 Hacker Theory, Why this is possible. |
  |
 Jan 26 2007, 11:57 PM
Post
#17
|
|
|
Hoppelhäschen 5000  Group: Members Posts: 5,807 Joined: 3-January 04 Member No.: 5,951  |
Sure, you can do it - it's just a transparence problem, and some of it is already implemented.
And there you are, exchanging storage consumption for network traffic... when it makes sense: In a world where storage is cheap and near-infinite, it still will be a healthy balance. |
 |
 
|
|
Jan 27 2007, 12:04 AM
Post
#18
|
|||||
|
Prime Runner Group: Banned Posts: 3,732 Joined: 1-September 05 From: Prague, Czech Republic Member No.: 7,665 |
Well... that's not actually addressed in the core book. Your hardware's response limits:
But it doesn't ever say that the programs you are running are specifically being cranked through on the chips inside your Commlink. Indeed, some of the programs (Track and Browe for example) are clearly not run in any meaningful way on your Commlink. The rest could just as plausibly be running on your jacket or belt buckle and directed by your commlink as running on your commlink itself. Indeed, whether your commlink is actually running your programs or simply subscribing to them is not an addressed point in the SR4 book. And of course, it generally doesn't matter. Of course, if your commlink is actually running all the programs you are using, then the Agent Smith attack is broken, while as if it is simply directing all the programs located on other servers and devices then it isn't - so you know which way I tend to read it. --- From a character standpoint, the primary difference is that with an entirely decentralized system, hackers are going to drag in wearable devices to provide extra computing power when they infiltrate sites with a closed wireless connection (thanks to WiFi suppressing Paint this is easy and doesn't require running copper mesh through exterior walls). Personally, I like the idea of hackers filling up on pull-tab buritos and 60-day air fresheners before they head out - those little devices run a constant olfactory scan and emit a small puff of cyclodextrin when "bad smells" come into its range (which means that you can wipe their firmware with a tag eraser and then load them up with attack subroutines that your commlink can call up to be released at matrix icons you don't like...) -Frank |
||||
|
|
|
||||
|
Jan 27 2007, 01:13 AM
Post
#19
|
|||||
|
Hoppelhäschen 5000 Group: Members Posts: 5,807 Joined: 3-January 04 Member No.: 5,951 |
What you mean by 'run on'? Where it is processed? Or where it is stored?
A tag 'eraser' deep fries a RFID tag... if you use it on regular electronics, everything from nothing to smoke signals can happen. |
||||
|
|
|
||||
|
Jan 27 2007, 01:22 AM
Post
#20
|
|||||
|
Moving Target Group: Members Posts: 745 Joined: 2-January 07 From: Los Angeles, CA Member No.: 10,510 |
On the Matrix. On the PAN. On your P2P file-sharing program. Or on some other ubiquitous data-sharing anomoly. It's not like the information is stored or processed in a room or on a device somewhere. That's just crazy-talk. |
||||
|
|
|
||||
|
Jan 27 2007, 04:18 AM
Post
#21
|
|||
|
Moving Target Group: Members Posts: 718 Joined: 10-September 05 From: Montevideo, in the elusive shadows of Latin America Member No.: 7,727 |
...It's crazy. It's just a little bit obsessive too. It doesn't generally make sense; Welcome to Dumpshock Forums!!!! :) :spin: Cheers, Max |
||
|
|
|
||
|
Jan 27 2007, 06:18 AM
Post
#22
|
|||
|
Great Dragon Group: Members Posts: 5,537 Joined: 27-August 06 From: Albuquerque NM Member No.: 9,234 |
Really? So the citrix server farm that runs Cerner on the hundreds of winterms we have. doesn't really exist? |
||
|
|
|
||
|
Jan 27 2007, 07:02 AM
Post
#23
|
|||||
|
Prime Runner Group: Banned Posts: 3,732 Joined: 1-September 05 From: Prague, Czech Republic Member No.: 7,665 |
Right. See while you may think that you have a single set of massive computational resources that can be tapped into and utilized by any of the workers in your company so long as you have assured bandwidth sufficient to send the commands - that's just futurist propaganda. Actually you have a similarly sized unit of computational power allocated into each and every terminal in your company, and they just tell you that you are simply manipulating a bulky server farm in order to keep you from stealing the immensely powerful windows boxes sitting at each and every desk. /sarcasm Seriously, distributing processing power is amazingly powerful, and the only practical way to even attempt many of the high-end cryptographic schemes. So long as bandwidth is not a limiting factor, the thin terminal concept is immensely useful and prctical. In the case of cryptography, for example, the inputs and outputs use up a very small amount of bandwidth compared to the number of calculations that need to be done. Thus, they can be distributed even if the bandwidth available is very small. The Bandwidth positted in Shadowrun is not small. It's very very fast, to the point where the bottle neck of operating programs is your processor speed rather than your wireless signal's bandwidth. That's... a lot of bandwidth. Under those circumstances, distributing processor tasks is a no brainer. So much so that the total available processor power is virtually unlimited. That is, you can perform X calculations in a second with an object the size of a quarter without seriously impacting the available WiFi space, and then you can have a ten dollar roll of quarters that performs 40X calculations, and you can one in each hand, and one in each pocket, and so on and so forth... The only meaningful limitation of Shadowrun computations, therefore, is how much you can orchestrate together into a whole - which is in turn a limitation of your System. That is, you can jolly well keep stacking micro devices one on top of another until the "no appreciable weight" eventually adds up and breaks the camel's back - but there's still a hard limit of how much you can get to all play nicely together at any given time. That's the limitation on a Shadowrun hacker. Basically the only limitation on a Shadowrun Hacker, and it's the reason that they are constantly seeking out illegally modified software and hardware to be able to Shanghai more and more programs and devices into simultaneously working on their behalf. -Frank |
||||
|
|
|
||||
|
Jan 27 2007, 12:22 PM
Post
#24
|
|
|
Running Target Group: Members Posts: 1,498 Joined: 4-August 05 From: ADL Member No.: 7,534 |
Well, I see that this concept would work in reality. But unfortunetely it would not work as a game concept as it is too complicated and would require too much micro management.
|
|
|
|
|
Jan 27 2007, 01:34 PM
Post
#25
|
|||||||||||
|
Hoppelhäschen 5000 Group: Members Posts: 5,807 Joined: 3-January 04 Member No.: 5,951 |
Winterms? I would call that a fatclient. ;) It would be interesting to know if it is done for reasons other than licencing issues...
And it will never be as swift as a normal desktop box. Such decisions are made on a cost basis, and with the low prices for computers today, it's usually cheaper to just use network and servers for storing the home directories of users. BTW - the star structure you are referring to isn't really a good example for the mesh structure you are proposing...
You mean like all those CAD boxes sporting certified GPUs that cost more than a whole consumer PC?
That's not entirely true - it's near-instantaneous. Transfers normally complete at the end of the combat turn they are started (unless the GM determines otherwise for exceptionally large files), so it may take about three seconds. Essentially, memory and bandwidth are equally distributed in SR4. Going for solutions focussing on one aspect would be pretty inefficient.
Oh, there are simple implementations... just add everything together. The problem lies elsewhere: The very concept results in a spiral of escalation. |
||||||||||
|
|
|
||||||||||
|
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 |
||
|
|
|
||
|
|
Lo-Fi Version | Time is now: 30th July 2026 - 11:46 PM |
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.