiders are pretty much it. Snakes are very rare to see unless you travel to the bush. Spiders though, are common outside.

Once you've learned not to turn over rocks in your garden (with your bare hands anyway), you're fine. :)

> Once you've learned not to turn over rocks in your garden

You have no idea how ridiculous that sounds to me.

It's the kind of thing you learn as a kid in Australia. It's super common to have "red back" spiders under rocks, logs, or pretty much any object that's stationary on the ground for more than a few days.

Red back spiders are poisonous, evil looking little buggers! ;)

https://en.wikipedia.org/wiki/Redback_spider

Funnel web spiders are a thing as well, but they seem to be less common here in Victoria (south eastern Australia). Red backs though are everywhere.

https://en.wikipedia.org/wiki/Australian_funnel-web_spider

That is creepy. I could not relax in such a garden, let alone let the kids play in it.

That sounds very interesting (I am also in Australia). Would you mind sharing (roughly) where you live?

Let's say, beyond 42deg south.

Oh, based on your comment I was also expecting more or less a place in QLD. What were your main reasons for Tassy? Climate? House prices?

Climate, house prices (rather, value - acreage overlooking the ocean surrounded by trees, for less than a suburban shitbox on the mainland), remoteness, no traffic, etc. etc.

Many places along the entire east coast would match. And Tasmania.

BTW while checking that latitude (southern half of Tasmania), I found this interesting website.

https://birrraus.com/

"(Better Internet for Rural, Regional and Remote Australia)"

I just moved out of Sydney to somewhere regional (not rural). NBN installation is turning into a joke (probably a month or two off due to tree roots growing through a cable) so I have turned to 4g. Telstra now do 200gb a month for $75, which has allowed me to stay here working (work only data usage is 100-150gb a month). Didn't think I'd be grateful to Telstra ever in my life, but here I am.

It’s better than the DSL AT&T provides in some parts of SF where “you’re too far from the CO, you’re maxed out at 12”.

LOL, try Verizon-level of rural infrastructure neglect. I paid for a guaranteed 3,300Kb/s down, 768Kb/s up. I usually got 900Kb/s down, and 150Kb/s up. to quote the linesman who had the shit job of patching together the literal 1930's(!) phone lines," I was told by corporate to absolutely not replace any lines unless they're taken out by a drunk or a tree. the cable on that spool I pull behind my truck is going to go white from sunfade long before I ever get to use it."

I non-jokingly asked him which tree I should cut to 'accidentally' nail the worst sections of wire. his response was essentially 'they're all so bad you'd need to cut down a lot of trees'. Verizon DSL was my only choice beside Hughesnet or dial-up. I ended up moving instead.

Man that's bad, here in Switzerland my Line was upgraded to 10Gbs from 1 and i was mad that the router just had four 1Gbs and one 2.5Gbs cable ports...so i had to buy the business model router with fiber LAN (witch i like much more..thinner cables).

Nice, what canton are you in?

Zurich, but outside city.

Don't you have unlimited data plans in the US? Or local Wimax ISPs?

I can't speak for the rest of the nation, but a couple years ago the big ISPs in the midwest started phasing in 1tb data caps.

Pretty hard to eat up a 1TB Monthly Data Cap using a cell phone (the OP's situation).

You're right, totally missed that.

doesn't matter if there's no LTE (or even 4G) out there.

4G is very hit or miss geographically, and "unlimited" data plans tend to throttle you at usage as low as 2GB/mo, and will refuse to renew your contact if you use too much too regularly (now that contacts are rare, that often means justa few consecutive months over an unwritten amount will get you terminated.

Lol here in .it I get a 1 Gbps ftth connection for 24,95 €/month

I was paying in excess of $112 US a month for the priviledge of DSL that had about a 60% uptime on fair weather days, and 0% uptime on rainy days.

I also forgot to mentiom how my DSL speeds got that low; one section of telephone wire on my rural street had literally no usable pairs left (they had all corroded into nothing). the linesman had to reroute our phone line about 1.5miles out of the way in the opposite direction of the CO to get us reconnected. after that, if you picked up the phone, all you'd hear was a ghostly whistling instead of a dial-tone. the only reason I still had a 'working' DSL connection was from modifying the firmware on my D-Link DSL modem to do horrifying things. :|

The more I grow up and read about the US, the more I am glad I was born on Europe.

Absolutely! As someone who lived in a third world country for 2.5y (2014-2016), I had to resort to satellite internet for semi-reliable uplink. I have to admit that I’m not too familiar with the differences between Starlink and what’s already available, but what I do know is that I had to settle for 4/0.5mbit and was really happy with it.

Reliability of the connection was much more important than the throughout.

Main advantage of Starlink over existing satellite is a (much) lower latency. Starlink constellation is flying by with a new satellite overhead every 10 minutes or less, and they're operating just above the atmosphere. Musk has claimed 20ms latency, which if true, would be absolutely a smash victory for rural Internet users everywhere.

That would have been absolutely awesome. If memory serves me right, my latency was around 150ms on average, which was ok-ish but not great.

On satellite right now. I'm getting ~600ms latency :(

The article says the beta testers are seeing 30-90ms.

> The article says the beta testers are seeing 30-90ms.

That's actually pretty impressive.

Why is that impressive if the original press announcement said it would be < 20ms?

Satellite usually gets several hundred to nearly 1000 ms of latency, so being under 100 is a big deal.

Satellite latency is so bad most people with smart TVs can't stream to them because most streaming buffers empty faster than the network responds with more packets. They are usually stuck with their computers to pre-download content or increase buffer size/decrease video resolution.

I didn't say the latency was as high as GEO. If a car commercial said a car was $20k, and you got to the dealership and it was $60k, would you just throw your hands up and say it's still pretty good? The point is that the publicity was based on completely fake promises.

When you live in mountainous areas you usually have three options.

Dial-up - if its offered still in your area,

Point to point internet - if you have neighbours in line of sight also participating.

or Satellite - if you are in line of sight to the satellite

Many people I know up-country usually all they get is satellite, so yeah starlink may not the promised 20ms, but its hella lot better than what traditional satellite is providing (also hopeful starlink provides better reliability and tech support which in satellite is quite atrocious as well)

Heh - as far as what's promised vs delivery, read the fine print on satellite internet plans, they have wording saying the hope is to provide the advertised speeds but there is no guarantee that they ever will.

Everything you just said applies to starlink as well. Why the double standard? Still need line of sight, they still won't provide advertised speed guarantees ( no ISP does).

And most people don't care about latency. Bandwidth is far more important.

> And most people don't care about latency. Bandwidth is far more important.

Nobody who has ever used existing satellite says this. Latency becomes a huge issue at 500ms.

For real time games, yes, for the vast majority of traffic, no. Streaming and web browsing are plenty interactive enough for this to matter. Even VoIP does perfectly fine if it's tuned properly.

Browsing on 500ms is terrible and 500ms delay on voice ruins conversations. Most streaming apps don’t even have deep enough buffers to deal with the terrible latencies of satellite.

A single web page load of a “modern” website requires 10s of TCP connections to load resources from CDNs, each of which requires TLS handshakes taking 1.5 seconds, which are not initiated until the initial page is downloaded enough to see resources to load.

I had to live on this Internet for a summer. It was awful and was only good for long downloads. Everything else was literally better on dial-up. And this was 15 years ago.

You keep mentioning this, and I can't come up with any announcement that it would be under 20ms, just quotes from Elon saying the target was under 20ms. Is there a press release you're aware of with that quote?

> Musk stated that Starlink's 20ms latency is designed to run "real-time, competitive video games." He also stated that Version 2, at a lower altitude, "could be as lows 8ms latency."

https://www.shacknews.com/article/118663/spacex-starlink-20m...

Version 2 is not at a lower altitude. In fact, most of the latency incurred is not from altitude at all. They are moving several hundred kilometers from their highest to lowest orbit, and if you do that math that's not accounting for much of the 20ms.

> Why is that impressive if the original press announcement said it would be < 20ms?

It's impressive that they are not too far off, while the service is still in beta. And compared to regular connections you can get on Earth, it's a very decent ping.

As a side note, press announcements are made to generate buzz and create headlines, so you should always expect hyperbole instead of strict facts. If you look at video games announcements, press releases constantly state that every new game is the best ever in their genre.

I think the latency will go up after launch. They don't have much of the complexity in the system they'll soon need.

There's an update to the article with more speedtests, including one hitting 20ms

That's better than the latency on my fiber connection

Location matters, so FWIW it seems to be from northern US / southern Canada to LA.

It would be pretty nifty if the latency and jitter is consistent enough to be used for basic SIP/VoIP. Under 100ms in general would be hard to notice latency much different than a mobile phone. Starlink could go as far as rolling their own VoIP service with guaranteed QoS. I would also hope during beta testing they are developing intelligent queuing to deprioritize bulk transfers, and optimize bursty traffic where latency would be most apparent.

Where did you live?

Cambodia.

If I can get those kinds of speeds in rural areas, I would legitimately uproot my life and move. I can work from anywhere, and I reckon, there will be more and more people like me in the post-Covid world. This has the potential to reshape our world and turn the tide away from mass urban migration.

My grandma in rural Arkansas would get 26.4 kbps over dialup. Cell service is passable outside so she could get a house mounted antenna but the data caps are so low it's not worth it. Sat internet was crazy expensive last time I checked.

Also she refuses to learn computers so I gave up trying to get her internet but that's not related to this conversation.

There are ways to get unlimited service still, which is what my parents do. It's just a bit... hacky. Basically you get an LTE modem and SIM swap to it. The ROOTer mod of OpenWRT is specifically for using LTE modems and works quite well. There are some drawbacks, but it's certainly better than the 1.5 Mbps DSL alternative (which they keep as a backup / automatic failover).

On AT&T (technically a tablet plan) it gets deprioritized after 22 GB, but that's never been an issue for them, even the one month they did 500 GB. I've never seen their speeds affected. Costs about $20 or $25 / month. This is a riskier plan, in that AT&T could check the IMEI and shut it off at any time if it doesn't match (as they've done in the past).

The Sprint plan (hotspot plan with a public routable IP, $41/month effectively, but prepaid for a year at a time) does not get deprioritized, but it's probably not worth it unless you can get band 41. If you do get band 41 you'll see some very nice speeds though, at least download. But T-Mobile is beginning to shut down Sprint's band 41, and to my knowledge this plan is not permitted to roam onto T-Mobile yet. It's keyed in their system as a mobile broadband plan with a 20 or 25 GB bucket of data, however it has unlimited overage. I done hundreds of GB on it no problem. It exists only because of the licensing arrangement for Sprint to use the EBS band 41 spectrum.

I know someone on that tablet plan (which, BTW, you can no longer get - go check ebay) who lost it by consistently hitting 1TB every month.

What's the Sprint plan?

I'll throw a mention in for OpenWRT too. I tether my iPhone to my OpenWRT router when I've had meaningful internet outages. I've done it using wifi and USB/lightning. Being wired is more reliable obviously. I'm sure my carrier can tell it is tethered data though. Which has its own caps/throttling.

Maybe she's on to something with the no computers thing.

Reliable 10mbps is a stable full hd stream and 5ish mbps to spare. Easily one stable video call or two at lower quality.

It'll be annoying if you want to download the latest AAA shooter coming in at dozens and dozens of gigabytes, sure. But you can do a lot at 10mbps as 1-2 people needing the uplink. That's really nice.

If you're already used to scheduling this overnight, having it done in only one night is going to be a significant improvement.

Totally agree.

Also - my frustrations over the last few days compel me to point out that the latest AAA shooter is in fact 227 GB, plus >100GB of updates.

Has anyone come out with a reason for the size? I mean they seem to reuse a lot of the same textures throughout the game so I have a hard time understanding how it managed to swell to that.

People keep paying for it. That's pretty much it. There's no legitimate reason for a game at the resolutions common to modern consoles to be that big.

Then again, I guess you have to store all those textures for the $9.99 exclusive hot pink and baby shit yellow digital camo shotgun skin somewhere...

IDK if they do this for PC versions...

But a lot of these games not only Reuse textures but also duplicate them. They do this to improve load times for levels/etc (less seeking).

Playstation DOOM is an early example of this pattern. The WAD files were actually 'per-level' and contained all sprites/textures/etc used in said level.

Wouldn't duplicated textures get de-duplicated by any halfway-decent compression algorithm though? I'd assume the game is sent over the wire compressed

That's not how compression would work. Think about if they are adding a new level. The algorithm would have nothing to compare the files against except what is actually being sent. If the texture appears in each level, and each level comes in a separate update, there's no way for the algorithm to reference or incorporate the files that are already in your computer, so it has to compress it like it's a brand new file. That is unless you have a costume installer which duplicates the file once it runs, but that would probably be a ton more work for the developers.

I think AAA games do this now anyway to allow this like smaller patches and playing while the game downloads.

You have this completely wrong. "there's no way"? It's extremely easy to make a patch file that references existing data. This is a problem that has been solved many times.

And that's not relevant to initial install anyway, which can/should be a single download.

But thats not compression. That's downloading and running an executable that duplicates files.

What's your definition or compression? A .tar.gz will de-duplicate files that will be re-duplicated on decompression.

Having a reference dictionary is a type of compression.

You need a program that decompresses no matter what, and the ability to reference existing files barely changes it. It's not a post-processing stage where files get copied around. It's the ability to reference arbitrary chunks of data inter-file just like it can reference arbitrary chunks of data intra-file.

At the most basic level it's like taking a compressed file and chopping it in half. The user already has the first half, then they download the second half and "resume" decompressing.

That MIGHT be happening, honestly not sure.

There are possibly challenges in that however, my understanding of most compression algos (which may be inaccurate or incorrect) is that you'd run into limitations either with size of data overall or the memory requirements.

Of course, I'm sure there's other ways around that (And I have no clue if they are in use). One option would be to send over the assets in a master 'assets' package and then duplicate/assemble the data as intended during installation.

Most compression algorithms have a limited data reference window (dictionary size) - tens to hundreds of MB being typical upper limits. If duplicated data is further apart, it gets compressed again.

If the data isn't pre-compressed, or doesn't otherwise significantly vary, then using the right tools should de-duplicate the data. Things like lrzip, rsync batch mode, bdiff should work well. If large blocks are identical, many filesystem archivers/backup solutions should do good deduplication, and squashfs can be kind of nice.

I assume it must be a massive refusal to deduplicate. For example in the Verdansk map, many of the houses and buildings are cut and pasted. In theory, these could be aliases of the same model, but I suspect they are all baked in.

Other than that I just don't know. Maybe they used bad settings for whatever is their compression algo and they don't have any experts who could actually identify the problem? It's a preposterous idea but I can't explain it otherwise.

The house models themselves might use the same geometry and textures, but if you pre-bake any lighting then that’s a unique texture across every single surface in the world that can’t be deduplicated.

Depending on how the models are expressed in the environment, theres a good chance they end up as triangle meshes with unique coordinates by the time the running game eversees them.

Wouldn’t they want to keep reused objects instanced for performance, or is there some other advantage to just meshing everything out into a unique set of level parts?

Unless it’s changed a lot since I worked in game dev, models absolutely are instanced but poly counts, texture resolutions and number of textures have gone through the roof. You used to have a texture and maybe a light map, now you have separate diffuse, specular, bump, glow, decals, etc. etc.

Everyone’s doing physically based rendering, so now it’s base color, roughness, metallic, and specular. Plus the glow, decals, etc.

The amount of uncompressed audio in todays games is too damn high.

I can think of two reasons which are purely non-techincal:

- The bigger the game, the less space for other games.

- Much like adding metal to the inside consumer headphone purely to weight it down and give a premium feeling (they do this with other products as well). A larger program may be perceived as more significant.

That backfired, I chose my last premium headset over the other because it was lighter.

Who wants a rock attached to their ears? In their carry on?

People who buy Beats headphones.

Here are some reasons I can come up with (3 is weak):

1) Fancy updaters with bindiff and compression take time and money. As a first step, you need repeatable builds.

2) A broken fancy updater results in a reinstall, support calls and refunds.

3) If the updates are pushed by the store, there is no incentive to reduce the cost.

With those three, if you can get away with shipping everything that's changed, you're going to do it.

Games have always had this dilemma. Content creators and level makers will simply try to push in as much as they possibly can until something says stop. Historically this has been the size of one CD but nowadays there are no such clearly defined limits.

If the engine starts compressing better then the map creators will just use that newly gained savings to add even more content.

Exactly. The limiting factor isn’t megabytes, it’s how long the customer will be willing to wait for a download.

Probably because bandwidth is cheap and artists are expensive.

Google Stadia should ship 1TB disks filled with 3-4 games to Starlink customers.

Edit: these customers would be unable to use Stadia properly, because of bad latency; at least you can capture them as customers.

In KC you used to be able to get free Google Fiber at 10mbps. Free as in beer.

I never needed anything better than that to watch Netflix or program or enjoy basic internet browsering. It was awesome!

To get that out in a rural area and have it reliable is a game changer for a lot of people.

This right here. In 1997 I resold a 56k connection over dialup and did a bang-up business until the local phone company put me out of business.

As someone who lives in a city where ~15 mbps is the max you can get... this is accurate. Streaming in 1080p with high quality may not work. Downloading a game over 80gb requires a couple of days. Oh, and I have just 1mbps upload, so syncing a video from my phone to Google Photos may take a couple of hours as well. ISPs need competition.

And 30-100 ping is usable as well - wonder what the packet loss rate is.

I am with one of the world’s biggest telcos, Telstra, on a fully owned and operated fibre-to-the-premises network. I live in a decade old estate half an hour from the centres of the third and sixth largest cities in the country.

The maximum upload speed currently on sale on that network is 1 Mbps. And it’s a monopoly. Inequality in Internet access aside, that maximum is below the average fixed upload speed of 170 out of 170 countries listed by Speedtest.net.

1 Mbps. I did three months of complaining and got 5 Mbps. The local regulator doesn’t care because even though they’re not actively selling it, said monopoly network is “capable of 5 Mbps”.

Contention aside, this and 5G will be important for those jailed by infrastructure monopolists. Here an extreme example, but the same is true across North America too, for instance. Rural users sure, but this can extend beyond that.

> I am with one of the world’s biggest telcos, Telstra...

You have my condolences :(

That's so bizarre to have that kind of low speed on fibre, though. I'm half an hour by car / bus from the 4th biggest city, and I'm reliably getting 10Mbps upload over the copper lines. (I was getting the full 20Mbps of my plan, until NBN recently botched the repair of a temporary line fault - their repair guy even said it was the legal required minimum speed and he wasn't going to try any harder than that.)

Is iiNet not available where you are? Though since it's all NBN now, the reseller telco shouldn't make any difference.

That 6th biggest city is awesome though - if I didn't live here, I'd move there. 3rd biggest is a nice place too.

It’s “Telstra Velocity”. It just got a special exemption to not make it operate like the NBN renewed because, and I paraphrase the federal government’s words: “because they couldn’t be arsed and if we didn’t renew it they’d turn off the Internet and blame it on us”. I too would like to have regulatory exemptions extended indefinitely with the excuse being that I can’t be arsed, but elected officials can just tell people to use mobile data instead like my local one did. Charging taxpayers more than $30,000 over a year.

No other providers are available. Technically OCCOM and probably still Exetel, but they can’t offer faster speeds and have to charge more than even Telstra.

My most recent communication with the federal communications department included this wonderful phrase:

> “Please note that this legislation applies only to the capabilities of the network. The plans offered to individual consumers may be limited to slower speeds.”

Unfortunately it seems like Starlink in particular probably won't be useful for urban areas with poor competition. I'd be interested if it were, but from the sounds of it, they won't be able to handle lots of users in a small area.

Likely, but Australia might have a bit less of that because all the big cities are very distant from each other. A Starlink satellite with a view of Perth, for example, is going to have close to absolutely nothing else on its horizon. Above the US, for instance, that’s unlikely with many people living in between too.

Wow, that's terrible. People like to complain about internet speeds in Australia, but I thought it was OK if you had FTTP. This is the first I've heard of these localised garbage fibre monopolies. https://www.itnews.com.au/news/telstra-gets-two-more-years-t...

I get ~2 mbps in my rural area (Canada). This is a major upgrade for me. RIP Bell.

> Even a reliable 10mbps

What are some applications that requires more than 10mbps speed in rural places?

Probably same applications that require more than 10mbps in non-rural places.

Same applications as anywhere urban. We folk living out in the sticks may appreciate the outdoors more, but there's still those of us who do use computers.

Example: Downloading modern software over anything at 10mbps is a royal pain- VMs, docker containers, bloated IDEs, gigabytes worth of OS updates, tens of gigabytes for games, you name it.

Streaming services cap quality, or stutter out horribly. Don't even bother trying to stream over satellite (even at 25mbs with hughsnet, for example), you'll blow your cap out of the water in a hurry and it'll make you resent leaving the 90's.

WINDOWS 10 OS.

The amount of updates you have to download weekly is insane.

cripples my connection to do anything else for hours at a time. Have to switch to "metered connection" to trick it into thinking im connceted to internet on a smartphone with super expensive$$$ dataplan, so to leave any updates as 'frozen'.

(for a non tech saavy person, they would be driven nuts... every day their computer would just grind to a halt doing network stuff and not know why)

My rule of thumb is to budget 3 mbps for every "real computer" to keep itself up to date. By my count, in my home, that's 15. Here that's phones, traditional computers, VMs, and Raspberry Pis. And about 1 mbps for every "kinda computer". So that would include smart speakers and televisions.

A family.

>What are some applications that requires more than 10mbps speed in rural places?

Why do rural areas need any lower internet speeds than you do in the city? What are you doing in the city that they aren't?

Agreed. Maybe it's gotten better, but when I used HughesNet in the middle of the woods in Maine, I would say it was comparable to dial-up except with worse latency and reliability. Anything in the ballpark of a 4G connection would have been phenomenal.

Viasat is already available and provides 12-100 Mbps service. Viasat only uses 4 satellites. HughesNet provides up to 25 Mbps service and I think has 5 satellites.

Caveat: Of course I've never used either service myself, so I can not speak to the actual download speeds vs advertised speeds.

I live in a rural area where both of these services are available. They both come with seriously outdated data caps. Hugesnet, if I recall, charges $150/month for their fastest tier, but after 50gb, your speed is reduced to 300kbps.

I haven't experienced either service, but I regularly read that they are riddled with disconnects and spotty service.

Currently we're paying a ridiculous amount, $200/month, to Unlimitedville for what is literally just a 4G hotspot in our house, but with an actual, usable data limit (1tb/month).

I have some serious concerns about Starlink, but until we decide that internet is a utility and rural people need access too, I guess I'm happy to see at least somebody is trying to make rural internets better.

"I have some serious concerns about Starlink..."

Curious, what kind?

>Curious, what kind?

Maybe that you might be forced into an NDA for a service that never ends up delivering on promises?

Both the speeds and the latency are worse than what was originally promised. there have also been no pricing for the plans, nor have there been pricing for the equipment. This has been the biggest downsides since day one.

I saw someone in this thread mention hacking a 4g tablet plan to work as modem doing a sim swap. You could possibly do something similar by setting up 2 or 3 separate accounts so have individual data caps on each and cycle through them each month. It sounds like a hassle but could save you enough to be worth it

But of course, remember that you need cell service for that to work.

Those services' satellites are in high orbit and therefore have very high latency (think in the 800ms range), so those seemingly fast speeds rarely result in a fast experience.

They're both slower than advertised, have tiny data caps, and have extremely high latency. The latency is the real killer.

I've used both services for several years, and they are perfectly fine. I think people have lost their objectivity completely here and somehow think starlink is a new technology that nobody else was capable of making.

This also makes remote work more viable in non-traditional locations.

Bandwidth is great and all, but satellite latency is a killer. Been there, done that.

This is mentioned on literally every Starlink thread I read, and the response is always that unlike previous satellite internet offerings, Starlink operates in low earth orbit, drastically reducing the ping time. Do you have more to add?

This ain't your grandpa's Hughes.

> latencies or ping rates ranging from 31ms to 94ms

That's on par or better than what I can get with either point-to-point wifi to small ISP or LTE.

31ms isn't too bad compared to the 20ms I get on bonded vdsl2.

Lucky the article includes ping times then.

What he said. From the article:

> While 60Mbps isn't a gigabit, it's on par with some of the lower cable speed tiers and is much higher than speeds offered by many DSL services in the rural areas

LOWER?! Cable doesn't even serve non-television use near me. Phone lines can carry 5 Mbps DSL if you're lucky. I happen to have a line-of-sight to a major radio tower so I get wifi from a mom-and-pop ISP, with a 10 Mbps cap. If I was on the other side of a hill, I'd be screwed.

Modern VDSL can reach around 200mbps at distances something around 3-5 km from exchange. DSL obviously lower but possibly similar to Starlink.

All that said you can those speeds with Starlink anywhere, for much less than installing a phone line.

Possible != commercially available. I talked to the guy doing DSL installs in my area, he says 5 Mbps max if you get lucky.

Starlink is leapfrogging the cable/telco industries by not needing the same infrastructure.

No, it's not. So far you don't know anything about the price or the actual speeds that will be offered. So far all we know is that the latency numbers that were promised of sub 20ms are clearly not going to happen.

> So far all we know is that the latency numbers that were promised of sub 20ms are clearly not going to happen.

What? How do we 'know' that?

This is only a beta, and we don't know how carefully the testing was done, and we don't know how far away the servers were from the ground stations.

And some users already managed results of 20ms.

Because the speed tests show it, and there are no inter satellite links. In fact, because of the limited coverage in a very specific area, is highly likely that what you're seeing is on the lower end of latency for the final system since they will have more hops.

Beta speed tests show that this is the best they can possibly do?

And I think you're being ridiculously unfair about hops there. If you're in a situation where your data could spend 20ms on starlink's network and 40ms on the ground, and they route it so it spends 30ms on starlink's network and 15ms on the ground, would you complain about that?

It seems clear to me that the 20ms number was about how much latency is unavoidable on starlink, the number you get without extra hops, because hops should be pure benefit. It's not like 20ms was supposed to include the path all the way to arbitrary servers, because that's blatantly impossible with the Earth being 130 light-milliseconds around.

You are arguing against exactly what Elon had said at launch. I'm not being unfair with a beta. I've lived this industry for a while; latency will go up with congestion. Let's revisit this comment in a year.

Which statements of his am I arguing against?

Also if buffers are configured to be very small then congestion does not have to add more than a fraction of a millisecond of latency. Bufferbloat is endemic but it is not unavoidable.

https://www.shacknews.com/article/118663/spacex-starlink-20m...

He said 20ms, and you are saying it'll be higher, and that's okay. You can't just shrink buffers to get rid of queuing latency. You introduce new problems with scheduling when doing so.

I'm not saying it will be higher.

This first generation of satellites doesn't even have inter-satellite lasers. So those extra hops will never be an issue. It's a moot point. The 20ms number is about the first generation.

But some day, when those links exist, if data goes halfway around the world via satellite, it would be stupid to say "gotcha! longer than 20ms!". That's a complete misunderstanding of the context, and complaining about something that makes the data go faster.

If someone was going to use multi-hop results to 'disprove' 20ms, they don't even need data. Just look at the speed of light. Bam, already disproven! But that's because it's not what the 20ms number was talking about. He wasn't promising to violate causality.

Nobody has ever claimed 20ms to go around the world, because that's physically impossible. He did claim ping times would be 20ms, which is patently false, even within CONUS. There is no "second generation" to date. They have not been launched or tested. So the more of v1 that are out, the less v2 you can launch without deorbiting a very expensive part of the fleet.

It's not patently false! Some of the testers got 20ms already AND this is a beta.

Let's revisit this. There's no point in arguing, but a single data point during a beta is clearly not what he was referring to.

Sure, we can revisit this later.

But if we need to look at the numbers later, doesn't that mean you're conceding the point about it being "clearly not going to happen" and "patently false" that it will ever reach the promised number?

Or do you think I'm delusional to say "maybe it will, maybe it won't" and you're just humoring me by offering to revisit?

I'm maintaining these are best case numbers and that they'll go up when the service is live. I would bet you'll never see 20ms when paying customers are on it.

Just to add some color to my comment: https://mobile.twitter.com/Megaconstellati/status/1294628802...

So SpaceX only has 2 PoP locations, and it happens to be the areas where they are doing trials. As soon as the ground system needs to backhaul the signal over long fiber distances, things change quite a bit.

More like 1-2km. I don't know which "modern" VDSL you mean.

In lab conditions with cat5 cable perhaps.

In the real world, the copper in the ground or on the pole is close to a century old, and probably qualifies as cat2 if you are lucky.

To me that frustration comes from a lack of understanding that living in a sparsely populated area comes with the good (less people around) and the bad (worse infrastructure)

Totally.

A couple of years back we lived in some mountains in Mexico and the best connection around for miles was less than 5Mbps. Most people had less than 1Mbps.

We then moved closer to civilization and I remember being amazed to be able to watch Netflix in 4K with a 20Mbps connection around 2015 or so.

Agreed. I have travelled to remote places where this speed would have been a godsend, not just for internet but for telephony as well with VoIP. Imagine the kids living in such places finally getting access to all the stuff internet has to offer.

The only thing to watch out would be pricing. I hope there is more competition so the price remains competitive.

> and it will just get better.

Won't it get worse when it's generally available? More congestion on the same channel means less per-user throughput right?

And when it's at scale it'll actually have to make a profit.

It's going to get worse. Right now it's completely uncongested. The terminals are not built to handle two satellites at the same time, so the throughput they are getting now is likely close to the maximum that they will be seeing.

>Even a reliable 10mbps would be a godsend in many places.

my first foray into 'rural' internet was during the late 90s, using a satellite+modem setup.

The bandwidth was fairly decent -- equivalent to a fast ISDN line -- but the reliability was horrendous.

Drops constantly, only to wait for a whole new modem negotiation and reconnect. Think about the idea of an analogue phone modem but over an unreliable aerial connection.

My point in this anecdote : don't take the upstream/downstream values at face value -- there are more things that go into a connection that's actually worth using.

That said , I have nothing but hope for Starlink. I still frequent rural areas that , if broadband is available, it's severely limited bandwidth. I'm glad there will soon exist an alternative -- let's hope that efforts are made to keep the astronomers happy so that this network can be a real boon to humanity.

You’ve got that right! I taught remotely all summer long on 6mpbs down/ 0.3 up. Dreadful. Just enough upload speed for PowerPoint + my voice. I’d love for something like Starlink to work.

My Xfinity "Up to 1Gbps" plan never tests above 60Gbps on any speed test other than their own. If nothing else Starlink should light a fire under Comcast, CenturyLink and others to start providing actual fast internet.

I'm living in a canyon about 30 minutes from down. Our fastest internet is about 5Mbit... I used to live in downtown SF where I had a gigabit.

As soon as I can get Starlink I'm on it.

When I was growing up, 56.6kbps was the hotness. However, in the boonies where I lived, the best I could do was 33.6kbps. Most of the time, I was lucky to get 28.8, but would never get less than 14.4. Turned out, that the central switch for our town was so old and we were so far from it that there was no way for it to provide a clean enough signal for >33.6.

When HughesNet came out, we tried it. It was horrible. Only if we were lucky would we get a signal better than our dial-up's download speed. The thing about Hughes is that we still connected with a dial-up for Tx. Only Rx came from space.

All of that to say HELL YES, those are amazing speeds. However, as an adult, I became cityfolk, and have full 1Gbps up/down via fiber.

I remember as a teen I'd go on Digg and open all the articles in separate tabs so that by the time I was done reading the text ones some of the images would be loaded on the ones with pictures. We've come a long way.

Oh, you had tabs! We had to use new browser instances! Uphill, in the snow!

Seriously though, I still do this out of habit, as it's just how I like to read. Skim a front page, open all the articles in tabs, and then go back and peruse them.

I remember when Internet Explorer 7 released with tab support. I'd been using Firefox since 1.0 (something like 5 years at that point?)

He whisked me over to his laptop and started opening tabs, before looking at me like he was the first man to discover fire

When I responded "oh, I've had that for years" he looked genuinely shattered at my reaction

Tabs in 2000 with Opera v4, before it even had a freeware option!

Actually the way 56k worked is that the ISP would provide a digital signal to the phone switch, and not all phone switches supported this. The problem might have been distance but it was probably the age of the switch!

Hell for some 'holler down in West Virginia or Kentucky a reliable 3mbps would be amazing. Many of those sorts of places have 384kbps-768kbps DSL or even dialup... in 2020. They just don't have the infrastructure and it costs way too much to build out last mile over vast stretches of land with few customers, and it's compounded in some areas by a lack of even stable right-of-way where you can easily run cable.

The market for this is huge. It's probably not going to displace high-end broadband in medium sized to larger metro areas, but it can clean up when it comes to anything outside of that. The only possible competitor is fixed wireless using next-generation modulations that provide incredible range, but that's hard to deploy over a wide area if the land is not really really flat. I guess 4G could compete too if they radically cut their prices, but I'm not sure the backhaul networks out in them 'thar hills would even be up to the traffic.

Our holler in Kentucky has fiber to the premises through a federal grant a few years back, but I’ll admit that’s unusual. :)

I had 5mbps at work. Then finally 5 years later they are able to get me 12mbps. A block down can get 200mbps from same company. My friend a mile away can get 2gbps fiber. And i have joyous 0.2mbps uploads WTFFFFF

How does it scale though? If want 20mbps do I just get two antennas? At which point does it max out for a single location?

"and it will just get better." is proven with any Musk co (that I know of).

It's about expectations. Yes, 10-50 Mbps is pretty acceptable. Today. It won't be in 5-10 years. 5G will make 1000 Mbps at 1ms, pretty much the expectation, and the apps of tomorrow will need it.

In 10 years people will be talking about 50MBps like they do 3G today. I cannot operate hardly a single app on my phone on 3G. It's worthless in 2020.

Satellite internet well above 10M has been available for many years already. This isn't anything new.

I've been a digital nomad for almost a year now. If I can get 5 Mbps down/up, I can do my work without interruption. In a couple places where I could only get 1 Mbps down/up, I could still do my work w/ a bit of patience on transfer speeds & good enough audio, as long as I didn't need to do much screen sharing & video uploading, and I could usually compensate when I really had to by tethering/hotpotting to my mobile phone. !0 Mbps would be fine for one person for most things. Might feel slow for a family during peak times.

this has been done with radio-based meshnets like the puget sound data ring for years, the government hasn't updated the legislation to allow encryption and commercial data on said meshnets, making them useless. we could already be providing free internet to millions if it weren't for the government and their dumb regulations.

I know encryption is controversial, but I cant say I've ever heard someone say that ham radio bands should be allowed to carry commercial traffic (other than people who want to just get rid of those bands all together and auction them off).

you misunderstand. I was saying that a special exception should be made for meshnets and only meshnets, and for no other application on the ham bands.

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.

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.

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.

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.

Starlink internet is not meant to replace hard line internet. There is no getting around the physics of cable being so much cheaper/easier. Even terrestrial cell towers are more efficient and cheaper than satellites.

This is 100% meant to supplement existing infrastructure not to replace it. You will almost always get cheaper and faster internet (for both you and the company providing internet) over a cable to your house. If you can't get that, then maybe a cell tower is better, if that doesn't work, then a land based beaming internet may be better, after that I'd look to space based internet.

It can probably supplant land based beaming internet, but I don't see this replacing ISPs/4g/5g anytime soon.

The target audience for much of this infrastructure is the places you can't put a cell tower, but want to get signal. Ocean based travel being a big one.

EDIT: I will contend. 5g could beat out ISPs at cost (which I thought I'd made more clear above). It's about the cost of infrastructure given the population density. I don't have as much knowledge in that area as I do satellite costs. But satellites won't beat out 5G at cost, not even with the reduced cost to space that SpaceX provides. It could with another couple orders of magnitude cost reduction in space. But there's a reason SpaceX is targeting only 5 million Americans and not 300 million.

Weren't they also going to track airplanes with it, to avoid new mh-370 cases where they don't know where it went and where it crashed?

Didn't it already have Iridium on board as well?

I read some claims that the power bus was unstable and thus it shut down the satellite radio multiple times before it could transmit much meaningful data. Which would imply that Starlink wouldn't have changed much in this regard.

We kinda know where MH-370 went, just not where it crashed.

If the price is 5 times cheaper though... what then?

There is a set capacity for these satellites, they can accomodate X number of users per satellite which would be a region (I'm making an educated guess) about the size of New Jersey. I think the size is larger than that which means fewer users but I'm trying to ballpark during my lunch break.

So if they expect to have 10,000 users (again ballpark) in New Jersey, but they instead get 40,000 they'll increase their costs. If they get 1000 users in say Wyoming, then maybe the costs in Wyoming will be lower.

The issue is they have a world wide capacity of 10,000 users at 11 MBs over the size of New Jersey. Again, I haven't run these numbers but these are the kinds of calculations GEO sats make to set prices for say, airplanes traveling from US -> Europe. The difference is that GEO targets specific areas more rather than blanketing the entire earth due to them being stationary.

The benefit of Starlink is that it'll have approximately equal signal all over the earth. So if you're in a city it can cover the same number of people per mile as it can in the middle of the ocean.

I believe the satellites are moving, so there's no "Wyoming satellite customers" in the same way as existing satellite internet. The Wyoming customers will be sharing satellites with several other places as they move in and out of their regions

The constellation is moving but they are moving in a formation. Meaning you have a rotation of 2-4 satellites above your head at any give time. Each satellite can service a radius the size of Wyoming.

So you have a limit on the number of satellites and the number of customers per satellite.

The article says 5 million in the USA. That makes sense given what I'm thinking. That's about 100,000 per state, but it's not going to work like that. It's a good ballpark for the service they'll provide.

My main point, is if it's 100K per (1/50th the US) there will be less bandwidth in cities than in rural regions.

This is a HUGE boon to rural areas and will be amazing.

This will probably not affect you if you are urban or even suburban.

Their current plans talk about a constellation of 40,000 says.

While there is limit to number of connections per sat that applies to any routing equipment really, there is not much limit to how many sats can be there .

Sure spectrum is limited today ,however if 5million people use it already, other spectrum could be freed up for this purpose , if there is demand .

The constellation today is not ready for high density usage , it is not that it will never be ready .

This type of constellation won't be able to do it.

They are planning on picking off the users who cable has to run the furthest to off of the ISPs not the close users. If we started spreading out evenly I could see this, but while we've got the population density we do, I don't see this happening.

Each house will interfere with it's neighbor, the more they spread the easier it is.

>The constellation today is not ready for high density usage , it is not that it will never be ready .

I am no sith, I shouldn't speak in absolutes. But, I'd bet that this technology won't disrupt ISPs within 50 miles of a city. I hope it would, but I don't see that happening.

That's not how it works. The FCC doesn't just free up massive spectrum from other places. We're talking about 500MHz chunks that just don't exist. The FCC favors cell technologies, so if anything, they would take it away from them.

Nothing would stop them from tracking access based on location.

The bandwidth calculations work out about the same, as the users are pretty much stationary - users will be in an area handled by ~one satellite at a time (with handover between satellites).

What you said at the end is not true. It does not have a relatively uniform signal at all areas of the earth (on purpose, since they know where the customers are), and it's not a benefit to be standing capacity where nobody lives.

I think you're underestimating the number of people that live in places where cell companies don't want to be arsed to build a tower.

Cell companies are also the last people you want to get your "main internet" from. Bandwidth caps, throttling, accounts shut down because actually what they meant is that "unlimited" means "200 GB/month".

I think you're overestimating how much those people are willing to pay.

It also raises the baseline across the board by injecting a new competitor to every ISP and wireless service in the world

This is 1000% my point. There is now a baseline amount of signal per KM. That signal is now the same everywhere. It will be (about) the same speed everywhere.

This is really great for more rural areas and over the ocean.

We already had a baseline amount from geosynchronous and Iridium. This is increasing bandwidth and reducing latency, but is not establishing the first baseline.

Gah - I got so caught up in this thread. Finally something I know a fair bit about.

Yes, there is a baseline. But that's like saying dail-up is a baseline. Everyone might have it but it doesn't give people access to the modern web.

Geo, can be great in some services but it's again not a blanket. They have broad beams but they're moving to spot beams instead and those beams don't move.

This constellation is really going to be the same bandwidth capacity across the globe within a given altitude band. Which is awesome.

That's not how it works. There are intentionally launching satellites at different inclinations so that they can cover more densely populated areas with more bandwidth. The speed and capacity is absolutely not uniform over the entire globe.

are the physics of cables cheaper and easier? Took me, what, over a decade to get fiber internet in the middle of Silicon Valley.

Yes.

All this beaming technology is available on earth (4g/5g) but we don't use it for our home internet... Why is that? Cost of bandwith per person to infrastructure.

If you're in Silicon Valley then you probably get GREAT cell service, why not use a hotspot as your internet. That signal to your phone is orders of magnitude cheaper than space based signal.

> If you're in Silicon Valley then you probably get GREAT cell service

You'd think so, but we have the same problems as everyone else. In fact, I get much better reception in Sacramento and Las Vegas than in the Bay Area.

No, you don't have the same problems as everywhere else. You said it took 10 years to get fiber, there are rural places where that's not an option.

There are places where there is no sewer and houses must have septic.

There are places where there is no water and houses must have wells / rain capture.

Those are the target audience for this technology. Not people who are waiting to go from 60 -> 100 -> 1000 Mbps. This won't compete there, not based on the specs of these satellites and their 10 year plan.

I was specifically addressing your claim that one could just use an LTE hotspot. I'm aware that there are rural communities that have worse infrastructure than the bay area.

You have better access to hotspot data than someone who is further from the nearest cell tower. I'm using a more physical infrastructure discussion to draw an analogy to what you might experience. It might make sense for someone in a city to build a well or water capture system, but in general it's not the same calculation as in more rural areas, where 1 mile of cable may reach 0.5 people vs 20.

What you have could easily be too many people per cell tower in your area, which would be the exact same problem that these satellites will encounter. The physics of adding a cell tower is much cheaper than adding more satellites.

You seem to be confusing posters and issues.

I'M the one who said it took 10 years to get fiber. I'm not convinced wired infrastructure is cheaper/easier to deploy, when it means digging up millions (billions?) of miles of trenches across the country.

Nobody claimed "having fiber in silicon valley today" is not better than "having crap internet in rural america"

I have good LTE at my house. That doesn't mean I can get terabytes/month of transfer for a reasonable cost.

And by the way, my ENTIRE POINT was that "the physics of wires" is NOT "cheaper and easier".

Reply to my parent if you're looking for an argument as to why wired is better than wireless.

The problem is that the target audience you refer to doesn't have the amount of money it takes to sustain a system like this. Otherwise any of the existing satellite providers could continue to launch more and more satellites and serve very expensive internet to them. Starlink will be just as much, or more expensive.

How much do you think it will be?

I expect for them to make money it's going to need to cost $150 starting for the low end. Although, if they get government RDOF money, all bets are off since they're burning through government money.

If you go just a few miles into the hills south and west of Silicon Valley there are huge dead zones.

I was arguing against the parent post that was claiming wires are cheaper and easier. Not terrestrial wireless vs satellite.

> over a decade to get fiber internet in the middle of Silicon Valley

This is not a physics problem. This is a local government problem.

The comment was "physics being cheaper/easier"

I'm just not convinced the "physics of trenching and stringing cables from poles" (whatever we're taking this to mean) is cheaper and easier than wireless points popped up, either terrestrially or in space.

Let me be super clear about what I meant in my first comment.

Satellites are VERY susceptible to overcrowding, this is what I mean by physics. Co-Channel interference and Adjacent Satellite Inteference play a big role. The more signal you use the more of a problem it becomes.

Think about it like this. Your car radio, has how many available stations? Have you ever gotten multiple signals at the same time? This is all a really big problem for space based communication you just can't solve with more signal.

Don't get me wrong we can do a lot with our spectrum, but there is a limited amount and if we both want to communicate with a satellite and we live 100 feet apart, our signals will interfere much more than if we live 1000 feet apart.

I'm not saying stringing cables is cheaper, I'm saying laying fiber from hub to hub and then connecting to those hubs is cheaper if you're within 50 mile radius. Maybe cellular is cheaper, maybe a wire to your house is cheaper. It depends on how many people you have close to you.

There just isn't enough spectrum for everyone.

Would this improve the internet quality on planes? Just wondering how that could get better... its absolutely crap most of the time

Poor quality of internet on planes is usually due to too many people on the plane trying to do high-bandwidth things at the same time when the connection was never designed for that level of saturation. Domestic flights usually connect to ground-based stations to provide in-flight wi-fi, which probably have similar throughput capabilities to Starlink, so I don't think this would help much. International flights use satellite connections, though, and this would definitely help there.

I'm still a bit puzzled why ground based stations didn't catch on. Seems it should be much simpler - there's tons of towers already, flights paths are predictable and much lesser distance between plane and sat.

That said my experience with internet was very positive - anywhere from free 4K streaming to $20 per 17hr flight enough to do remote SaaS development. I guess US airlines keeps American tradition of depriving people of good internet.

> "You will almost always get cheaper and faster internet (for both you and the company providing internet) over a cable to your house."

Not true for me, and I live 10 minutes walk from the financial centre of a major world city. Because there's no fibre installed to my building, unlimited ~300 Mbps 5G wireless internet is both cheaper and faster than any available fixed alternative. Even if fibre was available, the 5G would probably still be cheaper!

>If you can't get that, then maybe a cell tower is better, if that doesn't work, then a land based beaming internet may be better, after that I'd look to space based internet.

It is cheaper for a company, to run fiber to your house than to build a cell tower that handles ALL the people within that region given a certain density of houses.

It doesn't mean it's cheaper for your building to do it.

In the context of space vs ground infrastructure. Ground (including cell tower) is much cheaper. And they'll add 5G to the financial centers of the world before they add it to Wyoming.

The reason companies don't "advertise" hostpots as an alternative to fiber/cable is bandwidth. It works great if you read HN, but if everyone did it, they'd start running more fiber, or cost of cell data would go way up.

> "companies don't "advertise" hostpots as an alternative to fiber/cable"

They do around here!

http://www.three.co.uk/store/broadband/home-broadband

https://www.vodafone.co.uk/gigacube/

Sorry, My experience is mostly US based. I looked into this heavily in my area 2+ years ago.

T-Mobile, AT&T, Sprint, and Verizon say their hotspots are not meant for this in my region. At least that was 2 years ago in Sub-urban Seattle and 3 years ago D.C.

Seattle has had a hotspot "broadband" internet provider for 10+ years now. And t-mobile advertises isp service: https://www.t-mobile.com/isp

Sydney?

London!

The sad part about what you said is that there is definitely fibre in your area. After all, there’s a 5G tower within spitting distance.

Cell towers are often connected with microwave links, though in a dense city, its more likely fiber.

In fact, I'm pretty sure there's fibre running down the road directly outside the front door of our building. Probably multiple different providers! There's conduits under the footpath and I've seen them threading reels of cable down there late at night.

That said, I'm very happy with 5G. There are some key advantages over fibre:

- Cheap. I pay £30/month with no fixed contract terms. Any fibre around that price is slower and comes with a 12 or 24 month minimum contract.

- Very high reliability. If a cell tower goes down, it will simply switch to another one. At my last flat, the cable internet would seem to go down for 15 or 30 or 60 minutes pretty regularly, perhaps every 1-2 months. And that's just the times I was home and noticed!

- Portability. Wouldn't be much use in a large household, but for personal use it's great that I can just pick up my battery-powered router and take it with me for fast WiFi anywhere. Including roaming around Europe (5G roaming in selected countries!)

The problem is you're basing your argument solely on the physical act of running cable/wires everywhere. When you look at the cost of just hiring a crew to dig the ground or string wire on a pole you are correct. It's significantly cheaper and more reliable.

However, add in the human factor, government. The mess of rules, regulations, and right-of-way make it considerably more expensive to run wires everywhere.

The cost of hiring a crew to dig up the ground runs to hundreds of dollars per foot. At $1MM for a satellite that can serve thousands of people is a lot cheaper.

If they all lived in the same location. If the cost is hundreds per square foot and each customer is 1000ft away from each other than thecosts become hundreds of thousands per address.

Line of sight or cell tower might work but when it doesn't a satellite will.

I worked at a GEO company. They have a HUGE department just to work with governments to buy or keep their spectrum. The rules and regulations for spectrum are insane, and not just build this cable here in my county/state/country, but now this beam goes across borders what do I do? Oh wait, this region is contested by China/India we have to sell differently to each government.

Cost of labor => It's a calculation of how many people can be served per 1 mile (any unit) of cable. If that number is 20, cable wins. If it's 5, hotspot may win (which also uses cable FWIW). If it's 0.5 then these new satellites win. 5G may try to take more market from ISPs but they know they'll have to innovate to keep customers. I can 100% see cell towers being cheaper than cable, but not in every situation.

It's about supplementing not replacing infrastructure.

A lot of what you're saying isn't actually true here.

Light moves significantly (47%) slower through optical fiber than through space. This drastically reduces the latency of long distance connections. We're talking NY round trip to Japan in the same time it takes NY roundtrip to the UK now.

The cost of backhaul and ongoing maintenance of cables is a lot more than SpaceX launching their own satellites by an entire order of magnitude.

>A lot of what you're saying isn't actually true here.

You're talking latency not bandwidth. There is only so much spectrum to go around.

My understanding is that there is only a certain amount of bandwidth to go around and there are diminishing returns on adding people beyond a set capacity. I am not arguing that fiber is faster, I'm arguing that fiber has is cheaper at the bandwidth/cost in more dense areas. I don't believe we will see satellite based communication in cities (at a large % of population) for a long time.

Now I'm willing to believe someone who's worked closer to this than I have and to learn, but I've done enough in the satellite industry to not take what you're saying without seeing more evidence.

I worked for an ILEC providing cable/dsl/fiber service and just the rough napkin math around SpaceX's launch costs makes it look significantly cheaper.

Besides the whole digging and laying cable costs, last mile work is expensive. You have to employ huge crews. Construction crews constantly dig up & cut your infrastructure "by accident" and the cost of recovering the fines from them is as much as the fines pay out. Having to send an emergency crew of three guys out for 18 hours of work to re-splice a 72-ct fiber every other weekend adds up. That doesn't happen in space.

>last mile work is expensive.

What do you mean there?

SpaceX isn't doing the install work work on these Starlink kits. They just sell you some hardware.

If you're a terrestrial-based internet provider, you need a fleet of installers and they also have to do constant maintenance work. They have salaries and they make (lots of) overtime pay. At a lot of the smaller companies (like the one I worked at) they have pensions. We had a lot of guys who started with the company out of high school, retired at 20 years, came back as contractors for a year and then did another 20 years and got a second pension from the same company...

And a couple of those guys are still contracting after their two pensions.

SpaceX can't scale regionally, they have to scale globally. The nature of their satellites is that if they want to accommodate 2x as many users in a city they have to double their capacity.

The push I'm discussing is the cost to get a baseline internet out there for 10k people per 100 miles.

I'm assuming the cost per last mile is higher in rural areas than in urban areas. If that's true right now the majority of our population is in cities so SpaceX will have trouble reaching them.

That's the core of my point. For Starlink to hit everyone in a city they'd have a huge amount of wasted capacity in rural areas.

Satellites also degrade naturally over time. They have a set lifetime and don't all last that long. Cables under the ground tend to last longer even if they require spot fixes. I'm assuming that bandwidth wise, the cables are more efficient at least until getting manufactured equipment to space becomes even more cheaper.

Yeah but we're talking about a company that has in their near-term plans the capability to _launch 4 rockets per day_.

They're completely capable of adding capacity like that and having a high satellite count.

I'm not saying that it doesn't have a cost, but they have no middleman. They don't need to buy space on somebody else's rocket. They also don't have to pay any ongoing maintenance cost beyond equipment replacement and it's much cheaper than doing it on land. I know that you have satellite experience but their launch costs are a fraction of what other companies have done till now.

I worked for a suburban, regional ILEC and our operating costs were around 100 million a year to cover like two counties of maybe 100k customers and the local businesses.

...

I actually do agree with you about costs delivering in dense markets as providing locally has certain "economy of scale" benefits. It's also not going to compete on the "Gigabit in every home" front, because I do agree that the bandwidth is more limited. Operating a local provider in general though is extremely expensive and doesn't really provide any additional benefit to the provider. SpaceX themselves have uses for this network. I also hope that I'm right here :)

The bandwidth here is 100K per state not 2 counties. That's my issue with your calculations, we're probably somewhere in the middle. They might stop servicing people outside of town. I'd probably use the metric, if you're on septic you're probably better served by SpaceX than Comcast. If you're on sewer then Comcast will probably be better (no not always, I'm no sith).

We'll have to agree to disagree. Honestly, I hope you're correct and I'm wrong. I think ground based infrastructure is still going to be much cheaper at medium to high population densities. You don't agree with me, but I learned a lot in this back and forth. Thanks.

Their target launch cadence has been missed by quite a bit for many years. At this point they are only launching their own satellites at a cost to them, and not launching very many paying customer payloads at all.

This is absolutely not true. They are still selling you a dish in a terminal that most of the population could not install themselves on their house. The majority of the installs will be professionaly done.

That's not what I'm saying. Forest for trees.

SpaceX isn't doing the installation work. They're not hiring the fleet of techs to do the work.

No satellite companies do that. They are independent contractors.

This thread isn't about what other satellite providers do. It's about the Cable/Fiber/DSL providers. And even when those companies use contractors to do the work (like Spectrum, Comcast, etc), they're still the ones paying those contractors.

Let's keep it on-topic.

Please reread this thread. This is the comment I was replying to:

> SpaceX isn't doing the install work work on these Starlink kits. They just sell you some hardware. If you're a terrestrial-based internet provider, you need a fleet of installers and they also have to do constant maintenance work

This is not true. SpaceX is absolutely going to be doing most of the installs through independent contractors. The original post implied they just ship you the antenna and the end user can set it up, which is wrong.

Absolutely not correct. Read the terms from what they sent everyone in the beta program:

"You are responsible for installing the Starlink Kit. Do not allow third parties, or those not associated with SpaceX, to access or install the Starlink Kit unless you obtain approval from SpaceX," Starlink says.

"Do not install the Starlink Kit at your home if you do not have the authority to do so. It is your responsibility to ensure compliance with all applicable zoning, ordinances, covenants, conditions, restrictions, lease obligations and landlord/owner approvals related to the installation location."

SpaceX is not doing any of these installs themselves or through contractors. They've been very up front about this and how it's not just going to be for the beta program.

Keep up the FUD though.

It's a phased array antenna -- it's really simple to install. People are already doing this themselves on top of RVs and work trucks. It has motors that are self-adjusting. Their instructions are "1) plug in socket, 2) point at sky."

Wrong. A beta program where they don't want things leaked is very different from the real system. They self-selected a group of people capable of doing this. Do you really think that millions of real customers are capable installing on their own? More importantly, doing this without an issue later that requires calling customer service. It doesn't matter if it's a phased array. It MUST be installed securely in a location where there would be no blocking. This requires a professional. Trust me, I'm very familiar with this industry, and Elon can't work magic. You are taking your own understanding of how the general population is and assuming they're all reddit/HN readers.

Please show me where they've been up front that after the beta it will continue to do this. Also remember that almost everything Elon has said about starlink has turned out to not be accurate or delayed.

The SpaceX approach also completely bypasses a ton of the nonsense local bureaucracy, stonewalling by existing ISPs, etc that a company would have to deal with to put in cables, towers, etc in the first place, saving both time and money.

Yup. This is actually one of the more important factors for consumers that is underestimated.

I've been pegged at 1000% excitement since I first found out about this project a year ago. This is the best thing to happen to this service industry since probably ISDN.

This is not a good thing. Eventually countries will heavily regulate satellites and then SpaceX will be the monopoly that keeps small players out. Instead of postponing the problem of getting rid of an evil monopoly, citizens should realize that last mile internet delivery is practically identical to water, sewage and power systems that can be owned by the city and leased out to providers.

FWIW, last mile costs only matter if your last mile isn't wireless, more or less. Terrestrial internet doesn't imply wire/fiber to the premises, does it?

The costs of installing and maintaining infrastructure scale down to nothing with population density. Also you can upgrade a switch but not if it's on a satellite. I am very convinced that the two approaches will complement each other in the foreseeable future.

Please post your napkin math. That seems to disagree with all the other analysts' math.

> Light moves significantly (47%) slower through optical fiber than through space

The what?

As far as I know c is a constant.

c is the speed of light in a vacuum

The constant c is the speed of light in a vacuum. Different mediums have different speeds of light within them; in fact, it's actually possible for light to not be the fastest thing in a given medium (for more precise definitions of all those words): https://en.wikipedia.org/wiki/Cherenkov_radiation

When you refract light, you change its speed, slowing it down.

https://en.wikipedia.org/wiki/Refractive_index#:~:text=The%2....

https://www.zmescience.com/science/physics/how-to-slow-trap-...

Altought the answers so far posted are not exactly incorrect, it is also important to note that the geometry (that is, a cylinder for optical fibers) of the waveguide used influences the group velocity of the propagating electromagnetic field.

This will always end up being slower than a plane wave in the same material, usually only somewhere between 10%-90% (and this also depends on the wavelength) of the free-space group velocity.

In a vacuum. See Cherenkov radiation[0].

[0] https://en.wikipedia.org/wiki/Cherenkov_radiation

Downvotes are unfair; this part of physics I either forgot, or never learnt, I can't tell. Thanks to those who pointed me at links.

SpaceX has to pay the same fiber cost in terms of latency as everyone else. They do not have gateways near every PoP, and will need to backhaul a nontrivial distance to add latency.

The current generation of Starlink doesn't actually have any optical inter-satellite links which would forward your packets from satellite to satellite around the world.

Your packets will go up to the satellite, immediately down again to the next ground station and from there to the destination through the normal fiber backbone.

A cell tower today is basically a selective directional broadcast of ~2.5GHz backed by a microwave backhaul, conforming to local topography and about ~3 miles away.

An internet satellite is a phased array broadcast of ~20Ghz with a microwave backhaul without regard for local topography and about ~600 miles away.

... I am not sure there is obviously a better choice here, if the launch costs no longer dominate.

>... I am not sure there is obviously a better choice here, if the launch costs no longer dominate.

I'm basing my entire comment on existing costs due to SpaceX efficiency. We'd need (IMO) more cost reduction.

What sort of physics?

You can multiplex wavelengths for wireless just like cable right? Granted you'd need to buy the licenses for those frequencies, which should be fairly cheap in rural areas.

I don't think Elon would approve this project if cable is always cheaper and easier.

Why not? He doesn't have to reach everyone, just some portion of the population that's profitable.

My argument should have been this is only (imo) going to get at most 10% of the worlds population.

That's based on the signal interference in space. Scale gets costly.

EDIT: Then I'm done replying for real. This thread has been fun but I need to stop myself at some point. Thanks for replying.

physics => Co-channel and adjacent satellite interference to be specific, the closer you put your satellites the more signal they have to handle. These things are built to scale for the worlds population and you can't just add more satellites without scaling the whole constellation. If it was built to scale for NY city then they'd need to support that population density on every single satellite. But if they build it for Wyoming and steal those customers then they need to only scale for that and allow ISPs to keep NY city residents.

While it's cheaper for cable companies to lay wire close to hubs, it's cheaper for satellites to hit people the more spread out they are. So the physics of this constellation are such that it's cheaper for them to give more rural people a baseline of internet rather than upsetting ISPs. I think that's wishful thinking by many people who hate ISPs.

I'm not saying it's cheaper for cables to hit everyone everywhere in the world, or this system wouldn't exist. As I said earlier, it's meant to supplement a weakness in our current infrastructure by blanketing the world with cheap internet that works at lower population densities.

I used this metric in another thread, but if you're on Septic Starlink is probably for you. If you're on sewer then you're already surrounded by a density of infrastructure that probably makes ground based communication cheaper. No I'm not going to bet that someone on HN doesn't have a counter example for me, but overall I think this would probably make sense.

Fiber optic cables have a much much higher usable bandwidth than the wireless spectrum. The useful radio spectrum for Starlink is somewhere between 10 and 20 GHz. Fiber optic cables have more than 4THz of usable spectrum.

And it's no affected by things like fog, rain, strong wind (at least if you don't cheap out and stick it on poles), etc.

But you’re missing the reality of the situation. Hypothetically cable is better and cheaper but in reality it isn’t because of corruption and the fact that cable has to occupy a physical space which forced the involvement of governments which makes everything terrible. Starlink will finally provide good competition to these bastards in rural areas... im in Des Moines metro area where cable sucks for no reason. I can’t wait.

One of the big excuses for the telecom/ISP monopolies in Canada is that they help serve rural areas (same deal with Canada Post).

I'm curious to see if this ends up having any effect on the current system.

Technically you could just directly subsidize access to the rural areas, for those who can't afford the much cheaper new satellite systems themselves, instead of giving these anti-consumer monsters free reign on the whole country in exchange for managing expensive infrastructure just to serve a small percentage of the population.

"5g could beat out ISPs at cost"

Fixed line ISPs share much of the same infrastructure with 5g/4g providers. The only difference between them is the last mile, so it's not like 5g providers magically dont need any backhaul infrastructure and could provide dramatically lower prices.

Honestly, this is just awesome. It's a working demonstration of high-speed satellite internet, showing this approach is a viable solution for "anywhere on Earth" internet.

For all those people who don't have broadband available, this will be life-changing.

It'll only get better from here, welcome to the future :)

(and remember that perfect is the enemy of good)

I’m still not sure we won’t get develop Kessler Syndrome (runaway debris situation in LEO) so let’s hope you are right.

LEO is not as much of a problem as higher orbits. At the height that Starlink operates unpropelled debris falls out of the sky relatively quickly due to drag.

The greater the surface area the quicker the decay. So if these satellites naturally fall after 3-5 years that means that if they somehow collided with each other creating millions of pieces of debris they would clear out with in less than a year because of the loss of energy during the collisions and the increased surface area.

I wonder what happens on a cloudy day.. currently the satellite based DTH television displays a cloud and refuses to work. Technically is there a work around to that?

Cloudy days will 100% affect internet speeds. But it's also possible for them to have better equipment at LEO due to lower launch costs to overcome that affect.

It's kinda funny because cloudy cover are a smaller portion of signal loss at GEO, (most loss is due to distance), but it affects that signal quite a bit once satellites are tweaked to get signal to Earth. Most GEO satellites try to have something like 95+% uptime due to average cloudy days in an area, which means they account for those really bad days where the clouds are just too thick. They use weather stats to project how much cloud cover each region will get so they can build bigger beams for that area.

In the past those beams would cover large areas (Like the entire USA). Now they're getting smaller (State Sized) so they can tweak how much signal could go to say Washington vs. Idaho. They're always trying to make this stuff better. But those newer sats at GEO cost 300+ million dollars per sat and are expected to last 20+ years. So it'll be a while before this new generation of GEO sat is out there.

The workaround would be. More signal power at the satellite (costly), or you as a consumer get a bigger/more powerful receiving antenna.

Well, on a rainy day both LTE and point-to-point wifi here go bad anyway, so it's still an improvement (more signal routes = more chance one works).

People keep comparing Starlink to their ideal non-rural internet connection. That's not what it needs to compete with.

Existing satellite internet doesn't go out with clouds, and only with heavy rain.

Having this kind of network is a good thing but the idea that any kind of satellite or wireless can replace wires is just fantasy.

Delivering just 10Mbps to 100 million customers is over 900 Tbps. There is absolutely no way a satellite network can clear that kind of downlink bandwidth. And that's just a tiny fraction of the market.

Physical cables/fiber represent several orders of magnitude more capacity and always will. So satellite is great for remote areas or unique situations but absolutely not a substitute for running cables.

And even in remote areas, building out VDSL2 networks with relatively decent speeds (200Mbps down and 100Mbps up[1]) would be better, honestly.

Most homes already have a phone line and can support a DSL network... and in some countries a phone line is mandated by law, so the infrastructure already exists.

But, DSL is not nearly as "sexy" as thousands of satellites...

[1] https://en.wikipedia.org/wiki/VDSL#:~:text=VDSL%20offers%20s....

I mean, you're not wrong, VDSL in rural areas would be great! But it also hasn't happened.

I haven't personally deployed VDSL2, but I have deployed DSL and other data-over-twisted pair systems including recent G.fast, G.hn and MOCA. I strongly doubt that VDSL can achieve those kinds of speeds reliably at the distances and densities needed to provide this service to the areas that need it most. In rural areas the copper lines are too long and too old to achieve these speeds (but they're fine for voice/POTS, so there's no mandate to upgrade.) Twisted pair is also really susceptible to noise and crosstalk, so in more densely populated areas once you start doing DSL to a lot of customers whose cables are bundled together they start to interfere with each other and everybody's service gets bad. In both cases the problem is signal to noise ratio - long cables mean poor signal, density means high noise, and twisted pair copper just isn't designed to deal with either very well.

What about VDSL2 Vectoring (G.vector)? Seems like it mitigates most of the crosstalk problems - but I admit I'm not an expert on DSL deployments.

It seems VDSL2 can achieve maximum throughput at run lengths under 1000ft from the DSLAM, with maximum performance beginning to degrade after. That sounds sufficient for most of the people currently stuck with slow access right now in the US (more rural areas outside major cities, but not actually remote as-in nearest neighbor is 1 mile away).

Also... surely solving this problem is cheaper than sending thousands of satellites into orbit every few years as they decay.

> It seems VDSL2 can achieve maximum throughput at run lengths under 1000ft from the DSLAM, and maximum performance degrades after that over distance. That sounds sufficient for most of the people complaining of slow access right now in the US

I would guess that this isn't right, but honestly I don't have the data either. My rough guess would be that only around 5-15% of people have <1000' of copper to their DSLAM, and I from what I've experienced with other technologies the speeds degrade very quickly after the max length, and only even work up to the max length in absolutely ideal scenarios. What I've seen of current copper infrastructure in the U.S. is pretty bad - lots of hand twisted copper lines with the ends exposed and rusting out, for example - that kind of thing would add the equivalent of at least several hundred feet of additional cable, and is very difficult to fix en-masse.

Honestly my gut feeling is that you pretty much have to run new cables to something like 80% of the U.S. to get good Internet service everywhere, and obviously if you're running new cables you should just run fiber. Looking at the cost of that compared to launching satellites - and accounting for the additional benefits of the satellites, like connectivity in extremely remote places and on the ocean - I'm not sure the satellites don't come out ahead.

EDIT TO ADD: A comparison point on the 1000 feet number is that this is in the same ballpark as how close a 5G tower needs to be to get real 5G service to your home or your cell phone. Current 4G towers are nowhere near that dense, and it's going to take huge amounts of money to get them that dense, even in the most densely populated areas.

Part of the reason this type of network isn't already built, is likely most folks living in very remote places simply don't need or want 1Gbps internet (or even DSL) - or at least don't know they want it.

There are folks with dial-up connections still, unfortunately, and not always because they have no other option. The more technically-inclined people here on HN wanting to work remotely and stream multiple 4k videos at once - they aren't majority of, say, AT&T's customer base. AT&T won't build this out if they think 1 in 100 potential customers will actually upgrade.

This will change over time, as a generation of folks that grew up with technology age and move into more rural areas. Hopefully demand will drive these network upgrades over time, and make it make financial sense to run new fiber in rural areas.

This could even be expedited with some sort of National Infrastructure Bill to fund it today, subsidized by tax payers since the benefits are ultimately so great.

I still think that's ultimately the better (and long term, cheaper) option verse thousands of satellites being launched every few years.

I don't think you're wrong here, but I would phrase it a little bit differently. I've worked in a lot of pretty rural areas, and from what I've seen in the last 10 years or so even folks in very rural areas want a lot of the same things from their service as techies do. Also, even places like the Bay Area with huge amounts of exactly the HN demographic the residential Internet infrastructure isn't exactly great and isn't improving very rapidly.

My take on the situation is that what most people want is more Internet for the same or less money. That means the only way to get infrastructure improvements is with competition. If AT&T did a bunch of upgrades and then gave their customers better service for the same price their customers would take it, but the customers aren't going to pay more than they're paying now and they're only going to leave if there's something better.

True, but their first customer is the U.S. Army which tells me this networks primary purpose is the same as the original intent of the internet itself. A mobile network that is available and reliable anywhere. Civilians usually get to do all the beta testing and provide feedback that improves the technology.

Also many of the places I have been looking at moving to don't even have POTS lines. This would fill the gap where POTS and VDSL is not available. I would certainly use VDSL if that were an option and use Starlink as a fallback in a very remote area.

> Also many of the places I have been looking at moving to don't even have POTS lines

Where are you, if you don't mind telling. I had thought, perhaps incorrectly, that in the US it was required by law to have POTS lines run to all homes - which led to it being installed during construction.

The land I am looking at is in the south western part of Wyoming. Could you link the law? My understanding is that once a line is run, there is a process to decom the line, but that no line is required for a home. Many homesteaders do not have phone lines or on-grid power. The only requirements I am aware of are around septic systems.

Interesting. I couldn't find the exact law to link to, but did turn up this WaPo article talking about 4 states starting to relax the requirement.

> The universal landline requirement has been repealed in Florida, North Carolina, Texas, and Wisconsin. There, new homeowners have no guarantee that they could order phone service at affordable rates, consumer advocates say.

https://www.washingtonpost.com/business/economy/landline-rul...

That seems more like a plan [1] to ensure that people can get an affordable line if they want it. I can't find any laws saying that all homes require a phone line. It also appears that recent discussions in the government were about changing money allocation from pots lines to internet specifically. I suppose we would need a land developer here to chime in.

[1] https://www.fcc.gov/general/universal-service

I've dug everywhere and can't turn up anything definitive.

The language, however, of "guaranteed access to an affordable line" does seem to imply a TELCO will have to run a line to your property at your request? Without fees (since that wouldn't be affordable)?

Perhaps, in practice, this resulted in all new construction having lines run since that was cheaper than doing it ad-hoc upon demand.

> I suppose we would need a land developer here to chime in.

Yes, hopefully someone out there knows more! Honestly curious now, since I had always assumed this was how it was.

There was a loophole in the various provisions created to ensure rural connectivity that allowed for companies to advantage of the fact that independent and competitive local exchange carriers were able to bill higher per minute fees than average. This led to a boom in free conference and long distance calling companies that would be spun up to terminate at one of these destinations, splitting the fee with the telco.

I think they serve different purposes. Satellites serve areas where they can't run wires or it's insanely expensive for too few people. Think very rural areas and war torn areas. Imagine being able to get unblocked internet in China by using satellites? Or in Belarus. Or the middle of Amazon (rainforest)?

Will Starlink even be able to legally sell internet services in China? Would they not have to comply with CCP's censorship policies and disconnect people at their will?

I strongly disagree with CCP's policies here, but that's not for us to decide.

I was thinking of the opposite use case -- an American traveling to China for work and getting uncensored internet.

But I mean once a Chinese citizen gets their hands on the equipment, I'm not sure how China would stop them. I suppose they could have roving bands of listening posts looking for the uplink signals.

I'm sure someone could figure out a way to take payments for them. Many rich Chinese citizens already have bank accounts outside of China.

Pretty sure an American company deliberately subverting another nation's laws to make a profit isn't going to be looked well upon, even if it is unjust CCP laws.

Regarding an American traveling to China - I think you're still beholden to the host country's laws, no? You can't break their local laws just because you're from out of town.

Not to mention the punitive actions CCP could take against Tesla (as they try to build out a business there) to put pressure on SpaceX to stop allowing Starlink access in China. That wouldn't be fair either, but Musk is CEO of both organizations and I'm not sure the CCP would distinguish them separately.

SpaceX isn't going to piss off China. The satellites are really just repeaters to the local base station. Starlink will be firewalled off from the world just like the rest of China (when the terminal is in China.)

>I suppose they could have roving bands of listening posts looking for the uplink signals.

I think this is pretty likely.

In the US the FCC has a whole set of systems designed to locate interference sources [1], and I've heard stories that they're extremely good at it. They even have a whole facility in Maryland dedicated to satellite monitoring!

[1] https://www.fcc.gov/over-air-spectrum-observation-capabiliti...

People keep bringing that up as a common use case, but how do you propose that people in China or wartorn countries will get their hands on the modems they would need to use Starlink?

Or even afford it?

People earning an equivalent of a couple USD a day, or nothing at all, aren't going to be able to afford to pay some US corporation for fast internet.

Fast internet probably isn't even a concern in wartorn countries, honestly.

Let's be honest - Starlink is almost exclusively going to be used by people living in rural areas within developed nations.

Implying the US government won't pay for the cost of an open internet where people can search for 'bad people square'

I'm not sure I understand your comment.

Why would the US Government (aka. Taxpayers) pay for internet use in other nations?

We already have expensive problems domestically that people would rather spend that coin on.

Winning 'hearts and minds'.

You drop access to the internet as whole in places where the internet is suppressed or non-existent and the populace learns just how bad they have things then they might revolt on their own.

I wouldn't say I 100% advocate for the idea but thats the thought process behind it.

That would generally be considered a violation of another country's airspace, and some countries would even treat that as an act of war.

You would also have to deal with the real possibility that those countries would simply shoot down any planes observed dropping Starlink base stations.

nonprofits air drop usbs full of news in north korea, violations happen all the time

US government airdrops?

There are a bunch or articles that talk about people using balloons to send material to North Korea.

The Human Rights Foundation even collects flash drives to load them up with content to smuggle into North Korea: https://flashdrivesforfreedom.org

For private groups, or individuals, sure you can get away with this.

For a government to do this? How could that not lead to an armed conflict?

How do governments airdrop supplies into war-torn countries today? I’ll admit I’m no expert but I see headlines all the time of neutral nations sending supplies to other countries who are fighting a war and that neutral nation isn’t getting dragged into the fight.

Governments don't airdrop supplies into war-torn countries today, excepting designated UN or MSF refugee camps, and they do so with the knowledge of one or both sides of the conflict so that the supply planes don't get shot down.

I would very much like to see a headline of a neutral nation airdropping supplies into a country engaged in war without getting dragged into the conflict. I'm not aware of that having occurred in the past 4 decades.

Because if the government is the USA, who is really going to go to war with them over this?

You're not wrong, information/propaganda warfare is aggressive. Sometimes it makes more strategic sense then open warfare, and sometimes states can get away with it when they can't get away with open warfare.

Do people not understand that satellite internet already exists today?

We don't drop satellite modems into war torn countries now, and we're not going to start doing that just because Elon Musk needs to check off one of his PR claims.

They're really expensive and big. Also, we actually do sometimes drop satellite internet into places.

Unless the pictures on Google are wrong, a Starlink base station is significantly larger than a satellite phone and is at least as large as a field base station.

And at $2000/pop, they're at least as expensive as the equipment we're currently not randomly dropping around the globe because of the expense.

Possession of unauthorized satellite equipment is as good as getting caught red-handed at a dead drop. A paranoid security state wouldn't blink at executing every such person as a spy.

> And even in remote areas, building out VDSL2 networks with relatively decent speeds (200Mbps down and 100Mbps up[1]) would be better, honestly.

I live in the suburbs of Paris and the length of the phone line is still too long for VDSL2 (it's only available for distance less than 1 km) so I am not sure how feassible that is in rural areas.

Even if DSL can get there, reliability of the connection in rural areas is often very suspect. A tree falls over in a winter storm, a bad crash that takes out an exchange box, a pole just plain falls over due to shifting ground, or maybe construction takes out a main trunk... these problems happen in cities too, but in a city you have maybe 1 km of infrastructure to worry about. In rural areas you could have 100 or 200 km of infrastructure just waiting for a calamity. Even just as a reliable alternative Starlink would be a big boon to rural users.

> Even just as a reliable alternative Starlink would be a big boon to rural users

Starlink is susceptible to cloud cover, rainstorms, lightning and other things that would interfere with signal quality (or render it unusable for periods of time).

Seems buried lines (either POTS for DSL, Coax for Cable, or Fiber) is the best solution. We just need to figure out how to incentivize it being built.

As someone who lived 50 minutes inland in Florida, no way.

Our DSL and cable options were limited, Comcast said they offered cable, but when they got to the line near the house, the signal was so poor they were unable to make it work without 50k in labor.

Then ATT came with DSL, the connection has routine outages for hours at a time. There is a technician servicing the DSLAM pretty much every week, there is literally no way ATT makes money where I used to live. Everyone owns 20+ acres, there's basically 10 homes in 1 square mile. And then the Hurricanes... Every 4 years all of the lines get blown down and it takes ATT 1-2 months to get it back online... There is way too many advantages for satellite in rural areas.

If I had to guess, there was over 200 outages (anything over 10 minutes IMO is an outage) in the 6 years I lived there.

Sounds like you have experience with above-ground lines. I specifically said buried lines - which are mostly immune to environmental issues.

GPON done properly is much cheaper to operate than DSL or COAX, it's just longer cables, less equipment, less real state, less electricity consumption.

Digging cables is what's expensive. You can use telephone poles for fiber too, but it exposes it to more outages. You can secure the lines with steel cables but that ads cost, and if you are deploying to poles then you were looking for going cheap.

Given the cost of burying lines is mostly in the labor/right-of-way stuff, we should be incentivizing fiber all over the place.

Helps that pesky Huwawei/5G problem the US keeps complaining about as well.

In rural areas, we have LTE-based FWA.

10 MHz wide channel with 256-QAM can push 100 Mbps down.

I'd argue if you get 100 Mbps down, you're not all that rural.

European rural :) It is a valley, tho, and if there wasn't that one tower, there would be 0 bps.

The range just isn't there even for suburbs. The high speeds are only achieved with line length of less than 500m, and at 1600m it's similar to ADSL2... So not really a effective solution.

Economies of scale work massively in your favor in this situation. You definitely do not need to sell the bandwidth 1:1, and the more customers you aggregate the more you can oversell. With 1gbps upstream bandwidth you can comfortably provide 1000 customers with 10mbps Internet service and they will virtually never see speeds lower than 10mbps.

Your math assumes all 100 million customers will be downloading at 10Mbps at the same time, all the time. Obviously that is not the case.

Not "all the time", but at any one time - at the same time (eg: watching superbowl - but of course any kind of multicast/broadcast is a little different).

At any rate, it's mostly about the bandwidth each satellite has to route. I don't know what the likely number of subscribers for any one LEO sattelite will be? Will it really be on the order of 100 million notalso served by 4g cellular, dsl and other broadband?

Say for new York state its a large area and population total but the area where you might want/prefer satellite service isn't very densely populated?

Couldn't the satellites themselves cache content that are for world wide events like this?

Edit: Even further, if they're launching 12000 of these couldn't that be a CDN network itself?

No, because HTTPS caching is not possible.

What I hear you saying is, they'd need to build out an actual CDN service that individual companies would have to sign up and use. This service would allow Starlink to be able to use their own certs themselves or have the owner's certificate.

My understanding is Cloudflare's service works this way. I believe this would be a good partnership between both companies.

I'm guessing the other approach is similar to Netflix's open connect. Where Netflix is given the option to run what they normally run on their open connect boxes but in a virtualized environment on the satellites themselves.

Now, I know that wouldn't seem all that practical if the service was seen as something anyone could sign up and buy because it isn't like these satellites are the same as EC2 instances and it isn't easy to just rack a new one but my gut feeling is that a very tiny portion of content is a substantial amount of bandwidth.

Now that would be interesting - 42.000 satellites running a cloud computing platform in space....

Maybe there could be a "broadcast" mode where starlink users can "tune in" to a fixed number of "channels" which broadcast extremely popular content.

    There is absolutely no way a satellite network can clear that kind of downlink bandwidth.

Evergreen comment. Encoding and compression software and hardware continue to improve. The only constant in technology is change. Will we be able to do this sort of bandwith to LEO eventually? Yes. Can we do it this very moment? No one has tried.

Context is important, though. Network performance advances are always relative. If satellite fully-congested download speed improves 10x while hardwire link performance improves 10x, satellite will still be perceived as "too slow" by some segment. Our experience thus far has been that advances in applications tend to consume the bandwidth available. Whether that trend continues remains to be seen. (To be fair, I have never saturated my bidirectional 1GBps fiber link and I don't foresee ever doing so except for very brief periods of time.)

> Encoding and compression software

What does that have to do with bandwidth ?

Sending less data over starlink (which is not tcp/ip btw).

I’d you can make 10Mb suddenly only require 900k of actual raw bandwidth by say adding a fpga or something then you can increase throughout to the users.

Hopefully in 2020 most internet traffic is encrypted. If encrypted traffic is compressible you're doing it wrong.

It would help if it was compressed pre-encryption, by the server.

If that big compression rates where achievable why wouldn't we be already using them? They would be standard in any network.

Technology only improves compression ratios to a certain point, and we passed that point with lossless compression decades ago. We're approaching that point with lossy compression now.

Going beyond that point requires the laws of physics (more precisely, the laws of information theory) to change, and that's not going to happen.

There are underlying physical limitations to spectral efficiency. The best modulation schemes in use today can hit about 60 bits per second per hz of bandwidth. Meanwhile, single fibers top out above 40 terabit. To duplicate the bandwidth of just a single fiber link with a radio downlink, you'd need over 600 ghz of available spectrum. That'd require advances pushing deep into the thz gap, but beyond that, the physics of propagation are so different between 6 ghz, 60 ghz, and 600 ghz that it's hard to conceive of a single system covering so much spectrum. And that's not even getting into issues with geometric limitations of radio beam widths.

While change is constant, this isn't an argument that any change you can imagine is physically realizable.

Fiber is here to stay for a very long time precisely because it starts out with such massive fundamental physical advantages. It's also a mistake to assume innovation only benefits one alternative. We're still learning how to make and run fiber cheaper too.

For what it's worth, satellite is typically around 1-4 bits per second per Hz.

> 60 bits per second per hz of bandwidth

Between one transmitter and receiver. Starlink is designed to cheat Shannon by utilizing SDMA on a truly massive scale. Every antenna in the system, both on the satellites and on the ground stations, is a large phased-array antenna, capable of simultaneously emitting different signals on the same wavelength in different directions, and capable of discriminating between multiple different signals on the same wavelength that come from different directions.

This is how Starlink intends to push so much throughput over just ~50 GHz of bandwidth. Not by pushing more bits per hz per link, but by filling the sky with hundreds of simultaneously visible satellites, and having each of them to all use all of the available spectrum at the same time on every link between any of them and any of the ground stations.

Phased arrays aren't magic. The directivity of the beam scales logarithmically with array size / element count. The limitations are such that you should think of it as being able to send Nebraska one signal while you send Kansas a different one. You can't get much more specific than that.

Elon has been very explicit about starlink never being competitive vs telcos in urban areas.

And independent people have done the math too. The system geometry maxes out at less than a single fiber: http://www.mit.edu/~portillo/files/Comparison-LEO-IAC-2018-s...

Phased arrays are not magic, but they are better than you make them sound like. When the directivity of the beam scales logarithmically with the element count, it just means you need a whole lot of elements.

The beam sizes are in the FCC filings, and they will be able to fully re-use the frequencies for ground stations that are >20km apart. (Not with the same satellite, you want adjacent beam spots on the ground to come from different satellites, but with literally hundreds above you at any time, that's not a problem.) Availability of bandwidth does not really limit them until the satellite counts are up to really hilarious numbers.

The aggregate throughput numbers in the paper you linked are correct, for the initial FCC filing. Since then they have multiplied the size of the constellation by ~7 and the throughput per satellite by ~3 (for 2.0 satellites). (ISL's are still missing, though.)

> Encoding and compression software and hardware continue to improve.

But these would apply to both wired and wireless, no?

If encoding tech is specific to wireless then not really , just like improving cable medium , number of cable pairs etc will not improve wireless , better transmitter or receiver will not impact wired

What encoding technology is specific to wireless? (Hint: none)

You are thinking just software , Hardware tech is very different for wireless. The benefits from beam forming, MIMO or any of the advancements in 3G/4G/5G only benefit wireless. There has been similar technical improvements in wired which only benefit that medium, they keep cramming more signal on the same size cable every generation.

Even in software at L3-7 the encoding tech is probably same of course, however protocols L0-2 is very medium dependent. The loss , noise, signal strength patterns all change depending on the medium so does

Even at L4 and above UDP based stacks benefit in wireless compared TCP based transport, if your application can handle it of course. The next gen HTTP is getting designed around UDP in some part to reflect the changing way people access the internet , a lot more traffic is mobile and wireless today so designing around wireless makes sense.

Personally even as app developer who works only with L7(HTTP) and L6 (SDP) I have specifically developed non trivial optimizations, workflows and APIs for mobile clients which were not needed or used by web clients due to different connection parameters of wired/ WiFi as opposed to wireless 2G/3G .

So yes there is lot of difference and most people working at any layer cannot ignore the medium.

I'm very aware of MIMO and the other physical layer technologies. But the original topic was about how wireless could somehow progress faster in these metrics than physical mediums. Mediums like fiber have progress much more in the last decade by adding more wavelengths and different modulation technologies compared to space technologies. SpaceX is only using MIMO in the sense that they have a phased array antenna, but that's not the same as your typical wireless router at home with multiple antennas. The other issue is space is severely power limited, so they are not going to get really high modulation schemes that fiber or cable have enjoyed (256+QAM) for many years.

> Delivering just 10Mbps to 100 million customers is over 900 Tbps.

50 Gbps * 12000 satellites is 600 Tbps.

Most of those satellites will be over the middle of nowhere at any given time. But those customers don't use 10Mbps 24/7 either.

It also may not be limited to 12,000 satellites. They're seeking approval for up to 42,000.

Oct 2019: "SpaceX submits paperwork for 30,000 more Starlink satellites"

https://spacenews.com/spacex-submits-paperwork-for-30000-mor...

100Gbps microwave links have been shown on land. If you take that as a theoretical limit, 12k satellites can give you 1200Tbps.

That's not comparable to sending a signal from space.

One satellite cannot route 900tbps of course , just like your local ISP cannot handle that either .

However over 40,000 satellites that’s just 22.5 gps per sat , quite achievable

And yet such a satellite network is being built and deployed.

The people building this have obviously done the math and have a solution which they're implementing while you're busy saying that it can't be done :)

I'm pretty sure it could with that DIDO technology that Rearden Labs produced. The think the sub company they started it under was Artemis. I always thought they're tech would work very well with satellite constellations.

Artemis was vaporware. It never made it into the field.

Vaporware is not an accurate description. I'm pretty sure you they have manufactured hardware that you can deploy for a test deployment. Also pretty sure they're working with the military for weaponization and wireless power transmission.

They were supposed to do a large rollout with sprint, and it never happened. The technology itself was vaporware; it only worked in their lab setting.

10Mbps to 100 million customers equating to 900 Tbps would be more akin to serving 1 billion customers, since on average someone isn't even close to saturating their bandwidth.

More useful to me would be bufferbloat [1] tests. Any starlink beta testers on HN? Also useful would be nuttcp or iperf3 tests just in case they optimized / prioritized for sites like dslreports and other speed test sites.

[1] - http://www.dslreports.com/speedtest

Yep. It's pretty much a given that if you want to compare Starlink to a good terrestrial ISP, you'll be disappointed. But compared to any other satellite ISP, Starlink should be impressive, especially in terms of latency. So testing and media coverage of Starlink should focus more on latency than throughput, and testing latency properly means testing latency under load.

... and the network only has a handful of beta testers on it at the moment, so any load tests are going to be meaningless.

Some of these links start a speed test on page load beware. All of them collect IP info/metrics about you.

   https://fast.com/
   http://speed.ui.com/
   http://speedtest.googlefiber.net/
   https://www.speedtest.net

"...that said, we’ll make sure Starlink has no material effect on discoveries in astronomy. We care a great deal about science." Elon Musk

A long exposure of Comet NEOWISE by Photographer Daniel Lopez: https://imgur.com/a/rDI1Onn

N.B. The white streaks are Starlink constellation passthrough trajectories

Starlink currently has 597 orbiting satellites they plan for more than 12k to be deployed. Amazon just got an approval for their own constellation of another 3k low orbiting satellites. Facebook and Alibaba will probably join this trend.

Do the satellites in the photo have the new shades on them? I'd be interested to know how much the situation has improved with the improved anti-reflection.

Do they plan to retrofit the coating on the +500 deployed?

Are there regulations that prevent future constellations from Amazon or Facebook, or any other giant, to implement albedo reduction on their constellations?

Those won't be up forever. They're all in a decaying orbit. They will get replaced in about 5 years.

As has been stated many, many times, the satellites are in very low orbit and naturally degrade over time to burn up in the atmosphere. The ones in the first batches that do not have a shade will only be up for a few years (5 at most).

The visor satellites are supposed to be invisible to the naked eye. They're still orbit raising, so we don't have results yet.

A non-reflective coating on the satellites will not stop them from blocking the light from astronomical objects.

The concern astronomers raised was blooming caused by oversaturating their sensors, not blockage.

Once we can colonize Mars and set up astronomy observatories on Mars, Starlink satellites orbiting Earth won't get in the way of astronomy discoveries.

The starlink satellites orbiting Mars might!

The photographer has admitted that he used stacking to create the image.

But isn’t stacking used for astro-photography all the time?

Yes. The same tool can be used to either remove or emphasize streaks.

Yes, it absolutely is.

This image was pretty thoroughly panned as being either ignorant or malicious - there’s as much light coming off planes in the sky, yet astronomers don’t complain about that. That’s not to say the satellite aren’t irritating, but astronomers have mitigation mechanisms.

SpaceX doing well is also a net good thing, because it’s getting cheaper to put the telescopes in orbit, which is where we really want them to be. Earth’s surface has light pollution from cities, planes and drones; is dark enough for astronomy only a fraction of the time; and has constant occlusion from birds and bees.

> there’s as much light coming off planes in the sky

If you live right outside of an international airport, maybe, but there are areas where this is not true at all. Observatories are strategically built to be far away from such light pollution. StarLink encircles the globe, and there is no way to build around it.

That NEOWISE photo was 17 photos deliberately stacked to emphasise the problem rather than minimise it, so it's not really a fair datapoint. Astrophotograhy stacking software will generally remove items that don't appear in all photos in the stack rather than retain them (previous discussion on the topic: https://news.ycombinator.com/item?id=23926699). That said, I love taking photos of the night sky myself and had some Starlink satellites affect a single exposure photo of mine recently, so it is a real issue - just not quite as crazy as it looks from that photo.

> it is a real issue - just not quite as crazy as it looks from that photo

Aren't they planning to add a bunch more satellites?

Yes, but they're also trying to darken them so later launches should end up significantly less problematic than the initial ones[0]. I'm not trying to defend them here (in fact I'm quite annoyed from a photography perspective) but appreciate there's two sides to the story and they're at least trying to minimise the problem.

[0] https://www.theverge.com/2020/3/24/21190273/spacex-starlink-...

Since the ephemerides of all starlink satellites are known, shouldn't it be possible to program these into the camera and not expose the sensor when a satellite is overhead? Makes it more complicated of course but this can let everyone win

I just got back from an AirBnB where I worked remotely for a week and the internet was DSL. It was about 1.5Mbps up and down. Work was still possible. Even group video calls worked.

Uploading and downloading large files was out of the question however. I had to do a lot of work SSH'd into a remote server.

So, 11 to 60 is not great, but still useful.

How was the LTE signal with a good quality antenna? LTE is already faster than this, assuming you can get at least a half signal.

Of several friends, we all got about 1.5 to 2mbps over LTE there. But the ping was worse... So I used the DSL :)

Ya, the 11Mbps side of it is very disappointing. That's pretty in-line with slow DSL and existing satellite connections.

The latency will be interesting to see too, since that was the boldest claim Musk made with this venture. The article seems to have cherry picked the best speedtests to show in their graphic - I'm curious what real-world in-application latencies are, ie. inside Call Of Duty Modern Warefare or something similar.

11mbps is amazing for what this is. Sure, it's "slow DSL" speed, but if you can get slow DSL you don't care about satellite internet. This is for all the places in the world that you can't get 4G cell service, let alone DSL.

11mbps plus a ping time of sub100 ms means that most people could telework anywhere in the world, even in the middle of an ocean. Unless your job requires you to download/upload huge amounts of data, it's more than enough for surfing, working, zoom, etc.

OK maybe, but Elon promised 1Gbps speeds. 11Mbps is very, very far cry from that.

Ok, but at the same time this is the BETA period. Attacks on his claim should wait until the public availability of the service begins.

Fair enough... except everyone here is in awe of 11Mbps satellite service, as-if satellite internet is some brand new thing.

If Hughes had been serving 11-60 Mbps at 31-94 ms ping without caps, Starlink wouldn't need to exist. Starlink is absolutely not the same as the previous generation of satellite internet. This is a brand new thing.

OK, but they've touted each satellite in the cluster being capable of driving 20Gbps all by itself.

So why are people only getting 11Mbps? Beta or not, that's a very far cry from the promised specs.

11 to 60 is actually really good. If you look at the total throughput on a residential connection, even a gigabit fiber connection, they will almost never burst above about 30mbps download. For most people 4K streaming is about the most bandwidth intensive thing they'll do, and that only requires about 20mbps per stream. As you indicated, video conference streams really don't use that much bandwidth (although they are very sensitive to latency.) The only real exception to this on residential networks (even during Covid with everyone working from home) is big game updates.

When I worked on oil rigs in Saudi Arabia, ~300kbps seemed to be the minimum necessary for WhatsApp/Facebook messenger/low-intensity web browsing like old reddit or arstechnica. HN might work a little lower but probably not much before HTTP fails for other reasons (if your speeds are that low, likely your packet loss is far too high to keep things working over 2G/3G/4G).

Truly even 0.3Mbps made all the difference in the world to the people who had been stuck on those rigs for the past 30-800 days. Often it was the first time in over a week, sometimes over a month that they could talk with their families. It became a necessary part of my job to spend 6-12 hours/day rationing access to the WiFi hotspots so that each of the 60 or so personnel could talk with their families each day.

I purchased with my own money all the WiFi modems and a set of 8 different antennas for various frequencies, with very long cables. Generally took about 2-6 hours to align everything correctly to get any usable signal. Used public cell tower databases to find the GPS coordinates of the nearest towers and used vectors to those positions as starting points.

Some of the locations I never could get any signal and no one could talk to their families or keep up with the news. Hopefully Starlink eventually connects these people to their families on a much more regular basis.

What i'm more interested in is longer term (well, longer than a speedtest) consistency. 11mbps is fine if it's never less than 11mbps. Aiming for "20ms latency for gaming" etc is useless if the thing drops out even once in a game.

It should be fine unless a helicopter hovers over your dish, wind makes your dish unstable, the satellite reaches end of life, weather affects your frequency(rain/fog/snow), or predicted sun spots occur.

The price may be concerning if we compare to the current offers of competition. There are caps and bandwidth limitations. Satellite time is expensive.

If a helicopter is hovering over your dish, close enough to be noticeable amounts of wind (more wind than you say get normally)... you have bigger problems... Like a military group is fast-roping into your home.

just to do some napkin maths, the entire constellation is supposed to be 12k satellites. Some googling gives 20Gbps throughput per satellite, so that'd be about 240k Gbps per second.

Now of course the satellites keep exchanging information themselves and not all satellites are everywhere all the time but ignoring that, and let's say you can assume ten times as much bandwidth because not everyone is online at the same time, that still doesn't serve a lot of people right?

If you go with their eventual Gigabit speed claims that'd work out to about 2 million customers if I didn't make an error. That's only a single digit percentage of the American rural population let alone anywhere else in the world.

Bandwidth is over provisioned at 50-100x the amount. Sometimes even more.

If that’s the case, assuming 50mbs at 50x overprovisioning and $50/mo average, the 12k satellite system could provide service for 240m people and generate $144B/yr in revenue.

I don’t think that would still apply. I’m in a rural area and our local ISP speeds go down to 1mbps in the evening, from a supposed 15. Streaming services are used by a lot of people at the same time.

This will not be competitive with existing broadband offerings. Especially in competitive markets. Some people pay 70/month or less for 1GB currently in good markets. This will be competitive where the cost is currently 100/month for much slower connections.

Gigabit speeds are great, but 80% of the value of Starlink is just getting the 11 Mbps to the people who can't get even that speed in any better way. There are many (millions?) of Americans who don't have that. So let's calculate from that number (or 10 Mbps, to make it easy).

240K Gbps * 1000Gb/Mb / 10Mbps/customer = 24M customers.

I've heard that in most residential cases, people on average use about a tenth of their bandwidth. That gets us up to 240M customers. That's only about a tenth of the world's rural populations [1], but that is still pretty great.

[1]: https://www.un.org/development/desa/en/news/population/2018-...

Also, rural endpoints can be shared among multiple households, AFAIK small African villages often share one WiFi hotspot, and sometimes one phone/tablet/computer.

No, there are not millions of Americans who don't have that. Both viasat and Hughes provide 25Mbps everywhere in the USA, and up to 100Mbps in some cases. Your math is completely ignoring the fact that 80% of the satellites are over NOBODY at any given time.

Is anyone concerned about the effect this has on the night sky and astronomical observation?

https://www.skyatnightmagazine.com/space-science/starlink-sa...

Nope, we destroyed the earth already. Why should we worry about the sky?

My parents house has 2mbps DSL from ATT.

This would be a massive upgrade, and for petty much anyone in rural America 1.5mbps has been the dream for years now.

The number of grey market installs of this where people will use _any_ available US address to get the equipment and then take it back to wherever they are?

Wow. This isn't just rural USA, this is rural Canada, Europe, South America, Africa. Anywhere and everywhere.

Think back to US sat-tv providers selling into Canada. Think back to the original iPhone selling internationally.

Regulators be damned.

>The 35 best cities in the world for online gaming have ping rates of 8 to 28ms

That statement makes absolutely no sense. That would assume the server is always in the same city as where you are at, which is about never the case when you play different games.

Won't these speeds dramatically reduce once there are a lot of users saturating the connections?

If the network keep up in this state probably. But the plans are to launch some thousand more satellites

>the plans are to launch some thousand more satellites

1000 more in that shell, but overall 11,578 more satellites (most of them lower than the current shell). Their target number is 12,000 satellites, with possible expansion to 42,000 (!!)

The sheer scale boggles the mind

I suspect it will be really serving areas where that is unlikely to happen (rural areas and over ocean transportation).

I'd switch to Starlink JUST to spite Comcast.

If you have comcast you're probably in range of a T-Mobile tower. Maybe consider their fixed 5G instead. It's coming to my area and Comcast started launching a flurry of deals to try and lock existing customers into a two year contract.

While I totally agree with your sentiments, there is a large section of the population that don’t even have the bad option of Comcast. I hope they put some curbs on adoption by those who have alternatives. Source: farm owner in the middle of nowhere KS. We have fixed wireless, but if it was hilly here we’d be stuck with Hughes and their geosynchronous satellites. The latency is huge - the speed of light isn’t fast enough.

Fellow middle of Kansas guy here. Check for fixed wireless in the area. We just got KwiKom which handily beats CenturyLink DSL on price, speed, and latency. And our neighbors couldn't even get DSL.

Oh I see you already mentioned fixed wireless. You aren't me, by any chance?

It's just a limited satellite network right now. I suspect bandwidth limitations may be due to limited ground station exposure.

There’s no way to know how much per-user throttling is currently in place. E.g. a single connection might be limited to 0.1% of a given bird’s total throughout. There could also be artificial limits for the ground station‘s uplink.

Current belief is that each satellite can forward 20Gbps max. But I’m doubtful that a single client could ever access a significant fraction of that total.

Yes, they will.

I recall when Iridium was going up, there was discussion about the complexity of the software that managed the pattern of transmitting packets among the satellites. The description indicated it was a complex software problem and they were struggling to get it right. (I went looking for articles from the time period, but couldn't track one down, however, came across MIT/Motorola FCC application for what would become Iridium.[1])

It's likely(?) that the orders of magnitude more Starlink satellites, their LEO speed, and other factors that impact optimization (amount of data to transmit, type of data, current load on intervening sats, positional changes during transmission, handoffs, etc.) can make this pretty hard. Or... It could be a learning system via neural net, etc.

The notion that the system could get faster and more efficient over time, simply by operating and without the usual trudge of scheduled software releases and upgrades, is an intriguing possibility. Tho somehow, that's not a novel notion...

[1] https://web.mit.edu/deweck/www/research_files/comsats_2004_0...

IIRC, Starlink satellites operate at a much lower orbit, are much more numerous and are a lot closer together which reduces the tracking complexity a fair bit. Also the inter-satellite links on the Iridium network are radio whereas the Starlinks have 4 motorized optical links (laser). They're highly reflective objects against the background of space.

As far as I know, the current satellites don’t have any lasers for satellite to satellite communication yet. That will come in later generations.

Ahh, I had to look this up but you are correct.

That doesn't reduce tracking complexity. It makes handoffs more frequent and complex as the OP was pointing out.

What happened to the Iridium satellites?

After multiple bankruptcies, primarily offering nautical communication services.

https://www.iridium.com/products/iridium-go/

The latency here is the main takeaway for me. It's phenomenal for a space based system.

I spent some time at my parents cabin during the quarantine and the only option there is satellite internet. SSH with a ping of multiple seconds is rather frustrating. That and all the systems which have bad bandwidth because that much latency makes them think they've lost connection.

As my parents are both techies, they are desperately hoping they're accepted into the beta.

Once they have inter-satellite links I'd expect long-haul latency over those links to beat cable latency for equivalent hauls.

What is the pricing?

No one ever seems to answer this all-important question...

you already know why.

https://news.ycombinator.com/item?id=22940781

Surely less than $100 a month, which is how high some existing internet can be anyway

Shotwell has said that it will be competitive with existing Internet pricing at $80. She didn't say it would be $80, but competitive with that pricing. Read that as you will. I'm sure that's only the starting price, that you'll be able to pay a lot more for more bandwidth.

I wonder if it will have location-modulated pricing; $80 wouldn't be competitive where I am.

If anything, I can imagine them ramping up prices in higher-density areas (where other more competitive broadband providers are available) to avoid oversubscription.

$100-150/mo for satellite based service would be totally reasonable (depending on data caps, etc). When you figure the average home (wired) internet connection is in the $80/mo range, cell phones start around there too.

The people who need this service are the ones whom are so far out that they have no option for DSL, Cable or even 4G tethering.

Because that's what will make this whole thing fall apart. A motorized phased array antenna isn't going to be cheap to recoup.

Pings and upload speeds seem impressive (although as someone else said we'll see when under load). I come from a place (rural Canada) where a non-trival amount of people who live far away from town centers do not have access to decent internet. This will be a good alternative to existing satellite ISPs, or 5mbps (on a good day) land ISPs.

Also potentially amazing for people living off-grid.

I'm fairly confident that the Canadian FCC will find a way to outlaw starlink judging by their track record with existing telecom companies & banning international investments.

It will also put pressure on land based ISP to either give up customers, or pay to increase infrastructure to get comparable internet to those areas.

https://www.youtube.com/watch?v=giQ8xEWjnBs

According to this Engineering Explained video, it may be possible to achieve a lower latency using Star Link than over fiber if the points are far enough. People like high frequency traders might be willing to pay a ton for such a thing.

This is awesome to hear. This will enable more people to move out to more remote places and get some nice speeds. 60MB/s is good enough for most remote work and TV streaming services. I can't wait to see what the major players will respond to this if it's opened up to major cities.

Latency runs 31 mS to 70 mS in the speedtests shown. I don't buy their claims that you can game with >100 mS latency, that is not a good experience.

This is a great step forward for internet accessibility but I really wish they wouldn't oversell the capabilities before we're there.

As someone who lives in the middle of the ocean with ~70ms to the nearest major data center, you get used to it.

You’re at a disadvantage to people with low ping, but it’s certainly not unplayable.

It really depends on the game. I would say 70-100ms is the upper acceptable limit for FPSs (though some like CS:GO or Valorant require even lower), but even like 150-250ms is okay for casual MOBA play.

How low will it get once there's 100 (or more) people connecting from a single neighborhood?

High-quality satellite internet that isn’t hamstrung by ridiculous download caps is one of the things I’m waiting for before attempting a nomad lifestyle.

I can’t wait until it’s more available.

Some states in India ( Andhra Pradesh ) has started infrastructure projects for laying down fiber optic cables across, connecting all villages( nearly all), for both cable, telephone and internet services. This had tremendous impact on internet speeds in rural areas.

I am writing this message from one of the remote areas with roughly 60Mbps connection.

Can anyone tell me why this was not previously possible? Both latency and speed? What did Starlink do differently

I believe the innovation is that the satellites sit in low earth orbit.

Traditional internet satellites operate in geostationary orbit, which is around 35,000km above the surface of the earth. This was preferable because with only a handful of satellites you could give the entire globe internet connection. However you have much higher latency due to the distance. Not exactly sure why the bandwidth is so slow, distance probably doesn't help here either. Maybe those companies are also running outdated systems.

Starlink satellites operate in low earth orbit which is around 400km, which is way closer, leading to far lower latency. However, at that distance you need to basically cover the sky with satellites in order to always see one. That's why they're planning to launch thousands of them.

Another reason the lower orbit wasn't possible before is that atmospheric drag actually causes the satellites orbit to decay much faster. So they must be constantly replaced.

That's not an innovation. Iridium was doing that 25 years ago.

It is possible and has been done before in the past. The companies that tried it went bankrupt because the business plan doesn't work out. The ground stations are too expensive for end consumers. Starlink will likely see the same fate unless they get government money.

That's not necessarily true. SpaceX has a massive advantage that previous companies didn't – launch costs.

SpaceX is getting all of these satellites up for an insanely low cost.

As pointed out in many places, the launch costs are still being paid for by their lack of funding from other commercial launches. The bulk of the cost is not in launches; it's running an ISP and equipment.

My understanding is that the FCC is very partial towards existing ISP. If I'm correct, how likely are they to negatively impact starlink accessibility to end user? Do they even have such power?

I would very much appreciate if someone more knowledgeable could give her opinion.

The FCC prefers 5G. Starlink will likely never be able to use that spectrum in certain areas.

I super excited for the students at my small rural school district. This will be a game changer!

IMO, the throughput isn’t what’s impressive, you can get similar-ish speeds from hughesnet. What’s impressive is the latency. The bandwidth should be (relatively) easy to increase over time. If the latency had been 300+ms, that would’ve been a disaster.

This is better than my Verizon data plan. The only time I can get decent speed without any timeouts in most of the Bay Area and even Tahoe is either when no one is awake or during the lockdown. Otherwise, it's Verizon data is near useless for me

Does anyone know if these speeds will scale (well, remain constant) as the number of customers increase? For instance, the 4G network can and does get overloaded in crowded places, could the same thing occur to satellite internet?

I want to be a beta star link user - I am not in a totally rural area but in my area there are frequent natural disasters that take out power and comms.

Cell sites work until the generator runs out of fuel, typically 2 or 3 days.

11 mbps in the middle of a rural location in the heart of Chile ? Count me in!

If someone knows - what kind of cable does the unit use out to the antenna? My COVID project is wiring the house with Cat6 and it might be worthwhile putting an enclosed drop out on the deck for the future.

According to this Reddit thread, the dish only has one cable presumably with POE

https://www.reddit.com/r/Starlink/comments/hqu7p6/data_minin...

I followed a link from there to the FCC test reports. Pages 7 & 8 of this application doc appears to make it out to be standard POE:

https://apps.fcc.gov/eas/GetApplicationAttachment.html?id=48...

> Power Supply Rating 56Vdc from PoE adapter 0.3A

802.3af won't do that at any sort of appreciable cable length, so it's most likely 802.3bt.

Page 13 indicates it has a standard RJ45 connector, probably on the router to the customer's network. It doesn't explicitly say it's RJ45 to the antenna. But they're pretty ubiquitous and weatherproofing glands exist that will allow a RJ45 to pass through them. So .. probably.

Theoretically one family could even get 2 starlink services if they really needed faster link. No physical line is used so I can totally see some people doing that if they really want to and can afford it.

I wonder how easy it would be to build the antennas into a laptop screen. Though heck, adding a "sleeve" to my machine which meant it just "always" had internet would be amazing.

For reference these are LTE category 3 speeds with better than LTE pings. A good start but nowhere near the Gb they've promised. I'm cautiously optimistic.

Better than LTE pings? That makes no sense to me. LTE ping should be <25ms if the target server is close to where it terminates which it typically is with speedtest.

> better than LTE pings

If you're getting latency worse than that on LTE, consider changing provider (and checking that you're _actually_ on LTE). These would be reasonable for the better class of 3G (HDSPA) but not great for LTE. Just did a speed test on my phone with wifi switched off (to a server in the same city). 20ms.

Just ran speedtest:

Wifi = 97 Mbps down 9 (xfinity) LTE = 57 Mbps down (att)

Location = SF Bay Area

11 to 60 Mbps global coverage is totally acceptable speeds and if it holds up, would be a game changer.

I can't wait to go live in the wilderness with this. seems like pretty good speeds, let's see what price it will get. Not in USA tho...

This would have been awesome in Alaska, where we had only one option for ISP... we weren't in the boonies either. We just barely got DSL.

Plus you have it all to yourself. Anyone remember the early 3G "high speed" demos, when 3G was just coming up?

This is more momentum towards decentralization. Fantastic.

Little bit discouraged to see one comment stating 2mb/s for Canada, I wonder why it would be slower than the USA? Perhaps there is less overhead coverage there density wise?

I also wonder how you can ascertain upload/download speeds before signing up. I know traditional broadband over copper doesn't have this feature, but if feels like this could, if you have the terminal on trial.

Very interesting.

cox has a monopoly in My area of California. The download speed of 50 mbs we have is fine but expensive. The upload speed between 2.5 and 3mbs is what’s painful.

I’d have to pay around $70 or more to get those starlink upload speeds with cox. I’d much rather pay starlink.

we've been able to do this for years using radio-based meshnets but the government won't update the legislation to allow encryption and commercial data on them making them pretty much useless. you wouldn't even need to pay for it.

The latency numbers look really good- under 100ms is similar to ground based connections

What's the limit for how many internet users can use Starlink at the same time?

50Gbps aggregate per satellite (actually 25 since you have to go up and down). 12000 satellites in the constellation. Those satellites will be over the ocean or unoccupied land most of the time, so estimate about 1% utilization ratios to be conservative. Standard ISP oversubscription allocations allow about 1Mbps per customer.

That math gives me 300M customers.

What hardware is required to use the Starlink network? Would that be expensive?

I've read somewhere else that its in the territory of $250. But that was reporting from a while ago, and things might be different now.

Will having Starlink prevent asteroid detections from ground based telescopes?

I hope they provide global coverage despite totalitarian regimes !

That's great. The pings not even that bad..

what about upload?

aykm?? wasted the sky with this ugly crap for what we already have with 4g? somebody please kick Elon in the nuts for me.

There isn't a usable 4G signal everywhere. Not even everywhere in the US. On my 2 mile commute in Kansas, there's about 1 mile where I cannot even place calls or send texts, much less use data.

Oh i see I have angered the cult. A rare negative submission points for me by daring to criticize the holy one.

Once all tesla vehicles use starlink for sending back data to Tesla HQ I'm not sure much will be left for 'rural' broadband.

Nice to pick up the grants though.

For comparison, DSL is 1.5 Mbps. 4g is between 5Mbps and 12 Mbps (up to 50Mbps in burst mode), while 5g generally starts at 50Mbps. Cable internet is usually 25 Mbps, and up to 1 Gbps if you're willing to pay, and fiber is up to 10 Gbps if you're lucky enough to live in an area with fiber built out to your home.

So in other words, Starlink is useful if you live too far away from a major residential center to have access to better and cheaper forms of internet access.

There aren't enough people in the US or EU that live in rural areas far from decent internet connections for Starlink to be more than a tiny fraction of the ISP market.

On top of that, unlike existing forms of internet access, Starlink is susceptible to weather conditions.

> There aren't enough people in the US or EU that live in rural areas far from decent internet connections for Starlink to be more than a tiny fraction of the ISP market.

Oh, I'm not sure that's true. 41,000,000 Americans don't have access to broadband [1]. That's the entire population of Canada and then some.

[1] https://www.bloomberg.com/news/articles/2020-02-19/where-the...

Starlink can't handle that many people, so it's a moot point that there are a theoretical 41 million potential customers.

You can build out landlines far more easily than you can launch satellites; you just need the cable/fiber and some guys to dig a trench. The reason it hasn't happened is because it's not worth the financial investment without government subsidies (and this is why the federal government provides some subsidies for rural broadband buildouts).

Iridium tried to be Starlink a decade or two ago, and they ran into the same problem: the cost of servicing rural and remote areas far exceeds the money you would receive from rural customers without massive government subsidies.

> You can build out landlines far more easily than you can launch satellites;

Each satellite costs SpaceX about $250K to build and $500K to launch and services thousands of customers.

That same $750K might buy you a mile or so of trenching and cable. Many rural customers need multiple miles each.

Century link quotes me about $5K / mile of copper (I assume). But they will only give me dial tone, not internet, because their equipment is saturated (rural Montana).

That same $750K might buy you a mile or so of trenching and cable.

If you want to make this about the economics...

All studies on this topic report that it costs less than $250k/mile in metro areas, which cost more due to the need to get permits, plan around utility lines and existing infrastructure, dig into asphalt/sidewalk, and then repair the damage.

In a rural area where you can just dig into dirt and not repair anything, $750k should get you all of a town and then some, and I'm assuming we're talking about one scattered out across tens of miles. You're primarily paying for labor, and it's possible for a small team to dig dozens of miles of utility trenches in a day with common construction equipment. Once you lay the cable, your expenses are essentially fully booked.

A single Starlink satellite, assuming maximum satellite bandwidth and minimal user bandwidth would service at most 2500 customers. At $750k/launch, assuming a useful life of 4 years (per Musk's Techcrunch interview from April 2020) and maximum customers, Spacelink would need to charge at least $6.25/month to each customer not accounting for operational or marketing costs just to break even. You would need to at least double the price to cover ongoing operational costs. Plus, Spacelink would need to pay other ISPs for interconnection agreements, which is another 1x cost there. Add in the costs of paying off R&D and other capitalized costs, and you're looking at a minimum of 5x the putative minimum price of $6.25/month, or roughly $30/month.

Or in other words, Starlink would be charging no less than its current satellite and landline competitors already charge today.

Your operational and marketing and capital costs apply to all ISP's, not just Starlink. The net present value of $6.25 monthly at a 3% interest rate for 20 years is $1100. For 50 years it's $2000. So that's your breakeven point. You might be able to wire up an entire town for less than $2K per customer but you certainly can't do rural customers for that.

Shotwell has said that Starlink will be competitive with $80/month internet.

Plus, if these satellites are mass produced, then I'd expect the price to drop because they figure out ways to make them cheaper.

> Starlink can't handle that many people, so it's a moot point that there are a theoretical 41 million potential customers.

It's a critical point in fact, because you're never going to capture an entire market, you're going to capture at best a modest fraction of the maximum market (for numerous obvious reasons).

Those 41 million potential customers are actually more like maybe 10-12 million stable or semi-stable households. The 41m is an incorrect number to go on, that represents the total number of people rather than the household count.

You're only going to get a fraction of those 10-12 million households no matter what you do. So now you're talking about probably at least 1m, up to maybe 4-6m households / accounts. Starlink can make a large dent in that.

DSL might only be 1.5 Mbps where you are but it is standardized between (at least) 0.7Mbps and 300 Mbps. I personally have used it at speeds between 2 and 100 Mbps.

The fact that people can actually switch from their current ISP to starlink is a big deal though.

Most people just use the internet for social media and video streaming, which starlink provides more than enough bandwidth for

People can switch from their current ISP now, unless they live in a rural area. They can choose between DSL, cable, or go fully mobile and rely on their 4G/5G connection.

Starlink's bandwidth rates only apply to their beta test. Think about that: with minimal users, speed caps out at 60Mbps! 4G can handle that even when the network is fairly congested.

> "4g is between 5Mbps and 12 Mbps (up to 50Mbps in burst mode), while 5g generally starts at 50Mbps."

Yup. I actually regularly get sustained speeds well over 50 Mbps on 4G with carrier aggregation (which some phones call "4G+"). And anything less than 200 Mbps is a bad day on 5G. (Vodafone UK network)

https://www.speedtest.net/result/9898742367

So, I'm on 4G right now, getting 45/25 Mbps, in what is considered a rural area (20km from the city).

I lived in Birmingham, UK, a way more populated area, and got ~40/30 on 4G, as well.

I noticed many people say they're lucky to get ~15 Mbps on 4G and I'm confused, why? Is it just congestion or is there some difference in the towers/implementation?

4G is slower in the US, largely due to congestion.

4G in the US often matches the parent's reference speeds, it depends on where you're at. People often make poor comparisons with the US, they context drop, comparing eg the speed across the entire US or a state versus their experience in one large city in a smaller more densely populated country (and then cherry picking based on a good outcome).

Cleveland, Minneapolis, Detroit, NYC, Boston, Pittsburgh, Cincinnati, Baltimore are the best in the US, averaging at or above 30mbps across the cities in question. SF, Philly, Atlanta, Kansas City were also competitive with what the parent referenced (eg the SF avg was 27mbps, meaning people are routinely seeing the parent's ~40mbps in Birmingham).

https://9to5mac.com/2019/05/31/state-mobile-network-speeds/

> "I noticed many people say they're lucky to get ~15 Mbps on 4G and I'm confused, why? Is it just congestion or is there some difference in the towers/implementation?"

It can depend on the network, the device, congestion levels, and location. Some networks are notoriously bandwidth-constrained (ahem Three UK) and will always be prone to bad 4G speeds. Some networks have added capacity on more exotic 4G bands that are only supported by relatively recent devices. If you're connected to a macro site in a major city it probably has fibre backhaul with plenty of bandwidth available, but if you're connected to a small edge cell it may have radio backhaul with less capacity available.

And newer devices will support things like carrier aggregation and higher levels of MIMO that get faster speeds.

Interesting site showing how the UK mobile frequency is allocated here: https://pedroc.co.uk/content/uk-commercial-mobile-spectrum

Note that only fairly new devices will support the more exotic bands like 32/38/40 (and, of course, 5G).

The US does have a lot of people in rural areas where 5G coverage won't be good. And even if you have good 4G coverage, take take a look at the 4G home internet plans from Verizon: https://www.verizonwireless.com/home-services/lte-internet-i...

For $60/month you get 10 GB of data. For $150 you can go up to 40 GB. Overage is billed at $10 for each additional 1GB.

If I did my math right, at 12 Mbps you can chew through the largest monthly data allotment in under 8 hours.

In practical terms, for $150/month you can't even download one AAA game, and even if you buy all your games on disc you'll probably still go over the data cap just downloading mandatory updates.

Starlink hasn't said anything about data caps yet, but there's plenty of room to compete with terrestrial broadband providers on things other than download speed.

I get that people are placing their internet hopes and dreams on Starlink being cheap, but the financials they've provided just don't work out to them being any cheaper than any existing ISP unless (a) Starlink receives massive government subsidies (b) signs up a fuckton of commercial partners/sponsors, or (c) runs the network at a huge loss.

They've provided financials?

They've indicated that each satellite costs about $250K to build, and launch costs per satellite are about $500K. Each satellite lasts 5 years. Each satellite gets 50Gbps (25Gbps actual since you have to go up & down). They haven't released base station costs, but there are some estimates that it'll be about $2K per. The other big unknown is the utilization factor.

It's pretty hard to punch those numbers into a model and not show SpaceX making money hand over fist, but there are bears out there that have come up with some really contrived models.

It's pretty hard to punch those numbers into a model and not show SpaceX making money hand over fist, but there are bears out there that have come up with some really contrived models.

At $2000 just to buy in to the system, they've cut off a huge chunk of their purported market.

Unless they plan on using unicorn accounting again, it's pretty hard to punch those numbers into a model and show SpaceX even making its money back on the first wave of satellite launches unless they dominate the rural ISP market. In order to just break even on fixed and operational costs they'd need to charge at least as much as existing rural ISPs, and if they charge significantly more they won't acquire significant market share because their are faster options available at higher pricepoints that aren't susceptible to the weather. And most of those alternatives have roughly the same buy-in costs if SpaceX actually plans on making customers pay $2000 for a base station.

And then they'd need to do it all over again in 5 years when they have to replace their satellites.

Reply on news.ycombinator.com