Help - Search - Members - Calendar
Full Version: How much random info can a facility generate?
Dumpshock Forums > Discussion > Shadowrun
WeaverMount
Ok so in 2070 there are two instances i can think of off the top of my head where a facility could greatly benifit from the security of a one-time pad cypher where key distribution is relatively trivial. I was thinking of drones and wage-slaves. Both are tracked and in wire contact meaning that encryption would be good. Both also frequently 'dock' at a physical location where more 'pads' could be uploaded with the maximum security that site can.

So my question is, how much truely (or close enough for a one-time-pad) random data could you generate? Could a black box with a radio active isotope and REALLY good sensors generate appreciable amounts of random data? Could it at least generate enough truly random seeds to generate large pads? Is there there a better method altogether?
Tarantula
Call it strong encryption and find the rules in unwired.
kzt
If you want to assume they have invulnerable encryption it's easy to justify that. Random seeds are perfectly possible with hardware using radioactive decay.
Lebo77
QUOTE (kzt @ Sep 8 2008, 04:10 PM) *
If you want to assume they have invulnerable encryption it's easy to justify that. Random seeds are perfectly possible with hardware using radioactive decay.


Under the RAW the only "unbreakable" encryption is that employed by resonance beings (sprites and technomancers) between themselves. It's crackable by other sprites and technos, but normal computers see it as noise.

While I understand that one-time pads are "unbreakable" in real life (assuming you have a truly random noise source and your key distribution in not compromised, and you don't make the mistake of re-using a pad), in SR a "new math breakthrough" has rendered ALL encryption weak. If you go the One time pad route then many types of encryption will suddenly become un-crackable. (i.e. the PCs will want it too).

If you want to give the NPCs "plot armor" against hacking just say "It's unlike anything you have seen... it seems uncrackable" and leave it at that. Trying to apply real-world Info Tech to Shadowrun... that way lies only madness.

P.S. This is based on my experiences. I GM sometimes for a group of Mathematicians I.T. Pros. Better to just make things "magic" then try to explan them.
Tarantula
QUOTE (Lebo77 @ Sep 8 2008, 02:23 PM) *
"It's unlike anything you have seen... it seems uncrackable"


At which point I go "I let my decrypt program at it".
CanRay
QUOTE (Tarantula @ Sep 8 2008, 03:28 PM) *
At which point I go "I let my decrypt program at it".

It comes back with the complete works of Shakespear in the Original Klingon.
Shiloh
QUOTE (WeaverMount @ Sep 8 2008, 08:14 PM) *
Could a black box with a radio active isotope and REALLY good sensors generate appreciable amounts of random data? Could it at least generate enough truly random seeds to generate large pads? Is there there a better method altogether?


You can get true randomness out of RF white noise, too, I believe. Saves that messing about with radioisotopes: background hiss doesn't have a half-life smile.gif Another method I think uses thermal noise in sensitive microcircuitry.
Wasabi
"Dramatic Encryption" is an optional rule in unwired allowing unbreakable encryption.
"Strong Encryption" makes the interval so long its often pointless to stop and decrypt it during a 'run.

Of the two Dramatic Encryption seems the best choice for something absurdly difficult.
(And I play a TM most times so yes, I believe there are things a TM shouldnt be capable of decrypting)
Blade
Please stop with this "OTP are the ultimate encryption method!".
OTP are good. They also have vulnerabilities even if these aren't that problematic. But the biggest issue is that they aren't very practical to use. If there were, there would be used everywhere already!

First you have to store the keys somewhere, and even if you manage to send the key without any risk of interception, the storage itself can probably be compromised one way or another.
There's also the problem of the non-reuse: if you don't know which devices you'll connect to and how many connections you'll need to do, how do you know which pad to use for your next connection, and how do you know it hasn't been used yet?
Even when you do know which device you'll connect to and how many connections you'll make, it's very easy for an attacker to invalidate all your OTP in a few seconds, leaving you with no pad left.
Crusher Bob
The problem is not that OTP don't have their problems, but that all other encryption in SR is complete and utter crap. If you had good non-OTP encryption in SR, then people would use that instead, but there isn't so the OTP keeps comming up, since the OTP are not breakable. They have implementation difficulties, but when they are the only effective method of encryption available to you; you just grit your teeth and put up with the problems (and everyone becomes very good at implementing OTP systems).

[Edit]
It's kinda like playing in a Highlander game. The only way to kill another immortal is to cut of his head. Sure cutting off his head is kinda hard to but, it's the only way to get the job done. And after hundreds of years, everyone is pretty good at cutting someones head off.
[/edit]
Blade
Yes but the point is that it's not just "implementation difficulties": they are just impossible to implement in most common cases!
To me, they would just be a possible explanation for "Dramatic Encryption", which are rare and only come up in special situations.
Tarantula
QUOTE (CanRay @ Sep 8 2008, 05:33 PM) *
It comes back with the complete works of Shakespear in the Original Klingon.


BRILLIANT! This'll sell for a big wad of cash!
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Dumpshock Forums © 2001-2012