Hacking in today's world with a TL 10 Computer...

[Zeta Blaze]

Originally Posted by Snaps

I guess the real question for me is: Could a Complexity 9, High Capacity, Quantum Computer break today's best encryption in a reasonable amount of time?

the best probably not especially if we only use current math and not some possible future math like a proof P=NP and how to convert between them. Arguably the best algorithm is called a one time pad where the sender and receiver b have a block of shared secret random data that is at least the exact size of the message then A XOr s the message with the pad and sends it to be and then b XOrs it again to get the message. Then both throw out the key so people cant analyze patterns in the key from reuse. You can see why this is hard to crack in any conceivable situation but it is also very rarely used usually only for short messages to things like nuclear subs and spies. Most anti quantum algorithms have some major flaw like this one.

But and this is a big but most Encryption is doomed. It will crack anything using public key or RSA in a few seconds after download tops.
as for passwords and WIFI access depends heavily on how it is encrypted most passwords can be snooped on wifi or a tapped wire as they are not using one time pad or some other Quantum proof algorithm so anything you overhear that isn't specifically designed to block this is open (you will only see something like that in secure military or intelligence RF/wired and some high security corporate wired).

So yes assume any password you overhear is broken and any encryption meant for one person to send and another receive with no prior communication is broken but truly paranoid people with prep time will need to have there access code stolen or a lot of time used with TL11 decryption programs.

Note also that if you have the right skills and or software you could use this to gain control of most of the internet and by proxy overhear most of what you need to get into any network that is not physically separate from the internet but this will take time to set up and may eventually be discovered. Once you have it set up though you will be able to hack practically anything on the internet. But remember that is not everything especially government and military black projects are not usually on the net unless the project is trying to control or monitor it.

[Fred Brackin]

Originally Posted by Snaps

I guess the real question for me is: Could a Complexity 9, High Capacity, Quantum Computer break today's best encryption in a reasonable amount of time?

As far as Gurps canon is concerned the question is easily answered by looking at p.47 of UT. TL9 Basic encryption requires a C8 computer and Secure encryption needs C10. You have C9 + an effective 5 from Quantum for this purpose. Exceeding the required Complexity by enables decryption with no significant delay.

So you can read TL9 Secure encryption in real time and TL8 encryption will of course be easier than that.

You still can't do anything with one time pads.

You probably can't get by strong passwords by raw power but any time a password is sent by WiFi or other interceptable medium you can grab it then (even if it uses normal encryption).

[Anders]

Do they even have compatible code? Could you hack something from the 70's with a modern computer, or would the computer just say "I'm sorry, I don't speak COBOL*"?

*Or some language actually appropriate for the era, if COBOL is not

[malloyd]

Originally Posted by Anders

Do they even have compatible code? Could you hack something from the 70's with a modern computer, or would the computer just say "I'm sorry, I don't speak COBOL*"?

*Or some language actually appropriate for the era, if COBOL is not

You will need compatible codes, but TL10 software designed to hack low TL stuff can probably figure it out itself given some sampling time. The communications protocol rather than the high level computer language will be the part that matters though.

And this is really the biggest hurdle. Few systems from the 70s are set up to handle I/O over external com channels. If you want to hack one you will need physical access, at least to the cables that connect the terminals to the computer building if not to the machine itself.

Hacking really should be thought of as allowing you to compromise communications, not computers. That's what makes it feasible - a system designed to communicate with others has to be relatively vulnerable, because other machines, designed by other people, must be able to figure out what that pattern of voltages is supposed to mean in order to communicate. There must be some more or less known or accessible way to translate between whatever your machine does internally and an external standard. Once you figure out how to break *that standard* you're in to any machine that uses it. This is why security holes are typically specific to an operating system or browser (i.e. one of those external standards), and not say specific to the hardware manufacturer.

[Vynticator]

Honestly, anything a few years ahead within the same TL can probably run rings round any thing a few years earlier. By TL11, if you allow it to interface with the older tech then everything is fecked, frankly.

[Zeta Blaze]

Originally Posted by malloyd

Hacking really should be thought of as allowing you to compromise communications, not computers.

That is exactly right when you brake the system what you are doing 9 times out of 10 is faking legitimate communications and pretending to be someone who can legitimately do something remotely that you dont actually have permission to. The other 1 out of 10 you are doing something no one has permission to do but that due to a bug the system still allows to happen.

If you want to gain total control of another machine you use one of the above methods to gain access and then install a brand new program on the target machine this virus then lets you operate the victim as you see fit and can be so subtle that the victim never even knows they are infected.

On the topic of connecting to old systems I have worked with Legacy COBOL systems for a past employer of mine as long as you aren't looking at the internals the COBOL program has an interface comparable to any other program that is where <99.9% of software based hacks attack a system anyway. The only advantage to knowing the victims source code is it helps you to find bugs in there implementation to allow you to do things the system was never designed to do. Normally making someones system do for you what it is supposed to do for an authorized user is more than enough.

malloyd is also right in that any password sent over a network you can listen to can be assumed cracked. But anything encrypted with a one time pad or whose password you haven't snooped (even if encrypted) is probably safe.

Note if you have access to TL11 math as well as computation I would not be surprised if you could statistically analyze something larger than say 100kb sent under one time Pad find an unknown and unintended pattern in the pad and then exploit that to recreate the pad and read the message but i would bet that would take a few hours.

Good passwords that you haven't sniffed are still a beast to break.

[Zeta Blaze]

Originally Posted by Vynticator

Honestly, anything a few years ahead within the same TL can probably run rings round any thing a few years earlier. By TL11, if you allow it to interface with the older tech then everything is fecked, frankly.

RSA was developed in 1977 and is still the gold standard in public key encryption. If it weren't for the immanent development of quantum computers it would probably remain the dominant encryption for at least a decade. Though a proof of P=NP (which is probably false) would ruin it and most other encryption.
And as it stands non quantum computers have almost no ability to encrypt something that a quantum computer cant break except possibly one time pad some other algorithms can make it difficult but unlike conventional vs conventional it is just a matter of time not a practical impossibility.

[Sindri]

Originally Posted by Zeta Blaze

RSA was developed in 1977 and is still the gold standard in public key encryption. If it weren't for the immanent development of quantum computers it would probably remain the dominant encryption for at leas a decade. Though a proof of P=NP (which is probably false) would ruin it and most other encryption.
And as it stands non quantum computers have almost no ability to encrypt something that a quantum computer cant break except possibly one time pad some other algorithms can make it difficult but unlike conventional vs conventional it is just a matter of time not a practical impossibility.

There's no "possibly" involved with except one time pads.

[Flyndaran]

Originally Posted by Vynticator

Honestly, anything a few years ahead within the same TL can probably run rings round any thing a few years earlier. By TL11, if you allow it to interface with the older tech then everything is fecked, frankly.

Only if you have TL 11 math. Most codes, from what little I understand, are based on presently effectively unsolvable problems. But solve those problems, and the codes may become like cute substitution cyphers.

[Zeta Blaze]

Originally Posted by Sindri

There's no "possibly" involved with except one time pads.

if you have TL11 math and computation you may be able to find a flaw in the one time pad such that it actually contains a pastern then exploiting said flaw in the pad you could in theory brake it. That is why i say possibly. If the algorithm is executed flawlessly there is no way, but if the algorithm is at all flawed in execution it may be possible to brake. Generating a sufficiently large pad that is truly random is not a simple feat and gets harder as the pad grows. Furthermore the larger the a flawed pad the easier it becomes to find and decipher a pattern to exploit.