Reminds me of the Tom Scott video on the cable singe that leverages the moiré effect:

https://www.youtube.com/watch?v=d99_h30swtM

I worked for an L1 blockchain for a couple years and while the term "decentralized" is bandied about a bit too much, reading articles like this does underscore how vulnerable global traffic is.

e.g. some EVM chains have a majority of "stake" in either one country or several adjacent countries (particularly in Asia). A cable cut there, which the article points out has happened, can lead to the blockchain getting cut off or even halting.

I think people conflate decentralization with redundancy or resilience. IMO it just means you have multiple points of failure instead of one.

10 years ago the talk of hackers partitioning the internet to double spend BTC all seemed a bit far fetched. But now crypto has trillions in market cap and governments involved -- governments that sometimes cut network cables just to send a message. The "means" and "motive" exploded by 100x, did anyone find a way to close the "opportunity" or did people just stop talking about it lol?

You'd also need to take down fleets of satellites in space. Double-spends aren't really easy to profit from either.

No, you just have to give Elon a call and ask him politely to disable Starlink on your side of the partition. If he refuses, mention the size of your antisatellite arsenal. He won't, though, because he has already had this conversation and made this calculation and chosen to capitulate several times before.

It might be difficult for you or I to profit from double spend, but remember, the threat profile is entire countries, and if any of the heavy-hitters wanted to kill BTC the logistics of conducting a big gnarly double-spend would barely register on their existing cyber budgets.

Back in the day, these scenarios were fanciful because crypto was small-league and the viable threats were big-league. Since then, crypto made it to the big leagues, but I don't think it ever became capable of handling big-league threats. At some point it will have to reckon with that.

There are multiple satellite ISP companies and there are even Blockstream satellites that are fully dedicated to broadcasting Bitcoin blocks.

I'm not sure you really understand what a double spend is. It's not actually a big deal in the real world. An attacker capable of double spending is simply capable of reversing blocks, therefore reversing transactions.

So, for example, an attacker could buy a house with BTC, and then when the papers are signed, reverse the transaction to get their BTC back. But in the real world, the victim would just go to the police and get their house back.

The only plausible "big league" threat would be a government investing billions in ASICs to perform a 51% attack and grind the network to a halt. But the Bitcoin community would probably quickly move to a new Proof of Work algorithm and render their ASICs worthless.

I think the issue is that centralization is more economical due to scale, but that assumes all else (geopolitics or other black swan events) don't impact things.

We saw a similar thing with supply chains and chip suppliers.

> reading articles like this does underscore how vulnerable global traffic is.

Then also remember that the internet haven't seen a global outage since forever, worst parts have been when some submarine cables break and traffic has to route around it, but never has the entire thing fully went down. Considering how most of it is built with bandaids, silver-tape and WD40, it's surprising it is as resilient as it is.

Not sure about other blockchains but with Bitcoin, you only need one node to connect via satellite and the chain is good. In fact, there are satellites that broadcast blocks from space (no ISP required). [0]

[0] https://blockstream.com/satellite/

Also BTC has a very conservative 10 minute block time that protects to some extent against a network partition (there is time to recover)

During an internship I worked on optical transport network equipment. Most customers were either submarine or long-haul across land. I was told that many clients would design redundantly, mainly due to people digging up the cables, chopping them in half with a shovel, and being disappointed when there was not much copper to steal.

aren't decentralized systems using the same infrastructure and are also relying on ip to route around missing routes? Just like centralized ones?

I work on autonomous submarines that image the seafloor, and let me tell you it is very hard to engineer deep ocean systems. Just operationally, testing things in rolling waves is a brutal job. Arguably at parity with space as an engineering design environment.

I forget who said it, but there's a quote about how _of course_ we know more about space than the bottom of the ocean - in space we only have to deal with one atmosphere.

>only have to deal with one atmosphere.

A difference of one atmosphere?

The one atmosphere that telescopes are looking through. It is nowhere near as problematic as looking through seawater.

But space telescopes exist where you don't have to deal with one atmosphere.

This is about sub sea pressure where it reaches a thousand atmospheres. So a space telescope in zero atmosphere doesn't help.

There’s only one atmosphere of pressure from inside to outside. Under the sea the direction is reversed and pressure is much higher.

Spacecraft are pressurized to much less than one atmosphere, because it reduces the stress on the craft, and the weight of the air supply.

I was curious and went to look up how low it was, and the references I can find online say that the ISS, the Space Shuttle, and the Crew Dragon are all at 1 atm.

It does say that the pressure in the Space Shuttle was decreased to about 70% atm for 24h before EVA.

Apollo was 1/3 atm 100% oxygen.

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

https://space.stackexchange.com/questions/5690/why-is-the-br...

https://www.google.com/search?q=pressure+inside+the+space+sh...

https://www.google.com/search?q=pressure+inside+the+apollo+c...

This is actually a terrible idea, because to keep the air breathable you'd have to raise the oxygen fraction, making fire risk much worse. (Inert gases in air suppress fires by carrying away heat, making it harder to sustain a flame.)

People live on mountains just fine.

This exact same problem exists on mountains, which is why there are no permanent settlements above 5,100m. The Apollo pressure mentioned above (1/3rd atm) is roughly equivalent to an elevation of 8,500m, well into the "death zone" for a normal atmospheric mix.

Yes, I know about the death zone. But air pressure is significantly less in Denver, and many town are much higher.

Likewise commercial aircraft.

Fry: "How many atmospheres can the ship withstand?" Professor: "Well, it's a spaceship, so I'd say anywhere between zero and one." s02e12

futurama comes to mind. the spacecraft can withsthand any level of atmosphere between 0 and 1

I (briefly) worked as a test engineer on a similar kind of contract. To meet the UX requirements for the system we were building, it had to be reliably operable in sea state 7+. The customer wanted touch screen interfaces. I exited that project when they were looking for volunteers to go spend time offshore in the Gulf of Alaska in bad weather, because that was the test environment they could find that met the requirements. Insanity.

Shoulda told them to test it on roller coaster with a splash zone. Way easier to get engineers to come to work at Six Flags than brave the briny deep.

damn, I would definitely have volunteered. But quite a crazy environment, seems to me mechanical switches with protective caps to prevent accidental activation on would be better suited to such things. Though it also depends on the possible consequences of the wrong input I suppose, if it were like a microwave or a clothes washer I guess its fine

I found that sea, space and automotive are the three most hostile environments, each with very different requirements that made it hard in different ways.

Seems reasonable. I'd like to hear your thinking on the automotive side. Is it the cost pressure at scale and durability requirements?

I have been trying, and failing, to buy a cross section of a submarine cable. I'd love one as an art piece for my house.

Everyone in the industry I can find including people who do recycling of submarine cables. It seems that cross sections are sometimes given away as gifts as part of a project but are really hard to find.

who have you asked?

I think it's like slices of tree trucks to build wall clocks (or whatever) kind of wall art out of: If you know a guy who knows a guy who works with trees, then it's easy to get a rough cut-off slice.

It may even be free. If they're already there cutting the thing with an appropriate saw, then slicing off one more bit and saving it for Bill's friend jgrahamc isn't a big deal.

But without that kind of relationship, it's hard.

(I want one, too.)

Underwater cables have been one of the more awe inspiring things for me. I know nothing about them but they seem SO crazy and yet we have them running all across the oceans. .·°՞( ̑ ᗜ ̑ )՞°·.

Ah, the “unusually unlucky” cables serving Taiwan and in the Baltics.

Like statistically impossibly unlucky.

At least for the Baltics, there is the unique aspect of the infamous "shadow fleet". Utterly derelict ships with barely qualified crews.

Ordered to swim back and forth with anchor down until they catch something.

Yes, I admire the restraint required to not put "unlucky" in quotes.

How practical would be a mesh of satellites shooting lasers at each other, compared to undersea fiber-optic cables?

Alternatively, is there a wavelength of light to which water is completely transparent? Can we shoot lasers of that frequency through open water instead of through cables?

Maybe something that could be mounted on blue whales?

Starshield (military arm of SpaceX/Starlink) does this for satellite imaging.

Normally in imaging, you first need to get the satellite over the location you want to image, at the right time - then let it orbit over a downlink location.

There are companies making a lot of money by having downlink capability in useful places (eg: near the poles) - combined with a fat pipe to the Internet.

With Starshield, the satellites talk to each other over laser - so the satellite doesn't have to wait to travel to a downlink location - just relay to of that is already there.

So for certain special cases it can help - but I think you'll find a cable stuffed with multimodal fiber had vastly more bandwidth.

Someone else with more experience could chime in but I would put it at less practical due to latency (sats are very high up), bandwidth (one laser vs a bundle of 1000s of fiber optic cables), cost (to overcome latency you need a closer orbit which requires more sats), single point of failure, lifespan, intercepted communications, solar flares… and that’s just off the top of my head.

I don't actually think latency would be that much worse.

Of course, it depends on the altitude of the satellite but speed of light in a fiber optic is only ~2/3rds of speed of light in a vacuum so you can afford some detour

> speed of light in a fiber optic is only ~2/3rds of speed of light in a vacuum

That made me wonder: Why can't we use vacuum cables?

(the sort of questions I should probably be bothering an LLM with.. someone should make a LMGPTTFY)

If you figure out how to make cheap vacuum cables you will probably become very rich.

But otherwise chains of microwave links does get much closer to c in applications where you want absolute lowest latency (or much more commonly, when digging a fiber line is too expensive)

> Why can't we use vacuum cables?

Hollow-core fiber is basically this!

> LMGPTTFY

"let me go prompt that for you" or "let me GPT that for you"?

> "let me go prompt that for you" or "let me GPT that for you"?

yes

Not vacuum but an inert gas: https://www.thefoa.org/tech/ref/fiber/Hollow-core-fiber.html

You probably can’t get under 500 km orbit, so 2000 km round trip is the shortest you can do between you and a ground based server. That’s 150 ms minimum round trip latency.

Light travels 2000 km in 7 ms. Maybe you took the inverse and got 150 Hz?

latency lower for satellites, trading firms have looked at this already, no idea if anyone has actually done it

bandwidth would indeed be worse

starlink sats are some of the lowest orbiting satellites in space - 330-370km for v3. this helps latency as well as easy deorbit at end of life. the v3 sats have multiple 400 gbps lasers for meshing... I can't decide if your comment is sarcasm or not actually because starlink is fully functional and doesn't have any of those issues

*Gbps

(Irritiating, but not as bad as those who insist on having milli-bit speeds I grant you).

Nice catch, gJim

A submarine cable is a bundle of hundreds of strands, so you need quite a few satellites to replace one cable.

Typical submarine cables at at absolute most 24 strands, more commonly 4, 6, 10 or 12 for modern ones. Terrestrial cables can easily be hundreds and that's quite mundane.

The high capacity in submarine fiber is achieved through DWDM and coherent modulation 100Gbps+ transceivers, not massive parallelism in fiber use.

Why? compared to the laying cost, an extra 10000 fibers should be pocket change

The cable contains not only the fibers, but also optical amplifiers to compensate for the losses. Even for 24 fibers, the amplifier is a substantial and expensive piece of equipment. It also requires electric power, which is not trivial to provide along the entire span between continents.

And in short distance unrepeatered coastal festoon fiber systems where it's within normal DWDM system reach, node to node with no active underwater electronics, those are also often in areas with relatively low population/traffic that won't need more than 8 or 12 fibers at absolute max on any leg in a chain or ring. Such as the Connected Coast project being built in British Columbia right now.

I'd be astonished if it runs into capacity problems with an ultimate upgrade capacity of being able to run at least 20 different 200 or 400 Gbps circuits per fiber pair if appropriately costly DWDM terminals are installed at each end of dark fiber. Not even if all the tiny towns and cities that it hits were to magically quadruple in population.

12 is ridiculously low without the repeater concern. It seriously should cost a rounding error more to put 400, maybe 4000 fibers in that cable because the cost of the fiber itself is literally nothing compared to the cable armour, boat hire and labour.

Depends on a lot of things, but most undersea cables have several fibers (typically around 10 fibers), and each fiber runs many wavelengths (typically 40 or 80). Whatever your bandwidth is per wavelength (there are lots of options, at different costs, up to maybe 400gbps) that's a huge amount of bandwidth to replace.

Space lasers probably don't do multiple wavelengths, and aiming a lot of laser transceivers is going to be challenging.

> Alternatively, is there a wavelength of light to which water is completely transparent? Can we shoot lasers of that frequency through open water instead of through cables?

Even if you could, that's only going to work for short hops, long distance fiber includes repeaters every so often (50-100 km) to deal with loss, it would be tricky to build those under the sea ... and you need to power them, which means a cable, might as well include fiber with your power cable.

Modern satellites, in particular Starlink, use optical transceivers that are identical to the ones used for the fiber optic links, with an addition of all the required parts to make them work across free space. So a link between a pair of satellites is very comparable to a single pair of optic fibers.

The cables can go the entire distance between the data centers. The satellites probably connect to their closest neighbors, and any long distance transmission would have to make multiple hops, and share the bandwidth with everything else in the system going through the same nodes.

Since there are hundreds of undersea cables, with a dozen fiber pairs or so each, the cable system has several orders of magnitude greater throughput compared to the current constellation of satellites.

If you care at all about this, you'll love Neal Stephenson's long article Mother Earth, Motherboard: https://www.wired.com/1996/12/ffglass/

I used to have a print copy of that issue. That single article was something like 50-60 pages long with all the (fabulous and informative) pictures.

I think about this often.

Scan from the print issue: https://archive.org/details/wired-magazine-04.12-1996-decemb...

My son's school computing teacher was convinced that these oceanic communication cables float on the surface - to the extent they had a bit argument about it in class. I'm not sure how the teacher thought ships navigated around them.

My history teacher in high school was wrong. He said he was right because the teachers' answer key said X. I replied that X was obviously wrong. He knew he was wrong, because he hid behind what his answer key said.

We didn't get along.

I found https://smc.peering.tw/ fascinating. It shows an up to date status of submarine links to Taiwan and you can see many of them are having outages. Even more interesting is the historical view showing that the system operates in an always partly degraded regime

I recently learnt about Transatlantic telegraph cable: https://en.wikipedia.org/wiki/Transatlantic_telegraph_cable

It was laid in 1858. During its brief three-week lifespan, only 732 messages (amounting to roughly 400 total telegrams) were successfully transmitted before the cable died.

Likely failure was due to not being insulated properly and Wildman Whitehouse:

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

He got into a technical spat with Lord Kelvin and it didn’t go well for him.

The Digital Antiquarian made fantastic "A Web Around the World" series of articles covering start of telegraph all the way to first Web Browser in a style of James Burke Connections.

A Web Around the World, Part 1: Signals Down a Wire January 7, 2022 https://www.filfre.net/2022/01/a-web-around-the-world-part-1...

https://www.filfre.net/2022/01/a-web-around-the-world-part-2...

https://www.filfre.net/2022/02/a-web-around-the-world-part-3...

First working cable was sort of Monty Python Swamp Castle deal.

This was a fascinating, well written article.

Ever seen a crab wiggle its way down into the sand?

Now that AI + Robotics is taking off, perhaps the industry will get something in the near future that can crawl along the cable and help bury it over time (perhaps mechanically or ultrasonically), down enough that it can't be dug up by the anchors of ships going back and forth in a grid-like pattern...

Considering that even the moon has been visited more often than the Marianna Trench, it's truly a difficult and scientifically demanding subject.