Cyberpunk Fantasy Heartbreaker: Magic and Technology

[kzt]

and vice-versa, though satellites are harder to go after.

Not really. Lasers that can burn holes in armored vehicles at 5 km can burn holes in 0.5mm aluminum at 600 km. Communication sats are in very well known orbits, because you have to point your antennas at them. And you can easily see how a company might accidentally point a 50 megawatt laser cannon at the sat instead of a 50 miliwatt communications laser, right? It's the kind of accident that could happen to anyone.

[Endovior]

and vice-versa, though satellites are harder to go after.

Not really. Lasers that can burn holes in armored vehicles at 5 km can burn holes in 0.5mm aluminum at 600 km. Communication sats are in very well known orbits, because you have to point your antennas at them. And you can easily see how a company might accidentally point a 50 megawatt laser cannon at the sat instead of a 50 miliwatt communications laser, right? It's the kind of accident that could happen to anyone.

No, that's pretty overt. Doing overt things against other megacorps is a good way to make a cold war heat up. That's not something the megacorps will do for each other... that kind of war is not profitable.

[Username17]

No, that's pretty overt. Doing overt things against other megacorps is a good way to make a cold war heat up. That's not something the megacorps will do for each other... that kind of war is not profitable.

That kind of war is extremely profitable, depending on who you are. Every time a satellite goes down, someone is going to have to pay to put one back up. With a number of countries and syndicates being outright against satellites on general principle and against you having them in particular, the lifetime of a satellite is probably pretty short in the absence of binding international space treaties. Low Earth Orbit is probably a junkyard filled with debris beyond reckoning. And every time something goes up or something breaks or something has to be fixed or replaced, someone is making a profit.

The space debris around Earth already look lie this, and in 2075 it's going to be much much worse. From wikipedia:

Out of the estimated 600,000 objects above 1 centimetre (0.39 in) diameter, only 19,000 can be tracked as of today. This leads to wide uncertainties in the estimated quantities of debris, and the predicted path of their orbits.

There are literally millions of chunks of things flying around up there and satellites run into them all the time, creating more. And the result is that corps have plausible deniability when they arrange for satellites to have accidents, so they do that all the time, which drives the space debris problem up even more.

The only issue is that you still want space stations, and the players are going to want to be able to go to Anteros on Mars, Chandra on the Moon, and the Belter Refinery. Because those places are awesome. So no matter how Cowboy Bebop the lower orbits get, it still has to be a regular thing to leave the gravity well completely and to get to the Lagrangian space stations in the Kordylewski clouds. Since the Kordylewski clouds are ten times as far away as Geosynchronous Satellites are, it seems plausible that whatever the fuck it is that people do to dick with comsats is something that is impractical to do to the actual space stations.

-Username17

[Endovior]

No, that would be the blessings of destruction fallacy. Megacorps, having economists on staff, know the difference between a cold war where you occasionally sneak around and mess with your opponent every now and then through an arm's length third party... and a hot war, where it becomes obvious that the one group is acting against the other, and a cycle of retribution gets out of control and lots of both sides' expensive stuff gets wrecked... and their rivals overtake them completely as they grind each other into dust.

You can certainly eke out a little bit of profit when things get wrecked that you get paid to replace... but every time that happens, the net amount of total capital available to anyone goes down, and this being a quasi post-apocalyptic setting, there really aren't enough resources available as is to push things too far. It doesn't matter how sneaky you are about your scheming and plotting... even if nobody connects the dots and sends guys after you in revenge, every time you wreck something the other guy has, he's now got less money to spend on things, including things you have. If he needs the thing that got wrecked, he might wind up buying it back from you... but then again, he might not. There's a hard limit to this kind of thing where sabotage makes things stop being profitable, and whole industries dry up and blow away.

Accordingly, you might be able to get away with having some pissant little automated satellite that looks like space junk fall out of it's orbit a little and crash into another satellite, with predictable results... but you can't just charge up your lazors and blow the competition's shit out of space. If you do that, they will have to respond, and that's a negative sum game that nobody wants to play. Hell, you can't even arrange too many so-called accidents... there's only so many satellites that can fall out of the sky before people give up on using them and switch over to cable.

[Lokathor]

Advances in flight, energy production, and human ecology have been substantial and there are now cities under the seas, on the Moon, and even Mars. And yet for all of these demonstrable advances, progress is nothing like uniform. Several major calamities have occurred, and many things we take for granted today are practically fantasy in the world of 2075.

Sounds delightfully like cowboy bebop to me.

[Username17]

Hell, you can't even arrange too many so-called accidents... there's only so many satellites that can fall out of the sky before people give up on using them and switch over to cable.

EDIT: I have no idea why the last sentence did that, but I can't seem to get it to not do that.

You had an extra "[ / i ]" (no spaces).

Anyway, there is no "switching to cable". Switching to cable requires that you physically fight to control every centimeter of the Earth between you and wherever you want to go. It's just not going to happen.

But I basically agree with your lesser point that the skies wouldn't be red with laser fire and that satellites would have a life expectancy that was not measured in minutes. But between the fact that every corporation is hiding their lists of tracked space debris from each other and actively attempting to arrange accidents for each other's stuff, and fucking Helheim will straight shoot down your satellites if they violate their "space space" somehow, you gotta know that people will be measuring their satellite lifetimes in months or even weeks. Which is a noticeable but not unprecedented change from today's comsats, that have a lifespan of 5 years.

-Username17

[Red_Rob]

Given that this is an Awakened world the threats to underwater cabling can be many and varied. Deep One Isolationists can destroy cables and disappear before retaliation arrives, Devil's Reef and Y'ha'N'thlei create "black spots" of sovereign territory that block connections between various countries. Krakens are attracted by the cable laying machinery and make laying cables an expensive and dangerous prospect. Megaladon's super-acute electrical sense is driven haywire by the current running through the cables and causes them to bite and tear at the fibre-optics. The appearance and disappearance of various mystic isles and the periodic raising of sections of the sea bottom plays havoc with the established maps of the seascape.

And this is before you get into intercorporation sabotage attempts and corp wars to drive competitors out of business by destroying the prestigious high speed cable line they just sunk millions into. So having limited long distance connections seems pretty plausible to me.

[Stahlseele]

Megaladon's super-acute electrical sense is driven haywire by the current running through the cables and causes them to bite and tear at the fibre-optics.
Wat.
What is this, i don't evn . .

[Red_Rob]

Megaladon's super-acute electrical sense is driven haywire by the current running through the cables and causes them to bite and tear at the fibre-optics.
Wat.
What is this, i don't evn . .

Are you saying fibre optic cables don't hold an electrical current?

Modern optical fiber repeaters use a solid-state optical amplifier, usually an Erbium-doped fiber amplifier. Each repeater contains separate equipment for each fiber. These comprise signal reforming, error measurement and controls. A solid-state laser dispatches the signal into the next length of fiber. The solid-state laser excites a short length of doped fiber that itself acts as a laser amplifier. As the light passes through the fiber, it is amplified. This system also permits wavelength-division multiplexing, which dramatically increases the capacity of the fiber.

Repeaters are powered by a constant direct current passed down the conductor near the center of the cable, so all repeaters in a cable are in series.

Or are you disputing that sharks can sense electrical signals?

[Username17]

The Cloud of Junk
“Because we left garbage in all the other frontiers.”

The Earth is surrounded by trash. Space debris zip through the void at eleven thousand kilometers an hour, and the strongest shielding will keep out only the tiniest grains. Orbit around Earth has literally millions of fast moving objects around, and even the most adept agencies can only track the movements of a small fraction of them. Different corporations and governments do not even share their information with each other, and satellites get struck by debris all the time – becoming even more pieces of erratically moving debris. In our time a communications satellite stays up for an average of five years, but in 2075 they stay up for more like five months. So the inter-city backbone of The Network is changing all the time – old satellites break and new satellites go up every day.

The clouds of trash are not enough to keep space travel from being a thing, many space agencies manage to keep flight paths clear enough that they can get vessels outside the debris belts. And once you get out to fifty thousand kilometers or so, space is pretty empty and safe. The lunar colony of Chandra is the closest inhabited region of space, though the L4 and L5 space stations of Tiangong-2 and Mir-5 are not much farther. Radio communications reach these locations in over a second and they maintain their own infrastructure to hold their mirrors of The Network. Flights to these locations are expensive and take about 10 hours. The farther reaches of space are more remote, the Mars colony of Anteros is on an entirely different planet that is on a different orbit. Depending on when in the year it is, a radio transmission takes between a bit over 3 minutes and around 22 minutes, and even with the fusion-powered ships of the late 21st century, it still takes between 4 days and nearly 3 weeks to make the journey to Mars in person. Communication with Anteros is mostly done with the expensive and low bandwidth (but instantaneous) methods of magical sendings and quantum entanglement communication.

The Cloud of Noise
“I can't hear myself sneak in all this chatter!”

There are more wirelessly active devices active in any city at any one time than there are people. By more than ten to one in most regions. And while the actual range of most devices is small, it is nonetheless true that at any given time you will be within range of thousands of devices, all clamoring for attention and wanting their bit packages handed this way and that. And so it is that nearly every device that participates in the wireless world also participates in handing packets back and forth for whatever other devices happen to be in the area. Very complex informational space sharing algorithms are enforced that make cooperative communication possible. Data packets are handed from node to node through a dizzyingly large array of links and this works to keep the different nodes from jamming each other.

The cooperative communications protocols normally keep devices from really “knowing” in any real way what their packets contain, where they are going, or where they are coming from. And that means that devices will pass data on even when that data would not be allowed to stay on the device. Machines will willingly be a link for data that they would not accept as a transmission to them. Indeed, checking the password verifications for every data packet being routed through to another node would be impractical, and it just isn't done. Even secure installations will have password checks only at terminal nodes where data is actually being routed to – intermediary links won't bother. Secure installations thus work to segregate their version of The Network from outside pollution with wireless blocking walls and direct connections to dedicated satellites.

But of course there is no way to be certain that a wireless device actually passed on a data packet without keeping a copy for itself, nor is there any way to make sure that no intervening link kept a record of what route a data packet traveled. Unscrupulous hackers can get a lot of information simply by saving copies of data that pass through their wireless devices on a continuous basis. This is called “passive hacking”. The ad hoc wireless protocols are quite robust, but they are still predicated on the idea that every machine is playing nice and sharing the space. Unscrupulous hackers can flood the informational space with data packets that clog the tubes by being sent in long circuitous paths, and they can deliberately have their data packets collide with others, causing information to be lost. This is called “disruptive hacking”.

The Cloud of Meat
“Can I run some of these programs on your sister? She's like a little porcelain doll.”

The mind/machine interface was being worked on feverishly throughout the late 20th century, and substantial progress was made. In 2075, the mind/machine interface is very advanced. And that means that technological devices can interface with the brains of people with little required setup unless they are being actively jammed. In a lot of ways, this is incredibly convenient. You can have your bathtub turn on because you decide to get out of bed. Not by pressing buttons or flicking switches, but literally just the decision itself, because your house can be interfacing with your brain. You can also send an email to your friend without actually typing things just by having the computer read the words you want to send out of your brain.

The interfacing can and does also go the other way, where your perceptions can be externally stimulated by electronic devices. A device that stimulates simulated sensation to a person is called a “deus deceptor” or “DD”. When computers implant sensory information to a person's brain the sensations themselves are called “virtual reality” or “VR”. If these induced hallucinations are compatible with the actual world around the person they are “augmented reality” or “AR”. If the induced sensations supplant real-world sensations, that is “Zhuangzi reality” or “ZR”. Machines can also produce real light and sounds that can be detected by peoples sensory organs in the normal way. These are called “images”.

Unscrupulous hackers can interface with your brain against your will. Indeed, without technological intervention the very same children's toys that allow one to turn lights on and off with your desires could report to someone keeping an illicit log what your desires actually were. This is called “trepanation”. Unscrupulous use of DD technology can generate sensory data and even cause other state changes in the brain that can debilitate or insert foreign thought. This is called “basilisk hacking”. Images too can be produced in such a way that they trigger physical responses from epilepsy to bowel movements. This is called “flashing”.

[kzt]

No, that would be the blessings of destruction fallacy. Megacorps, having economists on staff, know the difference between a cold war where you occasionally sneak around and mess with your opponent every now and then through an arm's length third party... and a hot war, where it becomes obvious that the one group is acting against the other, and a cycle of retribution gets out of control and lots of both sides' expensive stuff gets wrecked... and their rivals overtake them completely as they grind each other into dust.

Chicago rules. "You cut my $10 million cable and something bad will happen to your 600 million sat. Capisci?"

[RiotGearEpsilon]

Even if intercontinental terabandwidth cables are nullified.

Frank, this is slightly bizarre of me to toss in there, but I have some relatives who were among the first people to ever put a transatlantic fiberoptic cable in the ocean. Would it be useful to get their feedback on the plausibility of fi-com cables becoming impractically failure-prone?

[virgil]

Any validity from this for keeping the internet down? http://www.cracked.com/article_18453_5- ... oment.html

[Vebyast]

Any validity from this for keeping the internet down? http://www.cracked.com/article_18453_5- ... oment.html

Bots (and possibly botnets) may be a central gameplay mechanic for some character archetypes. We're already discussing DNS security and wirecutting, and things basically proceed as in the article's worst-case scenario. Net Neutrality and Megacorps are implied by the setting to also be worst-case scenario. The Internet is stupendously resilient.

[kzt]

Are you saying fibre optic cables don't hold an electrical current?

Currently there are in use submarine fiber links across the Caribbean that have no repeaters. There is a company today that promises 500 km running 60 wavelenghts at 10GB each per pair without repeaters, while back in the early 80s people would have needed repeaters at less than 10 km apart to support gigabit speeds (however they didn't actually use gig speeds). So it's perfectly reasonable to expect that in 60 years you can drive a signal across the Atlantic or Pacific with 60 channels of faster than 100GB sec per fiber pair without a repeater.

You could, in theory, also have repeaters via fiber optics driving optical amplifiers.

[Vebyast]

You could, in theory, also have repeaters via fiber optics driving optical amplifiers.

According to an uncited claim on Wikipedia, most modern fiber already uses optical amplifiers, with the amps being powered by a secondary laser not used for communication.

[Endovior]

Any validity from this for keeping the internet down? http://www.cracked.com/article_18453_5- ... oment.html

Bots (and possibly botnets) may be a central gameplay mechanic for some character archetypes. We're already discussing DNS security and wirecutting, and things basically proceed as in the article's worst-case scenario. Net Neutrality and Megacorps are implied by the setting to also be worst-case scenario. The Internet is stupendously resilient.

This is very true. There is no unified Internet; so what portion of metahumanity's collected information resources you can access literally depends on where you live and who you're buying your bandwidth from. And of course, the parts that they'll be showing you are, in order of preference...

1: Things that your NSP wants to sell you
2: Things that other people have paid your NSP to show you
3: Things that people have paid your NSP to store on their servers
4: Things that your NSP has acquired for free from other people's servers
5: Things that your NSP had to pay to get off other people's servers
6: Things that people who haven't paid off your NSP want to sell you
7: Things that your NSP's competitors want to sell you

The cheapest pages to view fall under category #1 (in fact, pages under this category are probably free), and the most expensive are under category #7. If you buy your Internet from Axiom International, they will seriously charge you a massive extra premium fee to even look at the homepage of their competitor, Globodex Telecom Inc. They can get away with this because everyone's doing it, and because they're quite possibly the only people selling network access in your hometown. If, on the other hand, you happen to live in one of the areas where Axiom and Globodex are directly competing... then the amount that Axiom will charge you to even look at Globodex's homepage is similar to the amount you'd pay for a month's normal browsing. This is done because it's found to be remarkably effective at stopping the majority of users from viewing Globodex's homepage and seeing that they're charging 10% less then Axiom for comparable service.

Of course, most of the stuff that you actually want to see is probably under category #5; your NSP pays for this stuff for a reason, but they don't exactly hurt for it... they simply pass the fees on to you, probably inflated still further as they take a cut themselves.

[RiotGearEpsilon]

#7. If you buy your Internet from Axiom International, they who will seriously charge you a massive extra premium fee to even look at the homepage of their competitor, Globodex Telecom Inc. They can get away with this because everyone's doing it, and because they're quite possibly the only people selling network access in your hometown. If, on the other hand, you happen to live in one of the areas where Axiom and Globodex are directly competing... then the amount that Axiom will charge you to even look at Globodex's homepage is similar to the amount you'd pay for a month's normal browsing. This is done because it's found to be remarkably effective at stopping the majority of users from viewing Globodex's homepage and seeing that they're charging 10% less then Axiom for comparable service.

It seems likely that this could be pretty trivially bypassed by renting data service through a proxy or the internet equivalent of a disposaphone, but Globodex would probably do it anyway. Most folks aren't too tech savvy.

Actually, worst-case-scenario, they'd just edit the website you get sent from Axiom so that it would look like Axiom charges slightly more.

[Endovior]

That's the trick, though... if you're also buying data service off a proxy, then you're paying two different companies for service. Yes, this can get you more balanced coverage, but it's a tricky margin to calculate, since you're seriously trying to weigh the fluctuating rates Axiom gives you for the content they're pushing at you versus the fluctuating rates that Globodex gives you for theirs, and trying to decide if you can save money by paying the monthly fees for both and using whichever gives you the content you want cheaper at that instant. If you're especially clever, this might save you a bit of money... but since you're essentially just paying about twice as much on your monthly fees for the luxury of getting a second opinion on content costs, it's not a viable or attractive option to the average user.

[Username17]

Encryption
“We estimate that we can crack this faster by waiting a few years for computers to become faster and then starting the project on the new generation of machines.”

Cryptography is a complex thing. But an immutable fact of it is that if you are handed a set of data that has been scrambled by a non-repeating transformation of comparable size, that you cannot decipher it. Not “it's really hard to decipher” or “It'll take you a long time to decipher that” but that in fact you simply can't do it at all. So anyone with sufficient time on their hands and dedication to cryptographic secrecy can make a system that cannot be decrypted under any circumstances. It's called a one-time pad, and while resource intensive it is actually unbreakable. But people generally don't really need codes that can't be broken ever, most people will settle for codes that cannot be broken any time in the next hundred million years. That's the kind of time frame that even the extremely long lived are generally willing to concede that their secrets of today won't matter much once it has passed.

So while it is entirely within everyone's capacity to go out into the street, turn the microphone on super high and record random discordant noise for an hour, then download that hour into their drone as an exceedingly long cypher to get an hour of unbreakably encrypted communications between themselves and their drone – the vast majority of people are willing to accept a less intensive system where their communications are merely unlikely to be decrypted before the sun peters out. The tradeoff of ease of use to message security is dynamic, with many available options. For simplicity, the assumption is that when encryption is used, it is either SE, which is just good enough to keep anyone who doesn't have a quantum computer from reading your mail, and EUE, which is good enough to keep quantum computers from reading your mail, but requires you to share your complete two-way cypher with both sender and recipient before you can send any encrypted messages.

SE: Standard Encryption
“You can't get in here unless you push. With your hand.”

Every message sent by every wireless device is encrypted with what is called SE, or “standard encryption”. This is an asymmetric algorithm, which is based on very large prime numbers multiplied together. It has many advantages because the key the unravels it is different from the key that encodes it in the first place, and the key that encodes it is derivable from a single (very large) number. What this means is that every node and every link in the path between any two nodes has their magic number that they broadcast, and every packet sent to them (either as a final destination or as an intermediary link) is first encoded using that number. And if the computer is supposed to pass the packet on again, it decodes the standard encryption that was sent to it and then re-encodes it with standard encryption for whatever computer is supposed to receive it next. And it's layerable, so when you send a data packet to a node through a link, your computer puts on the standard encryption for the target node and layers the standard encryption for the intervening link on top of that. And because the link will strip their version of the standard encryption off the packet and recrypt it for the next link, the packet sender does not need to know the number of any of the intervening links besides the first (and the final node of course).

It's a great system and it prevents random computers from stealing your data. After all, any computer that does not have the decryption key can't take the encryption off, and thus wireless traffic can be sent all over the place without anyone stealing your data. And that is why people have been using standard encryption and systems essentially like standard encryption for nearly a century. And it has been the ubiquitous standard for wireless communication since the 2030s. Even old devices from your parent's house are going to be using SE (although they may be using smaller numbers).

The problem, from a security standpoint, is that factorization of large numbers is not actually particularly difficult if you have a quantum computer. Anyone with a quantum computer can derive the decryption algorithm for any packet encoded in standard encryption in seconds. And this means that any information encoded only in standard encryption and sent over wifi is vulnerable to passive hacking by anyone in range who has a quantum computer. Quantum computers are currently fairly rare and expensive, the exclusive province of assets and agents working for heavy hitting governments and syndicates. Which is great news for the hackers themselves, because right now hacking equipment is still rare enough that people haven't abandoned SE (not that there are any great alternatives according to currently understood math).

EUE: Effectively Unbreakable Encryption
“We estimate that we can crack this faster by waiting a few years for computers to become faster and then starting the project on the new generation of machines.”

Most highly secure communications use Essentially Unbreakable Encryption (EUE), a system where the sender and the intended recipient both have a cypher that is overlain on the messages and subsequently removed. The keys used in the 2070s are of variable length, but generally are thousands of bits long, and cannot be expected to be broken by any sort of mathematical attack. In order to attain such levels of security the cypher itself must have been shared at some earlier point between the intended sender and receiver, and it can of course be stolen either during the hand off or at any time that anyone has direct access to any of the computers which store the cypher itself. After all, EUE doesn't make the message completely illegible to anyone but the intended recipient; it makes the message completely unintelligible to anyone who doesn't have the key – not the same thing once espionage comes into the equation.

The big limitation of EUE is that every single computer that sends or receives the message in any final sense has to have a copy of the cypher and that anyone who gets a copy of the cypher can read any of the messages anyone sent with that cypher. EUE is a symmetric algorithm, and while it is successful at delaying even quantum decryption for years, it has a potentially fatal security flaw for every target that is intended to actually read any of the messages. Getting EUE started can be something of a chicken-egg problem, since if you had a secure enough way to send a message that you could get the EUE cypher to the intended recipient so that they could read your message, you might be able to send your message the same way.

[jadagul]

Since this is related to what I do, I'll chime in here a bit. (Yes, go ahead and ignore this for game balance purposes if necessary, but figured I'd throw it out there anyway).

From an actual theoretical perspective, the key question for future cryptography is whether we can solve NP-complete problems. If we can, I'm pretty sure there really isn't any level between your SE (can be cracked by anyone who can solve NP-complete problems) and an actual one-time pad. (If there's an encryption and you can tell whether you've gotten an answer--which is the only way you can decrypt with any amount of computing power--then I'm pretty sure you should be able to encode the whole question as some NP problem and pull the answer out of a hat). So either you want no algorithm for solving NP-complete problems, or you have to bite the bullet and just make shit up.

If we can't solve NP-complete problems but we do have working quantum computers (entirely likely. Shor's algorithm will factor primes and compute discrete logarithms, so it breaks basically all encryption people use today. But we're pretty sure those problems aren't NP-hard), then currently used encryption schema based on large primes don't work. But there are a couple public-key encryption mechanisms that people think are NP-hard: see Wikipedia on McEliece or Lattice-based cryptography. You can, of course, just declare that all these problems turn out not to be NP-hard and thus to be tractable with a quantum computer.

Other things to consider: we can probably set up a public key cryptography for anyone who has a quantum computer--that is, you generate the public/private key pair with a quantum computer, but you can do the encryption and decryption classically. This would make it possible to securely send messages to anyone with a quantum computer. My cursory reading doesn't indicate that we know this can happen, but it's likely; you probably want to rule it out for game balance reasons.

Finally, you probably want a shout-out to why quantum key exchange doesn't work (easy answer: the way you want to do it requires an unbroken optical connection, and we've already established that those have all been cut. You can obviously exchange keys over your instantaneous quantum communicators; you can probably just claim that's so expensive it usually isn't worth it, though).

Finally, you can give up on EUE, and say that the encryption is actually unbreakable without access to somebody's key. If you did that you can allow NP-complete problems to be solvable without ruining anything.

[Username17]

Finally, you can give up on EUE, and say that the encryption is actually unbreakable without access to somebody's key.

How is that different from what I have now?

SE can send "encrypted" information to people you have never talked to before, and it can only be read by the intended recipient or someone who has a quantum computer.

EUE is safe from quantum computer attacks, but you have to exchange the key with the intended recipient before they can receive the message.

You seem to be saying that I can give up on EUE by making it exactly like EUE. What the heck?

The only thing I am positing is that "one way" algorithms turn out to not be hard in quantum computing land. So you can have encryption that people can't break in a reasonable amount of time, but you have to share identical keys to encode and decode with both the sender and receiver.

-Username17

[jadagul]

Sorry, I meant that you should replace EUE with something that isn't "effectively unbreakable, but that is actually literally unbreakable. Because if the encryption is theoretically breakable and you can solve NP-hard problems then it's also breakable in practice. But if you say that no amount of computer time would let you break the encryption, I don't think that changes anything in the game but it allows nitpickers to not worry about a lot of the math under the hood.

I guess the summary is: if you can solve NP-hard problems then EUE doesn't exist--an encryption is either breakable or not. If you can't solve NP-hard problems then public key cryptography almost certainly exists. For the game balance results you want, you can assert that there are no public key systems that are NP-hard. Or you can just throw away the "effectively" from EUE, which is probably simpler.

[Username17]

I don't see how it is possible for something to be actually unbreakable. Even if we're in "you have to try all the 400 digit numbers, one at a time" territory such that breaking it would never ever happen, it's still possible. Actual one time pads cannot be broken. But they cannot be broken because they have a key length longer than the message and are only used once. If they were used over and over again, they would be breakable through comparative analysis.

A key that was generated by a system could be computationally intensive or even computationally intractable to crack, but it is still in theory going to be crackable. I really don't see how claiming that some system is literally unbreakable is going to push peoples' willing suspension of disbelief less hard than saying something is effectively unbreakable. It certainly would be a much bigger strain for my own willing suspension of disbelief, and I doubt I'm alone there.

-Username17

[jadagul]

It's unbreakable if you have no way of knowing whether you've gotten the actual message back out. Like, with a one-time pad we have no information about the message other than maximum length, because it could be literally anything that length or shorter. With a secret key we can start weeding out messages, but as long as comparative analysis doesn't throw out too many possibilities then your decryption algorithm can find multiple possible plaintexts and you don't know which one is the actual plaintext. If you keep using the same key too often or too transparently then comparative analysis becomes strong enough that you can break everything given enough time, and we're back to EUE.

Which is fine, but you have to make two assertions: 1) quantum computers can't solve NP-complete problems (which is likely), and 2) no public-key system is NP-complete (which is unlikely). Because if you reject the former then EUE literally can't exist (see footnote), and if you reject the latter then SE is effectively unbreakable and things get boring.

Footnote for anyone who cares why I say EUE can't survive if you can solve NP-complete problems:

Basically, an NP problem is one where you can check the answer if it's given to you. So, for instance, if I hand you a godawful algebra equation in twelve variables there might be no sensible way for you to solve it. But if I also hand you a solution, you can just plug it in and see if the solution works. Saying we can solve NP-complete problems basically means that any problem that we can check the answer to easily, we can also solve easily.

Now, if we can break some sort of encryption--under any theoretical circumstances--that means we have some way of knowing when we've got the right answer. If we didn't, we'd have no way of knowing if we'd cracked the encryption. So if we can solve NP-complete problems easily then we can crack the cipher easily.