There is no nicer way to put it—that's complete bullshit. If you absolutely insist on holding to that ridiculous argument, then we can move the timeline up to 1984—still a good five years prior to SR1.

~J

Two, actually. And the issue is that you can now start with a higher-grade commlink than would be possible off-the-shelf, *and* have enough extra BP for either a specialization or low-end contact, which are things that are much harder to get once game has started.

You still haven't demonstrated that this is a problem. Yes, it's possible to get a top-of-the-line commlink for better prices than off-the-shelf, but why is that an issue? Is there some reason that hackers shouldn't be able to upgrade their signal and response at chargen?

The problem isn't deckers, the problem is everyone else. Because it's cheap, suddenly you have everyone running around with high-end commlinks. Also, there's the issue of running multiple commlinks; the cost difference might be small on just one, but it gets huge when you're buying ten. And trust me, running 10 commlinks with agents gets ugly very quickly.

Well, first of all, I don't allow anyone without Hardware skill or a contact with it to buy upgrades at chargen.

Second, I don't allow anyone to give commands through multiple commlinks simultaneously. That's patently ridiculous. Maybe you have multiple comms, but you need to spend time popping between them, probably switching in and out of VR (and thus losing extra actions) in order to tell them what to do.

Finally, I don't let upgrades take a device past 1.5 times its original rating. I'm not even sure that's a house rule. Sure, some of the pregen characters would imply that that's a mistake, but nearly all of them violate the "no more than 200 points in attributes" rule, so I'm not impressed by that.

That seems fair. As is requiring a Build/Repair Test to install an upgrade.

I think it would be a Simple Action to switch between open VR "windows" if you are going through one commlink to the other.

House rule. I'd be inclined to have more difficult Build/Repair Tests per difference. The rules could do with less multiply by 1.5 rules, not more.

Give an example of a sample character in the 1.3 PDF that still violates the 200 BP maximum? AFAIK they've now cleaned up the sample characters to the point that there is only a couple errors in the entire set. Or are you talking 1st print of the book? Actually, give an example in the printed book as well. I don't recall any of them violating that. Remember that Edge and Magic do not count towards that max.

It is.

If you're going from one commlink to another, that's fine, but if you have agents running on another node, like another commlink, to give them orders through the first commlink requires them to be subcribed to it. If they're not on the subscriber list, you have to physically switch to another network which they are subscribed to to give them orders.

That's the big problem with the "agent storm" that gets brought up around here. To give an agent orders through your commlink, it must be subscribed to that commlink, barring spoofing. An agent on a secondary commlink, even if that secondary commlink is subscribed to the original, isn't listening to the original commlink's orders. Running an agent down-network from your main commlink isn't any different from running an agent on another node, because commlinks are nodes.

Well, I dont' have any aesthetic issues with the number of x1.5 rules, and it's easier than trying to work out a new threshold on higher levels.

Yes, I do see that it's not in the main rules now; I thought that the 1.5 hard limit applied to attributes as well as skills, but in fact it's only for skills. I'm still sticking with it, not for balance reasons, but just because I think, much like IRL, there's only so much upgrading you can do to a computer before you meet the limits of its original construction and have to, say, switch out for a new machine that can handle SATA instead of IDE if you want better performance.

Fair enough. They don't separate the Edge/Magic expenditures in the BP totals on the profiles, so I didn't think to deduct those when looking at the pregens.

In addition to that being a house rule, there's a bunch of contacts that would reasonably have that skill. For example, someone might argue that his tech-oriented Fixer should have that skill. 2 points isn't really much of a limitation, especially when you'd be buying that contact anyway.

You don't have to. You can pre-issue commands to agents, since they have autonomous abilities. With a Pilot rating of 4, they can understand reasonably complex commands, like "attack anyone hostile to us", without having to worry about them attacking you or missing something.

Yes, well, so what? Someone who spends two points on a fixer contact has used up the points they would have saved on a cheaper commlink, and I still don't see a reason to care that people can buy them cheaper in the first place. I only require the skill or a cnotact because its uits me, not for balance reasons.

Attack anyone hostile to us? Is the agent going to initiate hacking actions on the commlinks of all security forces in the building? Is it going to attack all of the IC in the local Matrix when the Shadowrunners get inside, setting off massive alarms in the process? I'd be very leery of letting an autonomous agent run with such nonspecific instructions.

The problem is that they would have likely bought that Fixer contact anyway, so they get away with a freebie. I see what you're trying to do, but I don't see that it's very useful.

Well, I could have worded that better. Still, the point remains: you don't have to give individual commands to each and every agent. You can have all the agents follow one that you're in contact with, and order that one to attack whatever you feel like. If you're running with multiple high-end cheap commlinks, you can have a very disgusting Matrix Attack Pack in short order, with very little cost to you.

If your turn some of Joe Average's commlinks into zombies, you have an even nastier amount of power - at no cost.

The 'problem' of distributed systems is not to be solved by trying to limit character generation.

I'm not necessarily trying to do anything except make the game run in a way that I like. I don't think that buying slightly cheaper commlinks at chargen is a balance problem, I just have an aesthetic objection to doing so without justification.

And my point is that such a Pack is extremely likely to draw nasty trouble down on your head, and fast. If numbers were always advantageous, every runner team would have thirty people, let alone agents and IC. There's a reason shadowrunner teams are small, and for similar reasons it's a bad idea to try to mimic a corp net with commlinks strapped to your person.

Unlike people, you can fit an unlimited number of Agents into a very small node. By the rules, one commlink is as easy to detect as a dozen, so there's not much of a disadvantage to the multiple commlink trick, other than cost. What's more, unlike people who can't physically fit into one area, there's no limit to how many Agents that can work together on a Teamwork test. So, you can't have thirty people helping with picking a lock, but you can have thirty Agents help on a hacking test.

You can't. Running too many Agents on a Node will result in drop of Response, thus decreasing the effectiveness of the agents.

I do remember someone posting that the only effect from lowered response is that you get -1 to response TESTS for running more than (Response) amount of programs.

The reponse issue really bothers me because there is no clear explanation. Running 10 programs gives you a -2 to response on a response 5 commlink, and then what?

1. Do you count from the NEW response attribute and realize that you now run 10 programs at a Response 3 commlink? Oh, nice lower the response instantly to 2, rinse and repeat until you get response 0.

2. Do you count from the ORIGINAL response value of 5 so that you would get -3 when running 15 programs?

IIRC, you get a -1 to Response for every number of Agents equal to your System rating. I'll check my book and get back to you when I get home.

Sorry about the mis-attribution. Fixed.

I might be missing something, but I don't see anything indicating that a Node has a Response rating. Also, I thought that the program/Response limits only applied to your own commlink, not the nodes you were accessing. The thought here was to use multiple commlinks, running multiple agents, in one massive hacking attack on a single Node... have I missed something?

That's not what Rotbart is saying, if I'm reading him correctly. If I have a lot of commlinks with programs running, the host shouldn't crash under the weight of my programs. If I can make the host run a lot of programs, that's different; but I can't see how the Host's load rating can possibly affect the load rating on my commlinks, or vice-versa.

And that still leaves open the multi-commlink mass hacking attack. You get a bunch of agents, on a bunch of commlinks, playing "simon says" with either your icon or an agent you're in direct control over. You can then call for a massive Teamwork test. What's more, you don't even have to worry about an agent making a critical fumble, since that only adds to the threshold-- and thresholds are never applied in opposed tests, which comprise an awful lot of Matrix actions.

Yeah, but in the meanwhile, you've pretty much guaranteed to get Root access. As an admin, you can suppress the IC, kill the alert, and even identify the security deckers as the intruders (and your agents as IC). You just need to make another test, with the same teamwork bonuses as before.

You know, that brings me to another question. I've looked and looked, but I can't find rules for "pulling the plug" on a Martix VR connection. I understand that there should be some penalties, such as dumpshock; but as far as I can tell, you're not allowed to simply shut off your commlink-- you *have* to jack out every time.

If I'm running multiple agents on multiple commlinks, why can't I just turn off a commlink to shut off all the agents running on it?

Well, they differentiate between logging off and jacking out. I've just assumed that jacking out (which produces dumpshock) IS "pulling the plug".

First of all, the node has already gone on alert, and the other hackers, who also have admin access, have any powers that you have. They, however, likely have actions remaining before your next IP.

Second, you don't have any teamwork bonuses, because you can't give commands to your agents if they're not subscribed to your commlink. You might be able to relay commands through another agent, but that takes time, which translates into lost actions.

Third, again, you can't give an agent commands if it's not subscribed to your commlink, without spoofing, so you don't get to shut off the IC that's attacking you.

Edit: that's not to mention that no corp nodes, in my game, allow even an admin to shut off alert status without physical, RFID-based authentication on a secondary node. You could hack into that node from the outer area, but you can't just use an Edit command immediately. YMMV, though; not all GMs will run corp security as tight as I do.

Go for it. You drop out of VR, losing your bonus actions, and turn off one commlink with each simple action, while your icon gets toasted in the corp's Matrix, your datatrail gets followed, and by the time you're done you've been booted, your icon's been destroyed and corp security is on their way to your house.

As a runner, the numbers game is not in your favor. What you're proposing is effectively launching a massive assault on a corp node with a huge army, and then trying to get whatever you need done before their own army of agents and meat bodies finds you. It's no more likely to work than if a rigger bought fifty drones and then tried to blast a hole in a corp skyrise to extract whatever it is he wants. In a straight up fight, no matter what you've brought with you, the numbers are not on your side.

I may be reading page 216 incorrectly, but as far as I can tell, security deckers get security access by default. Things may be different in your games, of course, but I thought the whole point of admin access was to restrict the number of people who have total system access. Giving it to your entire security team-- which, as you correctly pointed out, might number in the hundreds-- seems to be a rather large risk for a megacorp to take.

You don't have to give them commands, though. Agents are autonomous enough to act on their own, with only general direction. If you tell them to do "Simon Says" with an agent under your direct control, they won't wait for further orders to come down the commlink.

And again, Agents are autonomous. You don't have to issue them commands, if you've issued them ones in advance.

Like I said, there should be consequences; and other than the icon damage, I think that's all reasonable. However, what I don't get is this: when you're under attack by Black IC, why can't you just pull the plug, yank the battery from your commlink, or something else along those lines? Why can't someone else do it for you? And if they do shut off the commlink, why wouldn't that terminate the matrix connection, preventing further damage to your icon?

In SR3, there was a rule that if you pulled the plug, your datatrail stayed open, leaving you wide open for a Trace. I accept that this is fair in SR4. What I don't get is the fact that apparently you *can't* pull the plug.

The general run of security hackers is going to have security access, but someone has admin access, and when you blew their doors open they noticed you.

You do have to give them new commands if you have a new thing you want them to do. And, as I said, the Simon Says protocol takes time to relay instructions.

Also, a further complication is that your subscribed agent has, itself, a subscriber limit. To get really crazy numbers of agents, they're going to be further tiers down, taking further actions to pass the info for the new combined action, and if you want the teamwork bonuses you can't begin until everyone's ready.

How exactly have you given advance commands to the IC that was running on the node before you hacked in?

That's what jacking out is, as mentioned. You can indeed just pop the connection, accept the dumpshock, and move on with your life. Generally you wouldn't if you thought you could defeat the IC, or if you badly need to cover your datatrail, but it's an option.

If you order Agents to log onto another Node, they will leave your Node and run on the Node you ordered them to.

Ya, it is just plain old getting dumped. They didn't write that explaination explicitly into the core book though this time, that anytime the device your persona is in or any the communication link between that device and you stops working you are dumped. You just have to imply it from some of the examples it gives.

It is tough for someone to physically pull their own plug though, unless they have removed the safety override they can't move while in VR. Someone else though could shutdown their commlink or physically break whatever connection there is to break.

The Command for the soft turn-off of the commlink by the decker would seem like an obvious thing for the Black IC to block as part of keeping the decker jacked in.

its not 100% impossible, but physical actions while under RAS was in the old rules at +8 difficulty...

but under the current rules you can go from VR to AR and back again with a free action. and i cant say if find anything about how black IC affects that (alltho i guess it would be about the same as jacking out).

hell, when i think about it, black IC/out/hammer and AR is nowhere coverd. its as if you go automaticaly immue to those kinds of attacks even tho your still using simsense to display all the elements. im guessing you can somehow compare it to the old tortoise mode...

edit:

hmm, now that i think about it, its much safer to get high level wired reflexes and go AR all the time. only time thats a problem is when your probing a node as with VR you get one test pr hour while AR gives you one test pr day. in cybercombat and similar your just as fast as the IC if you have the best ware, and your safe from black programs. ok, so VR gives you +2 dice, but i hardly this its a good trade given the risks...

Right at the end of the Black Hammer description.

The Black IC must block the switch of VR, or Black IC just become quite toothless and the part about forcing a jack out doesn't make much sense. Either they just forgot to mention that the switching is blocked, or switching from VR to AR is assumed to log you out of the node you are in which means the blocking of the switching is implicit in the blocking of the logout.

EDIT For some reason it is bugging me that there is something somewhere about performing actions while in VR. It even had a conflicting die penalty, but I can't find it right now. So they might have put in some sort of wierd duality there that is a duplicate of the SR3 situation (that had slipped my mind, thanks).

I was just try to give two possible situations to resolve the get out of jail free of just by a Free Action switching to AR, which is indeed silly. The first is an omission, the second also requires an ommision. That VR to AR logs you out of the node.

Switching from VR to AR logging you out of the node plain silly? It really doesn't say one way or another. It is unclear as well if switching VR to AR leaves you in the same "window". There is another option that you actually keep the VR window, with your VR persona in it. The Black IC being able to juice you even though the inactive VR windown your "active" window is AR. But I don't notice in the RAW anything about being able to have a mix of AR and VR windows open either, and it is pretty silly to think about getting VR fry jobs while in AR. System Failure sort of hints at this at the end, but that is more Blackout in ARish prototype system.

That is the first one. There was a second one I thought.

Nope, an agent is a PROGRAM and not a DEVICE therefore they are not using the personas subscription list. Any drone WOULD do it though since a drone is a DEVICE.

See the following from page 223

Issuing Commands
While online, you can issue commands to an agent (p. 227),
drone (p. 238), sprite (p. 234), or other device under your con-
trol with a Simple Action. Note that you can issue the same com-
mand to multiple agents, drones, or sprites at once with the same
action; dif erent commands, however, require separate actions.
Note that agents and drones will only take orders from
their controlling persona, unless another persona spoofs an or-
der (see Spoof Command, p. 224). If the controlling character
chooses, he can instruct the agent or drone to receive orders
from other specii ed personas.

You don't have to have them subscribed to one another, though. You can simply say: "Coordinate your actions with icons X, Y, and Z"'; as long as the actions are matrix-oriented, the agents should be able to reasonably follow that directive.

A "Simon Says" protocol basically means: "Do what you see this guy do". You order an agent under your direct control to do something, and all the rest should follow suit. You don't need to send individual commands; agents are autonomous enough to follow a simple directive like that.

What I'm confused on is this: Black IC has the ability to prevent a decker from jacking out *and* logging off. However, if your buddy with the 2070 equivalent of hitcher jacks (or a biomonitor) sees that you're under attack, he should be able to pull your plug. But that's also jacking out, which Black IC won't allow.

I see that dumpshock is still an issue, and while the rules aren't as specific on this as they were in 3rd, I can easily see that your datatrail would be left undefended. But I can't see your icon taking any further damage if your commlink is shut off for you, and I can't see why you would be stuck in cybercombat if someone else pulled your plug.

Okay, let me make sure I have this straight. If I upload a bunch of agents to run autonomously in the matrix, they count against the program limits of the Node they're on. However, if I keep them on my commlinks, they only count against my rating.

That still means that if I'm running multiple commlinks with multiple agents, none of which are overloaded with programs, I can have a disgusting number of active agents at any given time.

Yup, you got it right. Here’s the even more fun spamtastic action you could take. Have an agent load itself with the following program loadout:

Agent
Attack
Armour
Analyze

Upload agent into another Node with the following orders.

1.Upload copy of yourself until target node is decreased to response=agent rating
2.Analyze system for hostiles
3.Attack hostiles
4.Defend other agents from attack.

Another available action is to simply rinse and repeat option 1 until system goes down into Response 0 or another desirable rating while you hack it – in that case substitute analyze for stealth.

Yes, it is evil, but agents can carry programs, and agents ARE programs. Still, I do think tha corporations probably have tactics against it, like setting a maximum amount of programs that can be loaded at one time, thus setting a permanent response rating.