 Hacker Theory, Why this is possible. |
  |
 Jan 27 2007, 01:37 PM
Post
#26
|
|
|
Running Target  Group: Members Posts: 1,498 Joined: 4-August 05 From: ADL Member No.: 7,534  |
Well, "adding everything" together is not that easy when every single thing in SR4 has some processing power, and is running some programs.
|
 |
 
|
|
Jan 27 2007, 01:53 PM
Post
#27
|
|
|
Hoppelhäschen 5000 Group: Members Posts: 5,807 Joined: 3-January 04 Member No.: 5,951 |
If you go that route, it becomes quite easy:
You declare your whole PAN a single node, adjust response to the amount of devices and if you loose significant amounts of them, in addition to loosing response, your persona gets permanent damage boxes that last until you reboot... let's say 1 box for every 10%. |
|
|
|
|
Jan 27 2007, 03:38 PM
Post
#28
|
|
|
Running Target Group: Members Posts: 1,498 Joined: 4-August 05 From: ADL Member No.: 7,534 |
Thats a very coarse grained method . . . and not capped at all.
|
|
|
|
|
Jan 27 2007, 03:42 PM
Post
#29
|
|||
|
Hoppelhäschen 5000 Group: Members Posts: 5,807 Joined: 3-January 04 Member No.: 5,951 |
Indeed. That's the problem. |
||
|
|
|
||
|
Jan 27 2007, 04:33 PM
Post
#30
|
|||
|
Prime Runner Group: Banned Posts: 3,732 Joined: 1-September 05 From: Prague, Czech Republic Member No.: 7,665 |
OK, this assessment is something I just don't understand. With the mesh client, you literally don't know or care where programs are or what they are doing. The only important thing is that if you are jammed, your programs don't work on anything you lack a wired connection to. That seems... intuitively obvious. Where things get overly complicated is the other way. If programs are literally running on your commlink, then there's nothing stopping you from having 3 or 4 commlinks that are all running different program suites. I'm fond of having a "Medic Bot" for example, that simply runs an agent setup to interact with an unrelated node your icon is subscribed to that runs Medic on your icon every round. That's... complicated. But that sort of thing is also inevitable if the system limits of your commlink are literally the limits of what your comm can have on it at one time rather than the limits of the number of things your comm can coordinate. If it's actually important what processing or programming is located on each piece of computational hardware, then obviously micromanagement increases beyond the complexity threshold that I am willing to deal with. -Frank |
||
|
|
|
||
|
Jan 27 2007, 05:14 PM
Post
#31
|
|||
|
Moving Target Group: Members Posts: 745 Joined: 2-January 07 From: Los Angeles, CA Member No.: 10,510 |
I've been thinking of giving a -2 "distraction bonus" to using devices that are not plugged into your PAN, since I figure AR-networked has become the norm of 2070. I haven't done it yet; just a thought I've been playing with. Maybe there would be a quality or something for people this doesn't apply to. |
||
|
|
|
||
|
Jan 27 2007, 05:29 PM
Post
#32
|
|
|
Hoppelhäschen 5000 Group: Members Posts: 5,807 Joined: 3-January 04 Member No.: 5,951 |
That's great. Let's invent even more extra rules concerning electronics usage... :S
|
|
|
|
|
Jan 27 2007, 05:34 PM
Post
#33
|
|||||
|
Running Target Group: Members Posts: 1,498 Joined: 4-August 05 From: ADL Member No.: 7,534 |
I have actually no idea what you have in mind specifically. Could you outline somehow how you would balance the whole thing? How do you define "number of entities that your comlink can manage?". |
||||
|
|
|
||||
|
Jan 27 2007, 06:51 PM
Post
#34
|
|||
|
Prime Runner Group: Banned Posts: 3,732 Joined: 1-September 05 From: Prague, Czech Republic Member No.: 7,665 |
Sure. The number one balance problem in Shadowrun hacking is something that we shall call the "Agent Smith Problem" - in reference to a movie that references Shadowrun, thereby going full circle. Here's the Agent Smith Problem: Agents are a Hacker's answer to Spirits, Sprites, or Drones. And like the others in the game, Agents are limited in number. Unfortunately, while Spirits or Sprites are limited to inviolate factors such as attributes, and even Drones are very meaningfully limited by space and cash flow, the fact is that the limiting factor on Agents is apparently processing capacity. Agents take up computer memory alone, which is a factor so readily available that the rules do not bother to even measure it. Indeed, parallel processing is available in practically limitless quantities. There is literally no meaningful limit to the number of Agent copies that can exist at one time. You can have one on your commlink, another on your coffee machine, another on your credstick, another in your car, another in your helmet, another in your olfactory booster, etc. without limit or sanity. If each Agent can accomplish anything (and as long as programs like Medic and Attack exist, they can), then there is no limit to what one Hacker can accomplish. And there is no limit to what the corporations can accomplish either. Recall that you can keep an Agent prowling out of your wristwatch, and the mainframes in Shiawase central are essentially just thousands or millions of wristwatches networked together. If IC exists at all, it might as well be a cascading horde of thousands or millions of Agents each trying to get lucky with a low rating Attack - dumping your ass out of the 'trix as soon as the tenth-part of the Smith core gets even a single action. Obviously, that can't be how it works. ---- So how does it work? Well, that's where the limits of the commlink's subscription comes in. Sure, there are an infinite number of Agent programs available. But they only do what they are told, and they are only told to do things when your commlink has an open window of communication with them. So if you are benefitting from a lot of Agents, you experience slowdown because you are "running" extra programs. Even though those Agents are physically located on other computers (thr milk freshometer in your refirgerator, for example). Similarly, the mainframes at Shiawase have virtually unlimited processing power in total, but have a limited coordination power against any specific portion of the net. --- Running Trucks Through the Rules: Obviously, in order to get more network action you could jolly well run a better system. But the other way to get more accomplished is to throw down a second Hacker. Every additional reference frame comes with an entirely new set of maximum subscriptions, so having even a weak hacker backing you up is a potential launching pad for Agents, allowing secondary characters and security hackers to provide a very real benefit. While that could be construed as abuse, I actually like it. It encourages people to hack with the whole team on major hacks. -Frank |
||
|
|
|
||
|
Jan 27 2007, 06:57 PM
Post
#35
|
|||
|
Hoppelhäschen 5000 Group: Members Posts: 5,807 Joined: 3-January 04 Member No.: 5,951 |
You do realize that only Agents loaded in your Persona travel with it? So, unless those independent Agents leave their node and enter the same as you do, they are twiddling thumbs... the only thing they can do is the medic trick. |
||
|
|
|
||
|
Jan 28 2007, 01:26 AM
Post
#36
|
|
|
Running Target Group: Members Posts: 1,498 Joined: 4-August 05 From: ADL Member No.: 7,534 |
Personally I do not see a problem with lots of agents on the hackers side. For me, hacking is all about stealth. Once you involve lots of agents, they will find you and can just shut down the host.
The main problem with agents is on the corps side. How to prevent them from ahving a gazillion agents in their security nodes? @Frank My problem with your conception is: It is a must that some nodes can have connections to hundred or thousands of other nodes at the same time. Matrix cafes work that way, databases work that way and so on. So any restrictions that are below this limit are not really helpful. |
|
|
|
|
Jan 28 2007, 01:33 AM
Post
#37
|
|||
|
Prime Runner Group: Banned Posts: 3,732 Joined: 1-September 05 From: Prague, Czech Republic Member No.: 7,665 |
OK, now I'm confused. Nodes can connect to as many locations as they want. Indeed, in a wireless world it is entirely possible that a particular node will be within range to accept signals from thousands of devices. The limitation is only on how many programs you can orchestrate and simultaneously command. So while there is doubtless a Track program running somewhere, if it exceeds your personal commlink's limits it is not running for you. So long as there is a limit on how many programs are running within the conceptual space of any particular reference frame, it's all good. The Matrix has topology, not topography. Each acting party has a limit of how many programs they can act with at any given time. And that's the only limit you need, in addition to being the only limit you could possibly enforce. -Frank |
||
|
|
|
||
|
Jan 28 2007, 01:41 AM
Post
#38
|
|||
|
Hoppelhäschen 5000 Group: Members Posts: 5,807 Joined: 3-January 04 Member No.: 5,951 |
The RAW limits active connections to Systemx2... |
||
|
|
|
||
|
Jan 28 2007, 01:45 AM
Post
#39
|
|||||
|
Runner Group: Members Posts: 2,526 Joined: 9-April 06 From: McGuire AFB, NJ Member No.: 8,445 |
I would wave that for stationary servers, IE anything that's big enough to have to have a server rack, or else you wouldnt be able to have matrix websites worth a damn. For commlinks and other moble devices, yeah, that rule stands, IMO. |
||||
|
|
|
||||
|
Jan 28 2007, 01:56 AM
Post
#40
|
|||
|
Running Target Group: Members Posts: 1,498 Joined: 4-August 05 From: ADL Member No.: 7,534 |
Then what is the difference between having a track programm run for you by 1000 nodes you are connected with (which seems not OK in your interpretation) and 1000 nodes connecting to you for database access (which seems to be OK in your interpretation)? Both are just interactions. I dont really get the difference. Especially if you say that you give all these 1000 nodes the same orders. Thats just like giving orders to one node and forwarding it to all the others. Maybe I am not really getting your interpretation. |
||
|
|
|
||
|
Jan 28 2007, 02:27 AM
Post
#41
|
|||
|
Prime Runner Group: Banned Posts: 3,732 Joined: 1-September 05 From: Prague, Czech Republic Member No.: 7,665 |
The point is that it doesn't matter how many nodes are running the Track program. The Track is "being run", and if it's one of the programs that you are currently directing, you get the benefit of perhaps knowing where another datatrail is originating from in realspace. If you allow it to make a difference where and how that Track is running, then yes, noone can hide anything and the entire Track action is just a formality. People will Agent Smith their Track actions onto a thousand nodes and instantly pinpoint any datatrail. But if it doesn't matter, then it... doesn't matter. --- In short, the matrix is like a bar. There's lots of boozes (programs) available, and you can pour them into your glass at any time. But you have a limited number of hands (system limits on program running), so the fact that all the different booze is all available doesn't actually fill up your glass any faster. Now if you had a second bartender, you could actually hold another set of taps and fill up your glass twice as fast. And whether that was a good idea would depend upon how skillful the other bartender was. --- Example: Jack-O is a hacker running a 5/5/5/5 modified commlink filled with illegal modifications that allow him to spoof his datatrail and mimic other commlinks and hop onto other peoples' networks. He's currently standing in a Taco Temple, which means that he is completely surrounded by nodes, PANs, and computational resources. There is nothing stopping him from putting copies of his programs on any or all of the nodes in this establishment, and indeed he is even now Editting the afternoon statement "Good Afternoon and Welcome to Taco Temple. Are you prepared to sacrifice your taste buds on the Altar of Flavor?" into something obscene so that visitors to this establishment will be confronted with Elf porn twenty minutes from now. However, there is a System limit on how many programs he can tap into and use simultaneously. If he exceeds his System of 5, his Response will decrease. So while he can load Agents galore into all the nodes around him (and he will), he probably isn't going to be personally benefitting from more than 5 programs each IP. Jack-O is currently running: Stealth Analyze Edit Scan Spoof So he can't actually have the Agents on the various other nodes do anything without dropping some of his programs or experiencing slow down. -Frank |
||
|
|
|
||
|
Jan 28 2007, 02:39 AM
Post
#42
|
|
|
Runner Group: Members Posts: 2,526 Joined: 9-April 06 From: McGuire AFB, NJ Member No.: 8,445 |
Frank...
I can see and understand bits and peices of that entire thing but I cant wrap my head around it all, the last section especially. Agents, when ran outside of the Hacker's personna, run off of pre-set instructions. IE. I could tell Agent 1 to wait 20 minutes then change File 1 (the Welcome sign) out with File 2 (Elf Pr0n) which was loaded with the agent. Then abbandon the node to come back to the commlink. The hacker could hack in, drop off the agent, then go on his merry way and the agent will perform the desired tasks, based off if it's stats pass the required tests. |
|
|
|
|
Jan 28 2007, 02:48 AM
Post
#43
|
|||||
|
Moving Target Group: Members Posts: 745 Joined: 2-January 07 From: Los Angeles, CA Member No.: 10,510 |
I think I understood everything quoted above, and agree with it, and I don't get any of the other stuff starting with the bar analogy. But if I understand you properly, you're pretty much saying that it's the TASK of running the trace system that's important to know at the "player level" and the specific method of how that task is achieved doesn't matter. We could say the Trace is being run at such a high rating because my hacker has an awesome program, or we could say its because he has thousands of little virtual-drones searching for you, or we could say it's because he's subscribed to some kind of pre-paid wireless P.I. services, or we could say he has a wireless connection to God that is telling him where you are. The point is that there are any number of ways to justify how a Trace is run because the act of "running a Trace action" is just an abstract way of describing other stuff going on in the setting. Your character wouldn't be running a Trace action, but would have other terms and processes to describe what he is doing. I can't tell if this is actually what you're describing, or just bias from my own interpretation of how agents should work because I've been making this case for Firewalls on other threads lately. Are we on the same page or reading two completely different texts? |
||||
|
|
|
||||
|
Jan 28 2007, 05:11 PM
Post
#44
|
|||
|
Prime Runner Group: Banned Posts: 3,732 Joined: 1-September 05 From: Prague, Czech Republic Member No.: 7,665 |
Exactly. And it is that task that is limited in what you can juggle simultaneously. -Frank |
||
|
|
|
||
|
Jan 28 2007, 05:50 PM
Post
#45
|
|
|
Running Target Group: Members Posts: 1,498 Joined: 4-August 05 From: ADL Member No.: 7,534 |
But agian, why can you handle (direct, benefit, whatever) only 6 connections to 6 trace actions but can handle (direct,benefit) 1000 Database connections at a time (which is a necessity for any working computer world)?
What hinders you from giving 1000 agents the same order? Certainly this can not be restricted in any way. The point of agents is that you give them very simple orders and they do very complex task, giving you back very simple results. Its not like your node has to do anything to keep the agent going if it is running on another node. You seem to avoid that point. |
|
|
|
|
Jan 28 2007, 06:13 PM
Post
#46
|
|||
|
Moving Target Group: Members Posts: 745 Joined: 2-January 07 From: Los Angeles, CA Member No.: 10,510 |
Because the rules are mostly concerned with what a character can do, not how he does it. It's like rolling Unarmed Combat. Do we really need to know the exact vector and force behind your uppercut, or whether or not there was an uppcut at all? Before or after the dice are rolled, depending on play style, the GM or player will describe the event how they wish. What hinders a player from saying "I kick him in the gut 10 times" vs "I kick him in the gut"? As long as its one roll to hit, and one roll to damage, it works within the context of the rules, right? If players and GMs want to be more lenient in what they allow for description, the rules don't have a problem with that. |
||
|
|
|
||
|
Jan 28 2007, 06:22 PM
Post
#47
|
|
|
Hoppelhäschen 5000 Group: Members Posts: 5,807 Joined: 3-January 04 Member No.: 5,951 |
Only that kind of abstraction layer hasn't anything to do with SR4 matrix rules.
It's not a houserule, is't pretty much a complete rewrite... in which case, everything goes. |
|
|
|
|
Jan 28 2007, 06:25 PM
Post
#48
|
|||
|
Running Target Group: Members Posts: 1,498 Joined: 4-August 05 From: ADL Member No.: 7,534 |
Granularity. You have to decide on which level ob abstraction you are on. Once you have a system where you model single kicks, you can not prevent somebody from doing 10 kicks in a row. Once you model agents, you can not prevent somebody from ordering arround 100 angents. The only thing you can do it model the effectiveness of 100 agents vs 1 agent. And that is the path I am using in my house rules. |
||
|
|
|
||
|
Jan 28 2007, 07:11 PM
Post
#49
|
|||||
|
Prime Runner Group: Banned Posts: 3,732 Joined: 1-September 05 From: Prague, Czech Republic Member No.: 7,665 |
Of course you can. The Agent is an abstraction. And what i is, is a set of tasks that are being conducted under the direction of your commlink. Whether it's a network of 1000 minor processes running on 1000 nodes or a single monolithic computer program running on your own commlink does not matter. The game does not have rules for processor speeds or memory, we literally have no idea how many parallels are needed to run each individual process. If you introduce such rules (for example, the "one agent can run per device in yur PAN" that you seem to take as given), then yes, the game can immediately and easily be broken just by stacking more devices on your back. But there's no reason to add that rule. The game makes no guarantees or assumptions about where any particular program "is", and as long as you continue to handwave that point, things work fine.
Because the six trace actions are each running through 1000 database connections. Or something. Honestly, who cares? -Frank |
||||
|
|
|
||||
|
Jan 28 2007, 07:14 PM
Post
#50
|
|||
|
Running Target Group: Members Posts: 1,498 Joined: 4-August 05 From: ADL Member No.: 7,534 |
That is excatly what an Agent is not doing. It is not under the direction of your comlink if it is running on another node. It is, by design desicion, independent. It even runs if you shut down your commlink, go away, and come back later to gather the results of teh agents task. Thats the whole point of agents. |
||
|
|
|
||
|
|
Lo-Fi Version | Time is now: 20th January 2025 - 04:48 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.