 Hacker Theory, Why this is possible. |
  |
 Jan 29 2007, 11:49 PM
Post
#126
|
|||||
|
Running Target  Group: Members Posts: 1,498 Joined: 4-August 05 From: ADL Member No.: 7,534  |
Now I am getting angry. The following is very easy to understand: Hacker: skill + program + edge Agent instructed by Joe Normal: Agent rating + program -> agent rating is replacing skill What do skillwires do? They replace skill and forbid specialization and use of edge. Rules-wise 100% comparable. WHAT is your problem Rotbart? Do you just want to troll arround? |
||||
 |
 
|
||||
|
Jan 29 2007, 11:56 PM
Post
#127
|
|||
|
Running Target Group: Members Posts: 1,498 Joined: 4-August 05 From: ADL Member No.: 7,534 |
Why cant you load agents on the credstick and run software? Its a node. Running software and Agents is what nodes are for. What should prevent it from connecting to node X? You can even store data on your gun! (Isnt that mentioned in the BBB explcitly? Forgot the page) Read my "proof quotes" one page 3 or 4 of this thread to find the exact SR4 BBB quotes. Thats really how SR4 matrix world was designed. And it is good that way, except for the System < Response rule. You may say that you dissallow this or that, but thats your own houserules. I solve the whole problem by giving a credstick System 6, Signal 3, Response 1 and Firewall 6. Thats exactly what it should be, a high security but very low performance node. Of course that is a house rule (mostly because Response is lower than System) And not everything is a commlink, because a comlink has an interface, a camera, a gps a microphone and lots of other stuff. But everything is a node. And apart from the System/Response problem that is a very good and streamlined solution. In one thing you are right: They dont have the same functions. But thats only hardware functions. A commlink cant make coffe and a coffe machine cant make pictures, but both can connect to the matrix, and run programs and software and can be hacked. And both can run a persona (it doesnt even cost precessing power in SR4) when connected to a brain. |
||
|
|
|
||
|
Jan 30 2007, 12:22 AM
Post
#128
|
|||
|
Hoppelhäschen 5000 Group: Members Posts: 5,807 Joined: 3-January 04 Member No.: 5,951 |
There is the mistake - an Agent does not replace a skill, it replaces the operator, including the skill. This has certain consequences within the rules you choose to ignore to make your simplified 'comparision' work. |
||
|
|
|
||
|
Jan 30 2007, 01:10 AM
Post
#129
|
|
|
Running Target Group: Members Posts: 1,498 Joined: 4-August 05 From: ADL Member No.: 7,534 |
It doesnt matter. The SC is directing the agent roll by roll. The fluff description may be different but the result is the same.
I see no "consequnces". If you know some, name them (and I really wonder why you did not do this in your last posting. Do you want to be misunderstood, or what is your intention?) |
|
|
|
|
Jan 30 2007, 02:44 AM
Post
#130
|
|||
|
Moving Target Group: Members Posts: 745 Joined: 2-January 07 From: Los Angeles, CA Member No.: 10,510 |
Serbitar, Rotbart already gave a perfectly reasonable argument that you haven't responded to. I think you might have overlooked it because it wasn't structured like yours were. Plus, I think it was actually in response to my message. He postulated that skillwires replace 'memory' not skill. Thus a hacker might have memories of an expert hacker hacking a similar system, and based on that make conscious decisions about how to hack that system. He further theorizes that an agent cannot make decisions based on that prior history in the same manner that a human operator can. Personally, I don't agree with his interpretations, but I think they're worth at least a response and, if refuted, do tear a hole in your skillwire = agent skill argument (which I still think is a perfectly valid and striking point). But don't accuse him unjustly of trying to muddle the argument - he did put forth an opinion and arguments to back it up. (To Rotbart: If I am misunderstanding your previous statements, please feel free to clarify the issue. I just don't like to see misunderstandings in what is otherwise a very good debate.) |
||
|
|
|
||
|
Jan 30 2007, 02:55 AM
Post
#131
|
|||
|
Moving Target Group: Members Posts: 745 Joined: 2-January 07 From: Los Angeles, CA Member No.: 10,510 |
Assuming I'm right about your objections, Rotbart, I'm going to put forth a cause-effect stream. Feel free to identify which parts you disagree with and why: 1) Agents have Artificial Intelligence. 2) Agents operate independently of a hacker, once given instructions. 3) Artificially Intelligent Agents can make decisions in roughly a similar manner to that which "actually" Intelligent people are capable of. 4) Artificially Intelligent agents can be more intelligent than people. 5) For purposes of hacking, equating intellectual decision-making capabilities to the Logic attribute is not a far stretch. 6) An agent with high Logic attributes has the same decision making powers of a high Logic hacker. 7) Skillwires grant the 'memory' of the use of a skill. 8) Deductive reasoning powers to decide on probably outcomes and decisions based on past data is a function of Logic, and therefore applicable to agents with this 'memory'. 9) The agent can use its Logic attribute for Logic based tests. 10) The agent has a hacking skill. 11) Decision making powers (Logic), knowledge/experience (skillsoft 'memory'), and independent operation (which all agents have, more or less) are all that is required to learn a Logic-based skill. 12) Agents can learn a hacking skill as easily as a skillwire. --- Personally, I don't see the need to define any of this. I think the fact that we can record skills as programs is pretty obvious - whether you call these skills program ratings, device ratings, autosofts, or skillwires is pretty much irrelevant to me. For the most part that's just nitpicking over the interface. It's like arguing that you have a different document, just because one is handwritten and one is faxed. |
||
|
|
|
||
|
Jan 30 2007, 03:08 AM
Post
#132
|
|
|
Runner Group: Members Posts: 2,526 Joined: 9-April 06 From: McGuire AFB, NJ Member No.: 8,445 |
Agents are just full of If=X - Then=Y.
"If = under attack, Then = Perform 'attack back'; Perform 'print to system'" "If = Attack Back, Then = Perform 'Target'; Perform 'Attack Target'" Ect... The ammount of If/Then the agent has is dependant on what rank it is, with a Rank 6 having the most situations available to it. |
|
|
|
|
Jan 30 2007, 03:38 AM
Post
#133
|
|||
|
Prime Runner Group: Banned Posts: 3,732 Joined: 1-September 05 From: Prague, Czech Republic Member No.: 7,665 |
Whoa, so that's what an argument from Looking-Glass Land looks like. Are you saying that Rotbart is claiming that an Agent is not like a Skillwire system because a Skillwire has a flavor-text description about how it prevents you from learning anything while you're using it and an Agent that you're directing has no such limitation? That's not a argument at all. -Frank |
||
|
|
|
||
|
Jan 30 2007, 04:07 AM
Post
#134
|
|||
|
Moving Target Group: Members Posts: 745 Joined: 2-January 07 From: Los Angeles, CA Member No.: 10,510 |
I don't think that was said by anybody, now. Serbiter said (and, of course, I'm paraphrasing, and not accurately: ) "this is basically how a Skillwire works and it supports my argument that I was making before." Rotbart said (paraphrasing again, as best I can: ) "that's not what a skillwire is," proceed to give a very brief and somewhat vague opinion on what a skillwire is and state it as fact, and then reason that if the skillwires worked that way (which he assumed to be true) then the way a skillwire works does not support Serbiter's argument. The word 'memory' was used, but at no time did anyone discuss learning anything, or speculate about whether or not an agent can learn. (Although, now that you mention it, such speculation does throw a hole in my Logic=intelligence formula in a previous post.) He did, however, I believe, numerously mention that agents suffer from inherent limitations that hackers do not. These dintinguishing limitations, apparently, for reasons relating to viewing the skillwires as recorded memories, do not allow an agent to use skillwires. (I feel weird arguing somebody else's case. :)) |
||
|
|
|
||
|
Jan 30 2007, 04:40 AM
Post
#135
|
|||||
|
Moving Target Group: Members Posts: 101 Joined: 30-December 06 Member No.: 10,493 |
One cure to the "Agent Smith" problem is the "independant agent" solution: Once you get software entities intelligent enough to act a a person (e.g. a hacker), those entities stop following orders. Modern "expert systems" can do amazing things, but they lack the reasoning capabilities to act as a person. Webcrawlers are great for search engines and databases, but that's about it. Spambots can spam, but nothing else. Viruses basically just act as viruses. It is possibile that anything capable of acting as a human in many capacities at once, unlike the aformentioned bots, will need a strong AI. When your comlink spawns 6 million AI agent processes, those processes go off and do whatever the fuck they want. Too bad for you. So, how do SR agents get used? It's as Frank says: Dumb AI processes being coordinated by your comlink, and acting in their limited capacities. If you spawn 6 million of them, they do nothing because you can't coordinate them and they can't coordinate themselves effectively (because you can't tell them what to do fast enough). Obviously not RAW, just a thought. |
||||
|
|
|
||||
|
Jan 30 2007, 05:09 AM
Post
#136
|
|
|
Moving Target Group: Members Posts: 941 Joined: 25-January 07 Member No.: 10,765 |
I'm the new guy around here, but I'll weigh in.
Now, I'm not the computer geek some of you guys are, and I haven't (sadly) had much time to shake the book and see what falls out like you guys do, but I'll address at least two arguments that have come up repeatedly here as I see them... you know, from the outside and without falling into one of the two camps. Technomancers are 'broken' because they cost too much to play. Bullshit. Sorry, bull... fucking... shit. I'm not gonna pull your leg and say they are better, worse or what not, but they are hardly broken. Sure, he's not going to be running around with straight sixes in his hacking stats, thus no 6/6/6/6 commlink equivlent, then again a bunch of hackers are probably gonna think to themselves that they can wait for that six in logic... that last dice in teh pool just ain't worth it to start, ya know? Maybe, maybe not. On the other hand, I don't see a 6/6/6/6 commlink in the book either, meaning your brand new old school hacker is gonna have to wait until the game starts to get one too. What you ARE going to see is a more dedicated TM over a more jack of all trades hacker. And unlike the hacker, the TM can come up with CF's he needs on the fly. At a guess, you'll also see that when the Karma starts rolling in the TM starts to outstrip the hacker, since resonance isn't capped. Credsticks and coffee machines as uber commlinks: Here I think it's a case of reading WAAAAAYYYYYY too much into the fluff. I'm sorta new at this debate, but: If, as a GM, I had a player tell me his credstick was a 6/6/6/6 commlink, instead of just an incredibly well protected dedicated purpose device, I'd laugh at him. Flat out. I wouldn't even bother with the ratings for the other stuff, the Device rating represnts how hard it is to hack, nothing more, nothing less. Comes from not trying to read into things. Here's how I see it. Sure, you credstick MIGHT have 6/6/6/6, but it's designed to do one thing and one thing only. It's not designed to work as an interface (as compared to a node...) in the wireless world. Every rating it has reflects its singular purpose (protecting your money/ID and transmitting it when authorized). So, if you want to turn your credstick into a commlink, in addition to potentially wrecking your bank account balance due to failsafes behind the encryption, you'll spend an extended check and the money tryign to go back and install all the things that make a commlink a fucking commlink and not a credstick. You know, Sim rig (last I checked your credstick lacked that...), additional memory for your actual copies of software, an OS designed to support hacking, etc. Same thing with the coffee machine. Sure, your commlink may steal memory and processing power from the coffee maker to run shit... but the Commlink is the one actually doing the running, not the damn bean brewer. As to wether or not the rules are actually broken or not, well... I've got to do some research on the 'Agent Smith Problem' and other things first. I still suspect you guys are trying too hard. I've got more issues with the damn layout of the book than munchkins running rampant. ;) |
|
|
|
|
Jan 30 2007, 05:16 AM
Post
#137
|
|||||
|
Moving Target Group: Members Posts: 745 Joined: 2-January 07 From: Los Angeles, CA Member No.: 10,510 |
I don't know, according to Frank, it's all pretty ubiquitous... (Just kidding guys.)
Oohhh... we're having a bad influence. :P Welcome to dumpshock! |
||||
|
|
|
||||
|
Jan 30 2007, 05:23 AM
Post
#138
|
|
|
Moving Target Group: Members Posts: 941 Joined: 25-January 07 Member No.: 10,765 |
Thanks, It only took me about a week to get approval.... :eek:
|
|
|
|
|
Jan 30 2007, 05:24 AM
Post
#139
|
|||
|
Moving Target Group: Members Posts: 745 Joined: 2-January 07 From: Los Angeles, CA Member No.: 10,510 |
It's to keep out the rats. :) Yeah, I hated that process, too. |
||
|
|
|
||
|
Jan 30 2007, 07:04 AM
Post
#140
|
|||
|
Great Dragon Group: Members Posts: 5,537 Joined: 27-August 06 From: Albuquerque NM Member No.: 9,234 |
Actually, they can do damn near anything you want as long as it isn't subtle. You can tell them to do a denial of service attack or attempt to hack into a corp. For example, attach to this lists of nodes and try to hack in using this set of programs. Naturally you have each of them trying it a bit differently, but it's all done in a pre-programmed fashion. If you really control 6 million devices I would tend to suspect that you are going to at the very least make it damn near impossible for anyone trying to keep hackers out from noticing anything that is less subtle. And there are few things less subtle than 6 million agents trying to brute-force their way in. In most cases I'd expect you will get significant success in breaking in, at which point the agent promptly goes and edits the node code, hacks the logs, and sends a message back telling the controller that it broke in and how to exploit this. As the security guy at our place points out, one person walking around trying the doors of cars in a building parking lot has a limited likelihood of success, and hence a limited cost/benefit ratio. If you can automatically try every car door in the entire city every second your chance of success is just a wee bit higher. With a botnet, or a horde of agents, you can. That's why botnets really suck. And agents are essentially just bots. |
||
|
|
|
||
|
Jan 30 2007, 08:32 AM
Post
#141
|
|
|
Running Target Group: Members Posts: 1,498 Joined: 4-August 05 From: ADL Member No.: 7,534 |
cetiah:
once gain, my reasoning goes like this: Agents that are told exactly what to do (roll this, then roll that) have exactly the same effective game mechanics like skillwires, as they replace your skill, and prevent you from using edge, but still you have complete freedom of choice what to do. BUT you can have them to rating 6, although you can have only skillwires to 4. THAT is my point. Nothing else. That agents are not skillwires at all and have completely different fluff descriptions does not change that they have exactly the same game mechanics when looked at dice numbers. Memory and stuff is fluff. Dice numbers dont care about fluff. |
|
|
|
|
Jan 30 2007, 08:41 AM
Post
#142
|
|||||||||||||||||
|
Running Target Group: Members Posts: 1,498 Joined: 4-August 05 From: ADL Member No.: 7,534 |
Do the math. Post it. Post a starting TM and a starting hacker. Count BPs. Count Karma till the TM equals the starting hacker. Then we will see. I did the math. Often. I wont accept your reasoning without any proof.
Could you please post the section from BBB that tells you this? These are your house rules. So a credstick is not a node? Read my quotes further below. Additonal memory? In SR4 memory is abundant even in your gun. an OS to support hacking? The credstick already has rating 6/6 OS. With the simrig part you are right though, but thats hardware.
This is even mroe housruled. So you cant run programs on any node there is? When does a node qualify for running software? Does it need the "Can run software" tag? Please give proof from SR4. You can not just make statements and claim they are RAW when they arent. here is SR4:
Conclusion: Everything is a node. A node is a CPU. You can run software on it. The node does not care at all what software it is. You can not just invent stuff and say that it is obvious. |
||||||||||||||||
|
|
|
||||||||||||||||
|
Jan 30 2007, 08:55 AM
Post
#143
|
|||||||||
|
Hoppelhäschen 5000 Group: Members Posts: 5,807 Joined: 3-January 04 Member No.: 5,951 |
By what judgement if the SC does not possess the skill? The one making the technical decisions is still the Agent.
Actions are crunch, so had you hate them wenn boosted people have more in martix and vehicles.
Read again.
It's a completly different setup with possibilities different to each other, taking different Actions and introducing different weaknesses (Spoofing & Combat). The rules are not the same for giving orders and doing things yourself. |
||||||||
|
|
|
||||||||
|
Jan 30 2007, 09:34 AM
Post
#144
|
|
|
Running Target Group: Members Posts: 1,498 Joined: 4-August 05 From: ADL Member No.: 7,534 |
I am not talking about the oder by command software version. I am talking about the order by "order" version.
And again, please give an example. |
|
|
|
|
Jan 30 2007, 09:47 AM
Post
#145
|
|||||||
|
Hoppelhäschen 5000 Group: Members Posts: 5,807 Joined: 3-January 04 Member No.: 5,951 |
There is no such thing. You can remote control or issue orders, but both require the Command program.
No, that's the first time you are asking...
..for a situational example - which you are to give, since you claimed something. Let me set the specific situation: AR 'Hacker' uses Skillwires + Exploit to probe a target vs. AR 'Hacker' uses Agent + Exploit to probe a target. Assume all Ratings to be 4, 'Hacker' has 1 IP. |
||||||
|
|
|
||||||
|
Jan 30 2007, 09:55 AM
Post
#146
|
|
|
Runner Group: Members Posts: 3,009 Joined: 25-September 06 From: Paris, France Member No.: 9,466 |
Actually, I seem to recall that the Command program isn't required to issue orders. It is only used for remote control.
I'm sorry I can't give you the precise references right now, but I remember spending quite some time to read everything related to this and getting to that conclusion. |
|
|
|
|
Jan 30 2007, 10:17 AM
Post
#147
|
|||||||
|
Hoppelhäschen 5000 Group: Members Posts: 5,807 Joined: 3-January 04 Member No.: 5,951 |
You're right - so it actually boils down to this:
The latter is the only way to do so anyways, not using the Commad program. |
||||||
|
|
|
||||||
|
Jan 30 2007, 04:11 PM
Post
#148
|
|||||||
|
Running Target Group: Members Posts: 1,498 Joined: 4-August 05 From: ADL Member No.: 7,534 |
Yeah! You got it! |
||||||
|
|
|
||||||
|
Jan 30 2007, 04:18 PM
Post
#149
|
|||||||||
|
Running Target Group: Members Posts: 1,498 Joined: 4-August 05 From: ADL Member No.: 7,534 |
no:
Well, in this case the Agent is even better than skillwires, as it is always working at VR speeds, so you might consider a VR example. But concerning the example: Both roll 8 dice. The skillwire guy rolling hacking(skillwired) + exploit, the guy with the agent buy saying: Hey agent hack in there and letting roll the agent rating + exploit. Once a problem occurs, the hacker with skillwires just does what he wants to do, while the hacker is watching the agent via VR or AR and is telling the agent what to do. Would you now please post the problems with the: "controlled agent = skillwires" (again, rues wise, number of dice and game mechanics, not fluff) comparison? Otherwise I am really beginning to waste my time. |
||||||||
|
|
|
||||||||
|
Jan 30 2007, 04:47 PM
Post
#150
|
|
|
Moving Target Group: Members Posts: 941 Joined: 25-January 07 Member No.: 10,765 |
Serbitar:
I'm at work now but I'll get to the number crunching on the characters as soon as I get a chance. As for the rest, I wasn't really talking house rules, I'm talking an off the freaking cuff answer that popped into my head. If you take a hundred shadowrun GM's, if one guy actually allows someone to use a credstick as a commlink, that guy is an exceptional moron, and I wouldn't want to play with him anyway. Now, to clear up confusion, I didn't say the credstick wasn't a node. I accept it is a node. I said it lacked proper interfaces to be a commlink. Huge difference. It lacks the OS to run most applications directly. Its a dedicated server. Sure, you can steal it's ram (or whatever) to store data on, but thats about it. It's not a 'house rule' its common sense. Let me put it too you this way. I have an MP3 player. It's essentially a solid state external drive, yeah? Now, I can go buy a solid state external hard drive, set it up with an OS and run it as a computer... if I hook it up to a computer first. It's a bit harder with the MP3 player. In fact, it's impossible to use the MP3 player as a computer without ripping the bastared apart and doing all sorts of crazy shit to it... so much that it would be easier to use a damn computer in the first place. Yes, it has an OS on it, one designed solely to facilitate playing music. Yes, it has memory... no problems there, yes it even has an interface... again, one solely dedicated to playing music. It's an idiot savant of a computer, totally useless outside of its purpose. Same thing with credsticks. Common sense, that's all it takes. Same thing with devices. Sure... they are wireless, sure they are nodes. Sure, you can interact with them, store data on them, even run software on their ram (or whatever) as a periphrial in your net... but you still need the central processor with the OS and the interface (simnet/keyboard/datajack) to actually DO anything. Now, to me that means the simple fix to all of these debates about using credsticks and coffeemakers as commlinks is to insert one single line that says: "Must have a commlink to actually interact with the wired world." If only to replace the use of 'common sense' as a 'house rule'. For all I know, they DID say that somewhere and every one of us missed it because the damn chapter is organized like scrambled eggs. |
|
|
|
|
|
Lo-Fi Version | Time is now: 1st February 2025 - 08:44 AM |
Topps, Inc has sole ownership of the names, logo, artwork, marks, photographs, sounds, audio, video and/or any proprietary material used in connection with the game Shadowrun. Topps, Inc has granted permission to the Dumpshock Forums to use such names, logos, artwork, marks and/or any proprietary materials for promotional and informational purposes on its website but does not endorse, and is not affiliated with the Dumpshock Forums in any official capacity whatsoever.