In the '80s, satellite comms showed promise – soon it'll be a viable means to punt internet services at anyone anywhere Terry Wogan has a lot to answer for. From 1982 to 1992, he presented an eponymous chat show on BBC1 where he would often interview celebrities via satellite in front of an unconvincing backdrop of the Hollywood sign. Back then, everyone had a single word to describe satellite communications: slow. The interviewer would ask a …
...to have the same data limited packages on a different but slightly laggier connection all forced through the cheapest and shittiest router the ISP can find that for some reason hasn't had a firmware update for 4 years, even thought the service is brand new.
I will be at the forefront of rebooting my modem and waiting 5 minutes when it doesn't work on a different platform.
I'm also putting a tenner on satellite services still requiring a mandatory land line subscription...because who wants a world without BT OpenLeech except everyone?
Please...please...please...I just want an RJ45 coming in to the house. That's all I want. At most, just a modem...not a router, not a smart hub, not a guaranteed fucking mesh network...just a modem with an RJ45.
To be clear...ISPs...just provide the internet for me. That's it...I'll sort out the router, firewall and the wifi thank you.
I appreciate, and agree with, the sentiment but let's be honest. The vast, vast majority of people wouldn't have a clue what to do with an RJ45 connector coming into the house or, worst still, would plug their laptop directly into the internet and live in ignorant bliss with a woefully insecure setup. So many people want a single box they can just plug straight in to their phone line, tap in the wifi password printed on the back and start browsing that it's not worth it for ISPs to cater to those of us who don't. They don't want the support calls from someone who ordered a slightly cheaper package because it didn't include a phone line and now doesn't know how to hook up their whole house to a single RJ45 socket.
What's wrong with Big Brother? Winston Smith was a subversive. The Proles weren't complaining and enjoyed the security and protection by the kindly efforts of Big Brother. Winston wanted to take those comforts away so the efforts of the statement were and are necessary to protected the masses.
Vox Populi Vox Deus
a license to use the Internet, something most often used in the privacy of people's homes? Can you say Big Brother? ...
Gimme a break!
Guess what? IT IS NOT your house anymore. We are in the 21st Century. But then again, it never REALLY was your house. The powers that be can crush you - and your house anytime.
So... You need a license to drive a car, don't you? And must pass a test?
Perhaps the internet should be the same way.
"So... You need a license to drive a car, don't you? And must pass a test?"
But at least the roads are owned and operated by the government, and people can be directly killed as a result of vehicle use. Can you say the same thing about the Internet, which is typically used in someone's home, doesn't involve a lot of physical interaction, and usually runs through privately-owned data lines?
And no, the government has a lot less say about vehicle operation on private property. Thus a seven-year-old CAN drive a car on private property, so long as the owner says so and other laws aren't broken in the process.
The problem is that, as America can demonstrate, anything like that can be abused to exclude "unwanted" people, thus why there are laws in the books to prevent their use. It's the whole "We Just Can't Have Nice Things" problem. And even then, there's no guarantee it'll stay up there given today's political climate, which is virtually No Holds Barred.
A colleague has recently had the BT residential fibre service installed (genuine FTTP, not FTTC), and he has no landline at all.
While the service does come with a BT Wi-Fi "router", you are free to not use that and just connect to the RJ45 port on the internal box with your own firewall etc, which is exactly what he has done.
We're all going to have to do that over the next few years. BT is turning off the wholesale telephony service in 2025 so everyone is going to need some kind of VoIP terminal and everyone will need a data connection, even luddites. Luckily VoIP bandwidth needs are low so most properties should be able to get a DSL connection that's good enough of BT relax their lower limits they impose at the moment.
Not to be a smartass, but how are you supposed to get a twisted pair copper connection for DSL when the PhoneCo is phasing out their copper networks entirely in favor of fiber optic ones instead? *Sad smile, hands you a pint*
Drink up, it'll help drown out the frustration... =-j
OOMA Telo box here.
FTTP, RJ45 connected from the ONT to the input on my router, OOMA connects to one of the 4 GigE ports on the router and a POTS line comes out of the OOMA box. I ported my old POTS number to OOMA and all works fine. The OOMA even survived the transition from Comcast (cable internet) to FIOS (fiber).
OOMA costs me $6/mo. I think telephone is free on FIOS if you take their TV, which I don't, and I wanted an ISP-independent VOIP provider
Yep, Spent 9 months after graduating dealing with people like that. The ISPs have to make it as simple as possible hence why you have the options you have and nothing like what you want. Even then there are people who struggle, which to be fair can be either parties fault. We ahave all had to deal with Virgin Media. Im sure you have heard all the stories before.
So they could hire an expert to set it up for them. Why should the average person think they can understand how IT works when we spend years learning and studying to make it work for them? If they had to hire a techy they would end up with a better service that is actually secure.
depends on the techy they hire.
i once worked with a UK born and bread global support engineer on a unix based product. He had absolutely no clue what he was talking about yet confidently spoke to advanced user customers in global giants about sorting stuff on our product. After a few days of him getting it wrong me or someone else would sort it.
lots of people have lots of "certificates" but are no better for it.
"lots of people have lots of "certificates" but are no better for it."
Yep, that's me. I have loads of certificates. I'm just really good at passing tests. I got them so I could get hired to do the tiny bit of the puzzle that I do but I'm not the person that could do the whole thing even if the paper hanging on the wall says I'm qualified. I don't know squat about RADAR (and know that I don't know), but I have a RADAR endorsement on my radio license. I got my amateur extra certification on my first sitting (passed all three test levels). I'm hopeless for most things, but I can set up high altitude balloon experiment telemetry which is what I got the cert for.
The other reason for getting the certs is I got to go to classes for a week instead of working and skived off most of those to play tourist if they were in a good area. I read the books and materials and regurgitated it back on a test form and passed. The company then gave me a pay rise for being more valuable, silly wabbits.
"I read the books and materials and regurgitated it back on a test form and passed. The company then gave me a pay rise for being more valuable, silly wabbits."
back in my telco days we had a major problem with this kind of behaviour.
Competent and capable techs were heavily in demand to fix things. The incompetent ones were put on a desk and told to read. Guess what they did and where they got promoted to whilst the busy techs never got a chance to consider career advancement.....
"would plug their laptop directly into the internet and live in ignorant bliss with a woefully insecure setup"
It's a long while ago but when NTL first supplied broadband to my address, they supplied a long RJ45 terminated cable with it. Which I blissfully ran through the house and plugged straight into my PCs Ethernet socket. I was so blown away by the speed compared to dial up that I didn't have a thought about security. Can anyone remember if fully wired NTL even came with a router?
Later a second computer entered the household and I bought my first WiFi router (before NTL supplied them). While on the phone to NTL CS I was told that more than one device wasn't allowed on the home network, when I mentioned my router.
"Can anyone remember if fully wired NTL even came with a router?"
Yes, they a;ways supplied a cable modem right from the start. No security built in, no router facilities included. As for your own security, if this was your first ever internet connection, then yes, you may have no security at all. But if you were upgrading from dial-up, surely you had some sort of defences already on your PC.
That is, in fact *exactly* what Starlink gives. They do supply a router as well, but I threw that into the network spares box and plugged the "dish" straight into the back of my pfSense firewall.
We've not had a landline here for, oh, about 8 years, I think?
GJC
>Please...please...please...I just want an RJ45 coming in to the house.
That's so last century...
With the country moving to FTTP, I'm happy with an LC UPC Simplex OS2 Single Mode connection. Then practically any router with an SFP port can be used. Okay currently getting the right patch cable and BiDi SFP module combination that works with both the ISP's module and your router isn't quit as straightforward as will be necessary for Joe Public.
I'm also putting a tenner on satellite services still requiring a mandatory land line subscription...because who wants a world without BT OpenLeech except everyone?
I'll take that bet. Feel free to send it over now - StarLink doesn't require a landline subscription and never will.
That's the point.
Many USA telcos are shitting rather large housebricks at the moment regarding Starlink and spreading all kinds of FUD. Quite frankly I suspect BTOR should be worried too (the current BTOR stance rabbits on about 400ms latency, etc - making it clear they equate Starlink with GEO service)
Starlink claim 300MB/s for £85/month. uncapped - which is on par with a G.Fast connection
I've been suffering from rotten copper in my area for 20 years. BTOR have made no move to fix it or roll out fibre early. They're likely to be one of the many companies that wakes up one day to find a 900 pound gorilla on their front lawn eating their lunch
Whilst Starlinnk may not be "urban suitable", it's certainly capable of coping with suburban housing densities even in its current guise. Once the extra shells go in, sub 20m ground terminal spacing in cells would easily be feasible
> just an RJ45 coming in to the house
Errr, that's what we get on this side of the pond. Actually, it's usually a coax (or fiber) that goes into a DOCSIS (or fiber) modem, that has an RJ-45 on it. You can rent a modem and firewall/VoIP/router/WiFi-AP/pigeonroost from the ISP or buy your own.
That's more or less what I have got.
Virgin dropped a co-ax into my home along with a modem to a RJ connector and that's me sorted.
Added my own router and ported my long held BT/Openreach POTS number to a VSP (not Virgin) and I am sorted - a 4G USB stick is on hand to take my mobile SIM should Virgin go down.
The Openreach cable still enters my property, and still has 'Non Connected' dial tone on it but it does not cost me a penny.
.. And the squarial certainly didn't.
There's a lot of effort being put into active phased arrays, not just for LEO but also GEO (to eliminate the need for a mobile subscriber having to point accurately). It's the enabling technology for both systems.
"the Starlink ones are genuinely fully 'steerable' arrays"
The ancestor of these al this is the steerable radar array in the nose of a fighter jet that replaced mechanically scanned radar.
https://en.wikipedia.org/wiki/Active_electronically_scanned_array
30 years ago something like Dishy would have cost upwards of $2-3 million apiece and been classified military technology (remembering the way certain countries reacted when commercial landmobile systems moved away from crystals to agile PLL oscillators, I suspect many governments would prefer it stiill was)
The USA is unusual in having a lot of people with money served by a worse-than-third-world private comms infrastructure. The upcoming services from Starlink, OneWeb, and the newer GEOs will be able to offer a competitive alternative, without any additional ground infrastructure. That alone is a market worth chasing.
That's not entirely accurate. The US is 11th in fixed broadband, 25th in wireless...on par with France and Switzerland. The difference is that in the US bandwidth is extremely disparate. Populous cities have very fast internet and rural areas have....well, none, or almost none. Rural areas have 5G and 4G only along major highways, otherwise some have grandfathered DSL The USA is abandoning copper too and new installs are generally not available. The LEOs should be a huge success in rural America.
the LEO market in the USA comes with a major issue for the rest of the world
Pricing Starlink so it seems "reasonable" in the USA compatred with current Telco pricing makes it hellaciously expensive for most of the rest of the world
If they lower the non-USA pricing too much then there will be howls about dumping and if they lower the intra-USA rate the telcos will be lobbying hard to nobble them.
It's only select areas in very major USA cities which have anything like reasonable price/competition and the vast majority of areas have legislated local monopolies
I wouldn't be at all surprised to see State PUCs attempt to drop huge surcharges on cord cutters in order to "protect their telcos"
Bear in mind that there is LESS competition across most of the USA than before the breakup of AT&T in the 80s. There are fewer LECs and alternative LD providers than there were then. Regutaory capture seems to be the US national sport...
You're quite a long way out on Starlink's bandwidth. I've just run a test and got 207Mbps down, 37Mbps up (and 24ms latency, which is nice). I've seen speeds up to 350Mbps down, and 60Mbps up, but not both at the same time.
It is, I have to say, a very, very nice system. We did have one of the original bi-directional systems using a geostationary satellite, and the bandwidth was astonishing for the time (14Mbps in a age of 128Kbps from ISDN, from memory) but the latency was also astonishing, at over a second. Quite usable for web browsing, as they had an intelligent proxy that bundled up the whole page and squirted it to the user in one big burst, overcoming some of the HTTP latency sensitivity. Telnet sessions were hilarious, though.
GJC
"I've just run a test and got 207Mbps down, 37Mbps up (and 24ms latency, which is nice). I've seen speeds up to 350Mbps down, and 60Mbps up, but not both at the same time."
It sounds like you are getting a huge chunk of the available bandwidth and only sharing with tens of other people. Multiply the users by 100 and figure that up to 5% will be saturating their connection with gaming and others watching 2 Netflix streams in their home. In California where there are lots of Starlink beta testers, the speeds are much lower.
And this is what makes a difference, along with the bandwidth (20-50 megabits per second)
Is that total available or the per-user average? Because it's 'adequate' for one person but if that's the bandwidth you're making available to, say, London then it's pitiful. This has always been a problem when using wireless for unicast services. Unless you use very tight and precise beam shaping it rapidly becomes a contended resource.
"Unless you use very tight and precise beam shaping "
They DO use very tight and precise beam shaping. The ground spot from each satellite is targetted to be less than 500m across. In reality it's more of an elongated ellipse but from 550km up and path lengths in practice being up to 900km long, it's pretty impressive
that kind of beam pattern and enough satellites allows a great deal of flexibility in terms of bandwidth, plus the dishes are multiband (12, 30 and 60GHz), so bandwidth constraints are going to end up being more driven by frequency regulation than user demands
This isn't your grandfather's [FT]DMA multiaccess radio system and I think the killer applications for it haven't even been dreamt up yet
Here (California, but basically anywhere in the US, I suspect), you can readily find those long lags (tens of seconds to single digit minutes) reminiscent of "Live via satellite" pretty much anywhere, from broadcast TV local traffic reports to video conferencing. Without having to leave the atmosphere. LTE "phone" service often has a whole set of weird echoes and multi-second lags that remind me of satellite calls to Japan back when. I have to wonder if some bright spark decided that San Jose to Corvallis was best relayed via Myanmar.
BTW: buying one's own Cable modem is a mugs game. As soon as enough people have done so, the ISP "upgrades" the service to require a new Modem, so you can't amortize the CapEx.
I call BS. I have a docsis 3.0 Arris I bought in 2015ish I picked up a 3.1 month or so ago as a backup incase this one gets tapped by lightning or something. Not gonna put in in cause I can't get 2x2Gig anyway. Future docsis is backward compatable, basically buys you extra carriers over the same bands. If you are amortizing tech > 5 years you are stupid.
Indeed. We did roughly the same thing. We had a Motorola Surfboard for ~10 years before that. The upgrade only happened because the Surfboard finally was unable to keep up with the base package's speed.
(OTOH, at least one large cableco will lobotomize your equipment so you may not quite get the features you expected.)
". I have to wonder if some bright spark decided that San Jose to Corvallis was best relayed via Myanmar."
Least Cost Routing is done automatically -- in the short term, it doesn't require human intervention at all. The only 'bright sparks' involved are those approving the cost of the telecom contract on the buyers side, and those approving the cost of the telecom contracts on the providers side.
It was because encoding was so slow. With a GSO satellite the round trip delay is a little over a quarter of a second. For a double round trip (ask a question and get the answer) it adds barely half a second, not noticeable in a typical Q&A style interview.
It is 80s technology that was the problem, not the satellite.
There was no video encoding in the 80's. The world was analogue back then, live and immediate. Double hop satellite feeds were often needed, ie. Europe to right coast US, then another hop to left coast, then another two hops to come back again. So, yes, easily upto a couple of seconds on a broadcast video feed.
True it was analog so "encoding" was the wrong word choice but it was modulated, and that modulation was not instantaneous.
It is unlikely that an 80s BBC chat show was regularly interviewing celebrities in Hollywood two hops away. Most of the time they'd be elsewhere in the UK, where it would be a single hop up and a single hop back.
"It is unlikely that an 80s BBC chat show was regularly interviewing celebrities in Hollywood two hops away. Most of the time they'd be elsewhere in the UK, where it would be a single hop up and a single hop back."
In the 80s, it's not very likely they'd be using a sat link inside the UK. Far more likely to be bonded phone lines specially set up for the "event".
"There was no video encoding in the 80's. The world was analogue back then, live and immediate. Double hop satellite feeds were often needed, ie. Europe to right coast US, then another hop to left coast, then another two hops to come back again. So, yes, easily upto a couple of seconds on a broadcast video feed."
Not to mention standards and framerate conversion. Depending on the time period, which end did the conversion, and which technology they used, it could easily have added another second or so in each direction,, eg scanlines going via delay buffers used as "memory" and then read back into the destination frequencies etc.
"The world was analogue back then, live and immediate. Double hop satellite feeds were often needed"
I remember working as part of the press uplink crew for Disneyland's 35th anniversary. We had up to 8 hops for some round trips. Really stupendous mileage for those signals.
"but nobody actually wanted it, so it never became a thing."
As an admin at a small ISP who saw his bonded T1s fill up when users discovered alt.binaries, I did, but there weren't enough of us, I guess.
Regarding Starlink, two notes:
1) 20-50 mbps is underselling it. Downstream I typically get 100-120, and have seen as much as 170. Upstream I get about 20.
2) the satellites weren't the biggest barrier, electronically steered antennas on the user terminals were the big game changing tech. With GSO satellites, you carefully align the user's dish to point at the satellite, and you're done. With LEO, the satellites are moving across the sky relative to the user. The user terminal needs to electronically follow the satellite, then quickly "steer" to the next one as the first goes out of range. This is done with electronically steered phase array antennas.
3) (I have *three* notes) the ability to track the satellites and run the phased array antennas depends on having enough computing power in the user terminal (with enough left over to route traffic).
4) (among my notes are...) the cost of all of the above needed to be low enough to be economically feasible. Current estimates are that Starlink sells each user terminal at a loss of something like $1500. They're working on reducing cost and increased production levels should help.
I agree with all that but I gather that the stumbling block for the phased arrays has been the phase shifters and the manufacturing cost of the elements. There has been a lot of materials research in recent years... The maths for a phased array is easy and does not have to be done very often.
It's going to be interesting to see who does well. Starlink and OneWeb are inelegant and inefficient systems, spending 60% of their time over mostly unpopulated water. But they could be cheap overall.
However, the very high throughput GEO sats can be launched in very low numbers and immediately offer a comprehensive service, which is a good way of offsetting their high unit cost. Slightly worse latency is an issue, but not for most users.
I'm not convinced that either approach will dominate, but it does look like consumers will be getting a whole lot of good choices. Which is good news unless you're a telecoms company in the USA with a local monopoly...
"It's going to be interesting to see who does well. Starlink and OneWeb are inelegant and inefficient systems, spending 60% of their time over mostly unpopulated water. But they could be cheap overall."
Part of it being cheap is that they don't care about that over-the-ocean bit (KISS). Besides, they have ways to monetize the ocean, too, by offering the service to transoceanic airliners and ships.
"The maths for a phased array is easy and does not have to be done very often."
For fixed links, yes. (this was something I had to do in comms systems)
When both ends are moving it's a completely different can of worms
"Starlink and OneWeb are inelegant and inefficient systems, spending 60% of their time over mostly unpopulated water."
two of the largest markets which spring to mind are ships at sea and aircraft in flight.
It used to cost ~$2/minute in electricity alone to run the transmitters for HF voice circuit calls back in the 80s. Times have changed and Iridium proved there was a huge pent-up demand (not where Motorola thoughit it would be though).
Being able to have real-time high bandiwth linking anywhere in the world is going to drastically change the way shipping operates (remote Watch supervision?) - and small isolated island countries aren't going to know what's hit them (Island telcos are frequently abusive monopolies setup to line pockets and the cat-and-mouse games my customers played with various pf them in the 1990s to be able to get email through instead of using their (100 times more expensive) fax services were eye opening)
As an admin at a small ISP who saw his bonded T1s fill up when users discovered alt.binaries, I did, but there weren't enough of us, I guess.
Same here, back in the mid 90's we went to a satellite feeding our nntp server. I assume that was considerably cheaper than hauling in another T1. I want to say the port and line costs for a T1 at the time were north of $2000/month.
I used to have Comcast (suburb of Boston, MA, USA). I was paying $70/mo for 100/6. Last year, they announced they were going to institute data caps on Massachusetts customers. I had always said that if that ever happened, I would dump them for FIOS.
Coincidentally, FIOS halved their monthly charge and doubled their speed (from 100/100 for $70 to 200/200 for $40). Since I already had the fiber installed, I signed up. The installer came out, hooked me up and I moved the RJ from my cable modem to the FIOS ONT. Installer told me to expect 300/300, which is what I have been getting since Feb of this year, when I called Comcast and dumped them.
I don't know what prompted the rate cut on FIOS's part, because up until this year they had always been neck and neck with Comcast. And data over a CATV plant, even if it's hybrid fiber/coax, is a losing game. Comcast is a losing company, especially now that FIOS is so competitive, and I think FIOS smells blood in the water...
Oh, and no more Sunday/Monday morning outages! Apparently, that's when Comcast makes their network changes, and recently, I have had to reset my modem once a month or so. Very annoying, but not happening with FIOS.
I hope it happens this time - for all the remote areas of the globe (think smaller island states) where it will never be cost effective to run fibre. Surprised the author didnt mention the Iridium LEO satellites that were first launched in the late 90s? A $3k phone and $10/min call costs from memory and they ran into issues with jurisdiction when certain governments threatened to shoot them down as 'unfriendly aircraft'. There still up there though. I am assuming Mr Musk has sorted out all the politics for his birds.
Iridium is more than still up there, they launched a new constellation recently. They are going to be pushed further into their niches, because their data offering will be swamped by Starlink and all the other upcoming systems.
But those niches are really quite important eg the Pacific Ocean tsunami warning system relies on one of Iridium's very unique services which no one else is replicating. There might come an awkward point where they need to continue being funded for the sake of those niches, and the US DoD gets to pay for it (a bit like they run GPS and let us use it for free).
Only about 25 of the original Iridium sats are still on orbit at the moment, the ones that failed and are uncontrolled, the other 65 have been deorbited now. Nice to see the Iridium Next constellation is up and running though, you're correct that Iridium have some very specific niche cases that just cant be served by other constellations.
I'll miss the Iridium flares though, used to love spotting those as a kid :)
I get 50Mbps from my local WISP. Decent router too (I could plug in my own). Starlink is faster (at the moment) costs nearly twice the price per month (and that's before the extra $15 electricity), is damn expensive to setup and his higher latency. Plus I'd have to trust prices won't increase after Starlink has it's IPO.
If they get to IPO... The company stands a good chance of being technically and commercially successful, but it's possible Musk won't want to be even more beholden to the shackles of the SEC. And, if they end up being dependent on DoD contracts that would be something that could be kept quiet if the company remains private.
"Plus I'd have to trust prices won't increase after Starlink has it's IPO."
Given the "sticker shock" of their pricing in the rest of the world, they'lre under pressure to being pricing down, not raise it
Bear in mind that the REAL money spinner for Starlink won't be ISP services anyway - that's just a distraction earning pin money.
LEO satellites equipped with lasers have transoceanic data latencies about 20% lower than submarine cables. The real money is in providing linking between stock market hubs - these are organisations which have paid BILLIONS for their own dedicated submarine cables that were a little shorter than the existing paths
https://www.reddit.com/r/SpaceXLounge/comments/9fvuqq/the_real_moneymaker_for_starlink_is_high/
> and LEO satellites are not really the kind of thing one can throw on the compost heap or take to the council recycling centre.
Surely the whole point is they burn up? I suppose technically that makes them landfill, but it's a bit of a stretch complaining about that when it's in the form of ash over several hundred square kilometers and the incineration is done in the upper atmosphere.
There's an essential big-money-making feature to Starlink that they're kind of keeping quiet about, which is high-frequency trading. There's a straight-line subsea cable from London to New York being installed at a cost of hundreds of millions of dollars to shave a few tens of milliseconds off the latency compared to the more wiggly existing cables, because this makes big bucks for traders.
The propagation delay to and between Starlink satellites is the speed of light so about 50% faster than optical fiber; even allowing for the other latencies, Starling will be able to offer considerably shorter latencies than even the straight-line cable, and they can do it between anywhere and anywhere else -- so, all the financial trading centres. The trading houses will pay *huge* fees for such links, because they can make stupendous amounts of money by being that little bit quicker.
There was a number floating round, and IIRC it was tens of billions of dollars...
This post has been deleted by its author
The Common Sense Skeptic did a good breakdown of Starlink on YouTube. Elon is talking about 42,000 sats with a 3% (so far) failure rate and a 5yr lifespan. They think they can get the cost of the satellites down to $250,000ea, but they could cost as much as $500,000. Launching 60 at a time puts the launch cost at a bit over $1,000,000ea ($50-$60mil for the rocket plus payload integration and overhead). After you finish with all of the arithmetic, the claim of it being economic is hard to believe. With all other things Elon, there has to be more government aspects. Maybe he can charge navies huge sums of money for at sea services when the satellites will hardly be in use. The same for big shipping companies that want to track their ships and crews and cargos.
Some of the numbers that are presented in the video might need changing but I don't see changes that would lead to Starlink making sense. In the US, there is a whole bunch of dark fiber that was put in by companies planning on being data toll bridges. That might have worked until engineers figured out how to cram even more data down each fiber and the investments didn't pay off. Just like the railroad, there are long stretches adjacent to major highways that could easily be connected up to extremely high speed internet that are nothing but undeveloped land right now. It's just going to take a few companies to take the plunge and build those new communities/company towns.
Did they take the traders into consideration? Sure, the "last hop" data centers are regulated near their respective stock exchanges but as business is global these days, faster long-haul communication between exchanges can mean big business for scoops and other key bits of "insider" information.
It potentially represents exciting times for anyone who is either i) living in a not spot or ii) nomadic and sometimes/often in areas without coverage. This could be like GSM. I was lucky enough to first be in Mumbai, and later in Nairobi when the networks were new and we could see that transition between mobile phones being for the few and then for the many. The between period where you can see how the technology can improve the life of the taxi driver or courier (and later relegate them to Uber and Deliveroo, if we let it happen), It really can give a lift to the people at the bottom.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
