Unwired: Not Happy, Taking requests Options

[kzt]

Until someone in management realizes that 'programmed in' is so much cheaper than 'burned in'.
Dystopia, to the tech person, means that security models are based on cost, not on security.

As allowing programming results in people stealing service, which impacts the bottom line, burning in the code is based on a cost model.

[kzt]

Which means that a one time pad of sufficient size to encrypt it would be equally too large.

You don't need an OTP to get secure symmetric encryption. A OTP provides an ability to get theoretically unbreakable encryption, while a symmetric system can be decoded in only a few hundred million years. However the Verona project was decryption of OTPs, broken through attacks on the entire system used for communication and flaws in the way the agents and Moscow center used the OTPs.

[Cthulhudreams]

Oh.

Hrm.

I actually agree with you - maybe we do need OTP to be the only viable encryption method?

[kzt]

Not really, forcing people to exchange key codes to establish a secure session is a pretty big limitation. But it is possible to overcome this by having things like comlinks and credit chips having the keys burned in and linked to an essentially unlimited data base at the telecom company or the bank.

This means that nobody (other then the communications company - and you trust them, right?) could snare your messages off the air, but it also means that the only way you could securely communicate with someone over a comlink (secure from the communications company) is if you had met face to face and exchanged keys.

There are supposed to be certain approaches to public key encryption that are not effectively compromised by quantum computing, but I have no idea how true that is.

[Gelare]

Frank, here're some more random requests to consider: descriptions of exactly what AR and VR mean and do, in terms of bonuses to real-world or matrix actions or initiative passes, including how many initiative passes can be used in AR, and if the answer is "As many as you get," then the answer to the question, "Why would anyone use VR?"; the distinctions between cold-sim and hot-sim, how they are activated, whether you can switch between them, what kinds of damage you can sustain in each, the bonuses and IPs they give, and what becomes of your meat body when in VR, including perception tests; an explanation for why a lucky hacker can't use Jedi Trick once and have control over all of Zurich Orbital's systems; an explanation for how technomancers' constant Connection range keeps them from totally hacking the hell out of everything and everyone; exploration of the possibility (or explanation of the impossibility) of downloading information directly into your brain, skillwires-style, with your datajack; whether brain hacking includes Control Actions/Control Thoughts-like shenanigans.

Sorry for doing that sort of datadump-style, I just wanted to throw out a sort of checklist of various things. Two more suggestions/requests for explanation are why hacking an orphan device/brain from a distance is a better idea (from a game-design perspective, not from an in-game perspective, since you can invent fluff to support it any way you wish) than requiring handshake range (in the brain's case, contact range) because I always thought handshake seemed more reasonable; and second, why you don't want to go in Unwired's direction regarding Technomancers, and Dissonance followers being like toxic mages. I haven't yet read Unwired, but I'm just curious why you think that's a bad thing. Thanks for taking the time to write this stuff.

[WeaverMount]

Hello Frank, what I would really like to see are some tables (for various opposed tests, prices, capacities etc) also some parenthetical section (or page) referances with the text would make using and learning your rules much easier. While I'm wishing I'd love a table of content, illustrations, and an index (in that order if you are thinking about it at all)

[Aaron]

Hello Frank, what I would really like to see are some tables (for various opposed tests, prices, capacities etc) also some parenthetical section (or page) referances with the text would make using and learning your rules much easier. While I'm wishing I'd love a table of content, illustrations, and an index (in that order if you are thinking about it at all)

What, no pop-up book stuff or a DVD-ROM filled with useful GM tools?

[Sma]

I have been using a two sided cheat sheet in my games, which I´m reluctant to put up right now since the rules I´ve been using differ in some points form the ones Frank posted and it would be of little use when the new rules come out. Same goes for a printable laid out version including a TOC.

I intend to update them once the new version of the rules are ironed out. If you want to submit art I´d be more than happy to include that.

[FrankTrollman]

I like the crypto part, seems reasonable. The other thing that you could do, given that storage is free, is assume that every comlink has a symmetric key to a provider burned in, with the provider having a seperate key for each phone on file. Remember, this is a dystopia, the corps have absolutely no reason to allow unlocked phones to be sold. Your phone come with a subscription to the provider, when you stop paying it stops calling.

I would prefer to not go that route because it would make listening to phone calls be a matter of lookup tables. See the MSPs would keep lists of these otherwise unbreakable keys and sell those lists to licensed law enforcement officials. Hackers would need to put together "clans" to collect these keys and disseminate them to other hackers, which means that just to spoof IDs and listen to commcalls you'd need to be part of a group, which I don't want.

The other thing that occurs to me is that having a symmetric key with your MSP would be pretty much pointless, because people could spam your Youtube account with messages of known content, intercept the encrypted versions the MSP sent to your commlink, and then subtract one from the other to get your key of whatever length. I am pretty sure that for a symmetric key to remain secure it has to be used only for private messages between individuals who already have the key.

-Frank

[WeaverMount]

Yeah I fully realize that most of that was way off the deep end. I do think that a little more internal structure would have a big enough impact on the learning curve to make or break people learning and using the rule set, which is I assume the point (IMG:style_emoticons/default/smile.gif)

As for a request, something I would like to see is a little verity in program rating. Under RAW there is no reason to have programs with a rating less than your response. This means that all deckers with an r6 comlink should have identical loadouts; all relevant r6 programs. Including some reason to keep that from happening would be really nice.

A half baked idea I had to this end was to limit nodes to running r/2 programs rounded up at a time, but have software suites that counted as only 1 program. Suits would have a scaling cost depending on the total rating of the programs in the suite. If you have the source code, adjusting the ratings within a suit should be moderately difficult, but fairly trivial for a 400bp decker primary. I'm thinking about a logic + Software tests pegged to the build repair table. Just like pilots are restricted to one type of drone, agents are restricted to the 1 suite they are build around. The base cost per rating of an agent is greatly reduced, but is then multiplied by size of there integral. A really good search bot (is not a couple IPs away from becoming black IC, and it doesn't cost near as much. I'm not quit sure how to mesh this this Frank's program catigories, but I'm sure it can be done. Maybe you allow more programs, but suits must be from one category. Maybe just a +20% cost per category in the suite. I haven't looked at balance much, but if it isn't a deal breaker this does a couple cool things
-Agents loose functionality in a logical way, while gaining flavor
-Deckers have meaningful choices about there load outs
-Makes the software skill meaningful for a decker's who carrier, even in games with little downtime.
-Puts a hurt on script kiddies

[Rotbart van Dain...]

As allowing programming results in people stealing service, which impacts the bottom line, burning in the code is based on a cost model.

Nope. Raising the cost of each unit even one bit will hurt bottomline much more than some people hacking it.

[RunnerPaul]

Because that might make it so that you can't see out them?

Hell, if they can manufacture contact lenses with static images like this or this that do not unduely impede vision, I can't imagine how doing similar with a dynamic display on the outside of the lens would.

[KCKitsune]

Hell, if they can manufacture contact lenses with static images like this or this that do not unduely impede vision, I can't imagine how doing similar with a dynamic display on the outside of the lens would.

You most certainly proved your point.

[JoelHalpern]

The other thing that occurs to me is that having a symmetric key with your MSP would be pretty much pointless, because people could spam your Youtube account with messages of known content, intercept the encrypted versions the MSP sent to your commlink, and then subtract one from the other to get your key of whatever length. I am pretty sure that for a symmetric key to remain secure it has to be used only for private messages between individuals who already have the key.

-Frank

If you are going to use simple one time pads, and simple transforms (because the powerful transforms are broken) then you MUST use different keys every time. So, for that level of security you have to have exchanged LOTs of keying material. And then you use part of it for each message. Even with somewhat more complex transforms (although there are limits), current mathematics is such that if you use the same key for different messages you are very vulnerable to known plaintext attacks. One of the many tools the allies used to keep breaking Ultra was to use messages where they had good guesses about the content.

From the discussion, I had assumed that there were two things going on:
1) Most uses, including most Matrix commerce, used long keys with the "broken" crypto. That is good enough that it takes serious effort to get at the source.
2) Really secure mechanisms use a one time pad, where each part of the pad is used only once. But the transform is sufficiently basic (like xor) that the mathematics can not break it. This requires key material as long as the message.

Note that for storage encryption the second approach is basically useless. I need the key to read the content. And the key is so big that I have to store the key on the computer. So the key and the data are both on the same machine?

Note that one problem that needs to be avoided is inventing the details of any crypto system that is ostensibly used. If we try to invent such details, they will be wrong, and broken. (Inventing details of why something is broken, when we get to invent fictional math, is easy. Trying to invent anything beyond one time pads that actually works is very hard.) The answer therefore may be a whole bunch of fluff that says "this is the net effect, don't ask how" because we probably can not explain what we need to have happen to make the setting work.


Yours,
Joel

[Blade]

Note that for storage encryption the second approach is basically useless. I need the key to read the content. And the key is so big that I have to store the key on the computer. So the key and the data are both on the same machine?

Well it has its uses. For example you can encrypt a file with a OTP, store the file in a place, and then everyone with the OTP, and only them, can decrypt it. In that case, the file will be stored in a node and the user(s) will have the OTP on his(their) commlinks.

Actually it's one of the few possible uses for OTP (even if they're often shown as the ultimate cryptographic method, they're pretty useless).

[FrankTrollman]

I don't think OTPs will be used very often, the amount of times you have to transfer them back and forth makes them at least as vulnerable as more sensible cryptographic systems.

So basically your choices are:

• Symmetric Key: Invulnerable to decryption algorithms, but broken instantaneously if a black hat gets a copy of plain text and the encrypted work. Must be shared before data transfer.
• Asymmetric Key: Can be broken with encryption algorithms, but completely blocks out casual listeners. Can be used with people for whom their is no prior contact; and gaining a copy of uncrypted text and a copy of the encrypted text is not helpful.

Regardless, hardware can be direct hacked at line of sight and forced to give up the keys if it contains them. Retransmitters like Matrix Hubs don't generally have keys, so you have to go on Shadowruns if you want to break the stronger encryption that corporations use for sensitive internal communications.

-Frank

[FrankTrollman]

OK, on the Technomancer thing:

I find the writing and terminology in the technomancy section to be extremely grating. Fortunately, I think that a big chunk of the terminology can be dumped entirely. Frankly, most of that stuff will never be visited in any meaningful fashion in later publications. So here's a list:

Networks: This is a perfectly acceptable name for technomancer "groups" it may or may not appear in later books, but it's fine. We can avoid going against the grain here and just throw up networks as the standard name. t works fine. As a caveat however, it should be noted that guilds and parties is insulting and retarded: that is the name for clans and teams respectively that people use in fantasy MMOs. Obviously some jackass played a bunch of WoW and forgot that groups who play in futuristic settings (like Unreal) or even modern settings belong to Clans and not Guilds. So the terminology of Guild and Party is right out the door. Technomancers are inherently futurists, not a bunch of World of Warcraft hangers on.

Dissonance: The entire tirade about how Dissonance Technomancers are totally of the dark side and wicked through and through has no place whatsoever in the shades-of-gray world that is Shadowrun. Fortunately, I seriously don't have to care about that because even having Catalyst subsequently print Dissonant characters who are irredeemably evil would just be one isolated case of bad writing and wouldn't be incompatible with Dissonance users as a whole being complex people with realistic goals. All the subgroups of Dissonance users have incredibly stupid names running from merely retarded (infektors) to downright religiously insensitive (Discordians), fortunately it doesn't matter at all because those streams don't actually have any rules or coherent philosophies attached to them so they are incredibly unlikely to come up importantly in ongoing plot materials. So I can quite happily drop all that and replace it with something that is less offensive and shitty.

Streams: I'm perfectly happy to name the traditions "streams" instead of "songs" to retain forward compatibility. The Streams themselves are annoying. No one is going to name their entire philosophy an inane pun like "Sourcery" that's lame. Putting elektro-puns into words stopped being cool in like 1985. So while none of the streams are going to be called "E-scapists" because that's juvenile, there are going to be some tradition equivalents which will suppose be named "Streams." I don't think that I will be handing out different drain attributes, because Sprite differences and Greatform power choices differentiate streams enough under these rules without making Elves even better at Fading Resistance.

Sprites: Sprite names are totally immaterial, and the very idea of playing enough WoW that you think that having a Paladin and Tank sprite is a good idea is offensive to me. Not only do I not have to roll the new sprites into the picture, I'm really happy that I don't have to use the new Sprite names.

Paragons: It may actually come up that the Mentors of technomancers are called "Paragons" so that terminology may as well be used. It works at all, so I'm actually fine with it. The presented Paragons themselves are by and large dreadful, especially the Dissonance ones, and will not be used.

I think that about covers it, terminology wise.

-Frank

[Jaid]

i may not agree with a lot of frank's problems with unwired (but we've already been there), but i have to agree with you frank... some of the names are pretty lame...

i do have to say, the whole dissonance thing though, i'm not seeing how you're going to make that reasonable. and i'm not entirely sure why it would be worth bothering; it's not that they're evil, really. i mean, i could make a TM who uses a psychotropic blackout to program people to be his slaves, and he wouldn't be dissonant, just evil. dissonant TMs do stupid things like randomly infecting people's commlinks with a virus just to taint the matrix or whatever.

so yeah, i'm just not seeing how it's really salvageable to create reasonable PCs, NPCs, villains, contacts, etc other than essentially declaring it to be just another stream.

[PlatonicPimp]

He doesn't have to make it reasonable, just not necessarily evil. Dissonance not as evil, but as chaos. There's good, fun chaos and bad chaos. When I think of dissonance I think of 4-chan and griefers.

[kzt]

[*] Symmetric Key: Invulnerable to decryption algorithms, but broken instantaneously if a black hat gets a copy of plain text and the encrypted work. Must be shared before data transfer.

Sorry, been spending my time updating software images on hundreds of switches, so haven' been paying enough attention to this thread.

Any cryptosystem that can be broken by one piece of chosen plaintext is worthless. People find attacks that that require only several hundred million plaintexts to be pretty clever. For example (stealing shamelessly from Wiki on AES)

"In April 2005, D.J. Bernstein announced a cache timing attack that he used to break a custom server that used OpenSSL's AES encryption.[10] The custom server was designed to give out as much timing information as possible, and the attack required over 200 million chosen plaintexts. Some say the attack is not practical over the internet with a distance of one or more hops;[11] Bruce Schneier called the research a "nice timing attack."

This was a deliberately crappy implementation of AES and still took over 200 million plaintexts. Well done implementations don't have this weakness. Good crypto systems that are well implemented are really hard to attack, as the people doing them are usually as smart and more experienced then virtually anyone who would attack the system and don't make the kind of mistakes that allow easy access. That's why everyone (not in management) in the computer industry tend to spit on vendors who offer "proprietary encryption" instead of modern well known systems, as there are lots of idiots who think they know what they are doing and produce crap.

Anything that a corporation is willing to pay professional criminals to steal from another corporation, if it's encrypted at all, is likely to be encrypted by a decent system. Not always, executives and salespeople are oftentimes are a really bad combination...

Anyhow, the key distro issue is a real problem and it's one of the reasons why very few companies or governments actually have good wide-spread crypto in place. It's hard and expensive to do this right. There is always a customer for the easy and cheap...

[Sma]

Streams: I'm perfectly happy to name the traditions "streams" instead of "songs" to retain forward compatibility. The Streams themselves are annoying. No one is going to name their entire philosophy an inane pun like "Sourcery" that's lame. Putting elektro-puns into words stopped being cool in like 1985. So while none of the streams are going to be called "E-scapists" because that's juvenile, there are going to be some tradition equivalents which will suppose be named "Streams." I don't think that I will be handing out different drain attributes, because Sprite differences and Greatform power choices differentiate streams enough under these rules without making Elves even better at Fading Resistance.

One thing I´d like to look into is having the rolls for techomancer exclusive complex be composed of Logic + Skill instead of Resonance + Skill. And by I, I of course mean you since I´m already liking the idea and haven´t seen any problems in the one (1) game I´ve tried it in. I regard to Stream or song names I´d rather have them be the same as the Mentors or if that´s out having them be in the style you had in the ends of the matrix.

Sprites: Sprite names are totally immaterial, and the very idea of playing enough WoW that you think that having a Paladin and Tank sprite is a good idea is offensive to me. Not only do I not have to roll the new sprites into the picture, I'm really happy that I don't have to use the new Sprite names.

In the interest of compatibility it probably wouldn't hurt to draw up an equivalency list. But since you´ll pretty much have to rebuild any technomancer and hacker that comes with a boxed adventure anyway that should be really low on the to do list. What would come in really handy is some guidelines to choosing your sprites. I had to draw myself a diagram to be sure I covered the required complex forms the first couple of times I made a technomancer. This gets easier once you are familiar with the rules. But a short descriptive blurb with each sprite still wouldn't hurt the reader. On related note, the ordering of the rules is something I noticed made it difficult to reference at the table. I have no idea what the solution would be though. Should a a draft of the revised rules becomes available, I can make one of the players in my group who only semiinterested in the matrix read through it and give comments on where he´d look for stuff if you are interested. The players actually playing deckers are eagerly awaiting any rules updates. Just saying.

the other stuff

no quibbles.

[Sma]

Any cryptosystem that can be broken by one piece of chosen plaintext is worthless.

But I don´t want realistic encryption in a game I sit down to play on a Sunday afternoon. Having a system where you manage to trick your opponent into saying Beetlejuice three times and encrypting that with one of his OTPs, thus allowing to get into his secret datastore is fun. Having to do it 200 million times isn´t.

Using One Time Pads would allow you to have the electronic correspond to an actual physical fob someone carries in his pocket allowing the rest of the team to play a part in cracking the corp node by stealing the fob long enough to make a copy. This encourages teamwork and gives everyone his 15 minutes.


Which brings me to a different point.

A Decker who uses Intercept and Impersonate on a drone is in a position to hand off this information to the teams rigger thus allowing him to share his progress in a meaningful fashion. Now what would make the current rules three kinds of awesome for me would be:
Ways for nondeckers (who bring dicepools of 2-4 to the table) to help a decker doing matrixworks without resorting to teamwork tests.
Ways for the decker to enable others to do matrixwork so both parties profit, and are able to do things without the decker making all the decisions.

[Heath Robinson]

He doesn't have to make it reasonable, just not necessarily evil. Dissonance not as evil, but as chaos. There's good, fun chaos and bad chaos. When I think of dissonance I think of 4-chan and griefers.

Dissonance is a belief or faith, exactly as Resonance is. There's actually a paragon that pretty much represents 4chan; Alias. Griefers are just people acting like people, taking enjoyment in being dicks to each other.

Dissonance, it's meant to be the Toxic analogy and, in SR, destroying nature is evil. Because the AmerIndians were right. Yup, that's the best reason I can find, white guilt is a founding part of SR.

[kzt]

Using One Time Pads would allow you to have the electronic correspond to an actual physical fob someone carries in his pocket allowing the rest of the team to play a part in cracking the corp node by stealing the fob long enough to make a copy. This encourages teamwork and gives everyone his 15 minutes.

You can do that by stealing someone's commlink. I don't know of anyone who actually really locks down their smartphone, though it's certainly possible. You swipe it for a few minutes you get their keys and login credentials. You might have to hack in right then, while he's "occupied", but it allows multiple people to get involved. I had a scenario I never ran where the Johnson wanted the documentation a guy who was a high powered security consultant had on a company's system, but also wanted nothing bad to happen to the guy and nobody to know it was stolen. Of course, he had a lot more than just that one file the Johnson wanted on his commlink.

[Blade]

I don't know of anyone who actually really locks down their smartphone, though it's certainly possible.

I don't know of anyone who does either, but I don't know of anyone who uses it to pay either. They use credit cards to pay, and these credit cards need a 4 digit code to work.
A commlink isn't just a smartphone and if you add this to the fact that you can have easy user authentication (biometric data, thoughts commands...) I guess that commlinks would probably be locked down more often than smartphones.