Help - Search - Members - Calendar
Full Version: Response and IC, I am confused
Dumpshock Forums > Discussion > Shadowrun
Pages: 1, 2, 3
Serbitar
What response rating does a mainframe have? Only 6?
Why does a credstick have a device rating of 6? This means that it has System 6, Response 6, Firewall 6, Singal 6?
Why does anybody buy comlinks if your credstick is so powerfull? Just modify your credstick.
How much IC can you load onto simple nodes that are not meant to work as "real" mainframes?
How much IC can you load on "real" mainframes?
Why shouldn't there be a huge difference between both kinds of system?
How does IC or an Agent affect a nodes response (as indicated in p. 228, in the paragraph "Intrusion Countermeasures")? Does every of the 4 attributes count as a program? Does only the IC count as a program? If yes, why? Do the programmes, the IC is running, count toward the response limit of the node?
How does a lowered response affect the standard duties? Does this even matter?


The fact, that almost everything now has a device rating, which must fit into the 1-6 levels, smartgun node in your weapon and supermainframe alike, imposes some huge problems.
Does anybody know a solution?
Is there any way to bring comlinks with system ratings of 6, hightech devices with devicerating 6 that are not originally meant to work as superpower processors, and real mainframes that can only have a rating of 6 together?

Do I just have to accept the ratings and roll on random tables for node responses because reasoning is futile?

Help me, for I am confused.
blakkie
Matrix nodes are not limited to rating 6. MATRIX ATTRIBUTES, page 212. More detail coming in Unwired of course. In the meantime just cluster up those cutting edge bad boys to groups, and handwave things like the main Ares mainframe.
Serbitar
Hm, I really hope that the writers of unwired try to model something which is at least consistent on first sight . . .

But the Problem at hand is: How many agents can a comlink, a credstick and your IngramSmartguns node run?
and
Is there any reason not to load every IC available at the first net hit of the node and track the hell out of everything?
Just having the IC loaded everytime does not inhibit anything, so tehre is no reason not to do load the full set at the slightest hint of danger.

I have no idea.

Unwired should have definately been the first book to be released after BBB. Not doing this is a very very big mistake.
blakkie
Here are a few someones posting somethings about hosts.

http://forums.dumpshock.com/index.php?showtopic=12458

As for IC on everything, there isn't really much need for it on most of your PAN items because they seem to get treated more like end devices. It is your commlink's IC that is going to come into play.

Certified Credsticks i wouldn't even bother give a real Signal too. They aren't what they used to be, the main retail cash device.
Glayvin34
I'm kinda glad that I got into Shadow Run for the fourth edition, sounds like third edition knowledge mucks up the understandings.

Anywho, I've read and reread the Wireless chapter many times, and there are some oddities there, but it's consistent throughout, for the most part.

You can run IC from your Smartgun or Datajack or whatever, but you need a response upgrade for that to be effective. Datajacks and Smartguns have ratings around 3, for every 3 programs (rounding down) you loose 1 response. Also, whatever IC you're running can't act higher than the modified response value. But the short answer is that yes, you can load IC on every device and bit of cyberware that you own. I guess you could load IC on a credstick, I don't think that credsticks are self-powered.

All devices have a System rating, and all you need beyond that to run a program is a response rating. Any character with some Hardware skill can make a response upgrade in a few days at half cost (see page 240), so that you can load a shitload of IC into your PAN.

The Hacker that I play spends his downtime putting together response 4 commlink upgrades for nuyen.gif 1,000. Then he copies over system 6 from his main comm, and, bam, he has an IC node that runs around 10 active Agents with rating 3. Set four or five of those in tandem without a wireless link between your comm and your PAN and no one's getting in. And if someone attacks you can have 40 agents run out and start to kick their ass.
Moon-Hawk
So then, if this is so cheap and easy, then everyone in your game does this, right?
James McMurray
My rigger /street sam has IC on everything he owns that can carry it and won't be overly affected adversely by it. With free program copying there's no reason not to.
Glayvin34
QUOTE (Moon-Hawk)
So then, if this is so cheap and easy, then everyone in your game does this, right?

Pretty much.
The GM figured it out, too, so if I set off the alarm in a node and don't get a good intiative, I get the crap kicked out of me by 30 IC agents.
Moon-Hawk
See, it's the "there's no reason not to" part that bothers me.
James McMurray
Why does that bother you?
Voran
Yeah it does get a little wonky when the old fairlight excal was a million bucks, and currently the best comlink can be carried by darn near everyone (at least every runner, hacker or not) and be pretty much hacker-grade forwhat, under 10 grand? smile.gif

Mainframes should at least have a couple points higher than rating 6. Or at least have a higher thresholds or be less negatively affected by running many programs at once than someone on a comlink.
James McMurray
Unwired will probably have commlinks with higher ratings and much higher costs.
Ophis
My take on it is to allow mainframes to break rules about howmany programs are running on them. I don't go for the IC running node idea myself I take the Node the IC is running in to be the one it is acting in. This prevents the why isn't everyone doing it problem.
Serbitar
QUOTE (Glayvin34)
QUOTE (Moon-Hawk @ May 3 2006, 05:57 PM)
So then, if this is so cheap and easy, then everyone in your game does this, right?

Pretty much.
The GM figured it out, too, so if I set off the alarm in a node and don't get a good intiative, I get the crap kicked out of me by 30 IC agents.

I would call that inconsistent and unbalanced.

Furthermore: Why does an IC only count as 1 programm and not as the 2 persona programms plus the programms it is running?

Furthermore: A mainframe sholud have at least 10 times the processing power of a mere comlink, if not 100 times. And certainly not just " a couple of points".
Say hello to response 50 mainframes with related initative.

I would call that unbalanced (but not inconsistent), too.

I am still confused.
Kanada Ten
I don't see why a mainframe needs more than a rating 9, it's really more of a hub and attached are nodes such as "off-line storage","coffee machine", "I/O Port". The mainframe is only a large PAN, it doesn't need to run lots of programs.
Glayvin34
QUOTE (Serbitar)

I would call that inconsistent and unbalanced.

Furthermore: Why does an IC only count as 1 programm and not as the 2 persona programms plus the programms it is running?

Furthermore: A mainframe sholud have at least 10 times the processing power of a mere comlink, if not 100 times. And certainly not just " a couple of points".
Say hello to response 50 mainframes with related initative.

I would call that unbalanced (but not inconsistent), too.

I am still confused.

I fundamentally agree with you on the Mainframe part, and if Mainframes had a rating of more like 20 then the whole system would be more balanced. That would make a Hacker with a upper limit of 6 only modestly capable, even with 30 rating 3 agents. Or you could just raise the firewall of Mainframes to 20, then not even a host of rating 6 Agents would do much damage if their threshold is periodically higher than their dice pool.

Apparently processing power and data storage is cheap in 2070. Nanometer scale circuits seem to be commonplace. So if you try to hack a credstick 4 or 5 Rating 6 Agents pop out and tell on you.

A better way to balance this would be to say that an Agent's Load counts against the max programs running. Attack IC needs at least Armor, Attack and Stealth Programs, so that would be 4 programs instead of one.
blakkie
QUOTE (Serbitar @ May 3 2006, 05:19 PM)
I am still confused.

I think it partly has to do with your assumption that a single matrix node must equate to a "mainframe". A "mainframe" could easily consist of multiple nodes at a software level, and very likely a cluster of hardware whose divisions don't nessasarily match that of the software divisions. That would really be a lot closer to real computer architecture.

If you look further down that link i found with Search there is a suggestion about the better systems being able to run some IC without putting a load on a node's resources. Will Unwired have that sort of stuff? *shrug* I don't know, but it seems like a workable solution.

QUOTE (Glayvin34)

The GM figured it out, too, so if I set off the alarm in a node and don't get a good intiative, I get the crap kicked out of me by 30 IC agents.


30 agents? WTF? Isn't that going to load the crap out of the node? They'll be tossing a die or two each. There are people that know that chapter's rules better, but having that many Agents load up all at once seems quite a stretch. Certainly a perversion.

Of course that is just another great reason to intepret the vague copy protection cracking rule as a one per cracking extended test copy instead of unlimited copying for a single cracking extended test.
The Jopp
QUOTE (blakkie)

30 agents? WTF? Isn't that going to load the crap out of the node? They'll be tossing a die or two each. There are people that know that chapter's rules better, but having that many Agents load up all at once seems quite a stretch. Certainly a perversion.

Rating 6 node with 30 agents:

6 Agents = Response 5
12 Agents = Response 4
18 Agents = Response 3
24 Agents= Response 2
30 Agents= Response 1

Ok, so the GM toss 30 agents against you. Well, first they have to find you. They need Analyze and will roll 2D6 against..say that you run 9 programs on a response 5 commlink so you have 8D6 in total. One of them need to find you in order to call on the others to attack you.

Their attack programs will be a rating 1 program and they attack with 2D6, you defend with 8D6- Each attack they do might do 1DV+ damage, you resist with around 8D6.

In order to defeat them just upload 6 agents of your own on the node to attack the other agents. Once you have uploaded 6 agents it will be 36 agents on a rating 6 node and it will reach Response 0.

Not a good tactic. He should stop with 12 agents rolling 8D6 each on attack tests with rating 4 programs.
Serbitar
I think even 5 IC that are loaded when the node gets 1 hit in detecting the intruder is overkill and ruins the whole game balance. But there is no logical reason not to do it.
Furthermore I still think an agent should count as "2 (System, Firewall)+number of programs on the agent" programs. This would even make IC count on response 1-6 devices reasonable.

AND I still think that a mainframe (realistically) should have response 50-100 if a comlink can have 6.
Even if a mainframe is running multiple nodes (why should it?) it has enough power reseves to do whatever it wants (and provide ultra high intiative for its IC).
Remember: A comlink has to be small. A mainframe can be the size of todays midi towers. It should be very easy to squeeze 10-100 times the processor power of an usb-stick sized comlink into such a mainframe at a cost that is much lower than the comlink. (Just compare todays PDAs, destkop pcs and mainframes)
Voran
I'm wondering at the math for agents/IC. After looking over the program section again for agents/IC, I'm wondering if I understand the math correctly.

An agent, if run in my persona, counts as 1 program. But while loaded in my persona, does nothing unless I give it a program to do. So lets say I give it Heal to do. That would give me a program count of at least 2, using both active at the same time. This could then be set to run heal on me every turn or something, freeing me to do other things. But this leaves me with 2 less 'slots' for programs if I want to avoid my own lag. Alternately I could dump the agent by itself in the node I'm at and load him with heal. He could still do the heal every turn of its turn, but now I have 2 freed up slots in my persona to run other stuff. He counts towards my subscription, however. The agent, now independent, can impact the node response, since its not running off my personal resources anymore.

Thing is, does the now free floating agent count as 1 prog or 2 (agent+heal) for purposes of impacting the node response rating? Likewise does an IC loaded with analyze, attack, blackhammer, blackout, and trace count as 1 or 6? (IC + 5 progs).

Yes in a way memory is functionally unlimited. But if they're putting in a mechanic of # of programs affects function, that still sorta implies that # of programs run within an agent/IC count towards total count too? Which would also lead nicely to less agents/IC being run at one time, rather than swarm tactics, unless you ran agents/IC with only 1 loaded prog each, or something.
blakkie
QUOTE (Serbitar @ May 4 2006, 05:24 AM)
I think even 5 IC that are loaded when the node gets 1 hit in detecting the intruder is overkill and ruins the whole game balance. But there is no logical reason not to do it.

Other than it is overkill? Thanks for that breakdown Jopp, but i was thinking more along the lines that it would take an IP per Agent brough up by the system that is not already actively running. So ever IP the system might throw another log on the fire, taking a full turn to get 3 up. Or is that 2 per pass? So you face this cascading dogpile if you don't get the hell out of Dodge fast, but it isn't a total dump on you right from the start.

QUOTE
Even if a mainframe is running multiple nodes (why should it?) it has enough power reseves to do whatever it wants (and provide ultra high intiative for its IC).
Remember: A comlink has to be small. A mainframe can be the size of todays midi towers. It should be very easy to squeeze 10-100 times the processor power of an usb-stick sized comlink into such a mainframe at a cost that is much lower than the comlink. (Just compare todays PDAs, destkop pcs and mainframes)

AND I still think that a mainframe (realistically) should have response 50-100 if a comlink can have 6.


The whole big cluster of the hardware that comprises the "mainframe" might have that collective power, but a single matrix -node- that you get to see on it from the inside out does not. A mainframe isn't really organized like a desktop. Besides, try to keep in mind that this is SR computing. Try too hard to rationalize it and your eyeballs will get sucked right out the back of their sockets. frown.gif

P.S. Balance? You keep using that word. I do not think that word means what you think it means.
Glayvin34
Hm. Now I'm confused about the rules. Does the IC need to be in the same node as you to initiate cybercombat or just subscribed to the node you're in? For almighty BALANCE, they would have to be in the same node as you, but I can't tell from the Cybercombat section in the book.

In this thread everyone seems to think that you can just subscribe 10 IC nodes to your main node for the aforementioned 30-40 Agents:
Skill wires rort.
mdynna
Although every "device" in the game has the same ratings, for sanity purposes in my games I have put them into 2 classes: device, and host (call them whatever you want).

The first class, "device" is mostly everything in the world: Commlinks, credsticks, vending machines, cars. They follow all of the rules listed, as standard.

A "host" (or mainframe) is what the whole Matrix "backbone" is made of, and what corps run for all of their stuff. Any Node classified as a "host" ignores the rule for Response Decrease. It still has all of the ratings, but has unlimited "bandwidth" or "multitasking" capabilities.

To take the Response Decrease rules and apply them blindly and universally to everything is just silly. Think about an MSP for a second, it must have a Response of.... 5 BILLION or so? I think those Response Decrease rules were written to prevent the "Hacker with an Agent army" problem.

I think a lot of people are still suffering from some SR3-rules "hangover." Under SR3 there were micro-management rules and number crunching for everything: Storage Memory, Active Memeory, I/O Speed, Party IC, Cascading IC, Trap IC, yadda yadda (did anyone get as far to use the Optional Bug rules?). This reduced the game to a crawl and made Decking an exercise in mathematics to the nth degree. Even then, players found "holes" all over the place and ways to exploit the game. Now, we have general rule "guidelines" and the mantra of SR has become whatever you think is reasonable. The rules, in many cases, are a starting point. Take them as such, and extend and build upon them as your situations arise.
GrinderTheTroll
Put aside trying to deistinguish between a "device" and a "mainframe" they are all nodes in SR4.

So as such, isn't the max device rating 6? System is capped at Response, so you could have Systemx2 programs before Response degrades by 1? So a Rating 6 node could employ 12 Agents before you'd see a performance issue? So once a 13th Agent enters, the nodes response drops 1 to 5 effective reducing any Agent's ratings of 6+ to 5. This also includes your agents, since they are "free" of you, they take on the Reponse of the node the occupy.

As long as my understanding holds true, standing in a Node rated 6, and it launched 20 Agents rated 6, they'd all suffer from Reponse degradation and become Rating 4.

Am I correct in this?
Moon-Hawk
The biggest reason, to me, that you have to distinguish between a device and a mainframe is that a device can only connect to system x 2 other devices at a time. An average home would have dozens, hundreds, possibly thousands of wireless devices in it that the telecom should be monitoring and managing. An office building mainframe should have hundreds of users logged on at one time. There's no way a rating 6 device can connect to that many people and run that many personas at once. Unless of course the mainframe is dozens of rating 6 commlinks wired together in parallel, in which case it could become a sort of virtual commlink that still only runs at rating 6, but can connect hundreds of users before it slows down, hey, let's call it a mainframe!
Serbitar
Mainfraimes are not nodes in SR4, because rating 6 nods have restrictions that no mainframe would have. Furthermore, as I mentioned before, a device 100 times bigger than a small comlink can easily have a rating of 50 and be cheaper than a comlink.

But still: Can anybody resolve the "why not load the maximum number of agents at first node hit" problem?
blakkie
@mdynna

Usually though a big iron system will set a choke point on system resources used by an individual login or virtual machine. A node isn't exactly that, but the reasons for limiting the resources remain largely applicable.

In some ways comlinks, in regard to the things that Voran mentioned, make a lot more sense under a mainframe type situation than with a handheld commlink. Because the commlink already has this slicing up of resources happening where just having multiple logins by Agents on a commlink appears to pull resources out of nowhere.

@Glayvin34 The rules kind of hint at on page 227, but don't come right out and say, that an Agent that is acting on it's own must be in the node it is performing its action in. Although it gets a little murkier with things like Track, i'd make the assumption there is the action is being performed in the node where the target icon is located. Not on the path leading back to the physical location, which would require the Agent to node hop back along the datatrail.

@Moon-Hawk

Being able to manage enormous numbers of I/O devices is certainly something that is a characteristic of a mainframe type system. Acting as the communication hub for perhaps thousands of user logins at one time.
GrinderTheTroll
QUOTE (Serbitar)
Mainfraimes are not nodes in SR4, because rating 6 nods have restrictions that no mainframe would have. Furthermore, as I mentioned before, a device 100 times bigger than a small comlink can easily have a rating of 50 and be cheaper than a comlink.

No rules exist in SR4 for anything other than a node. Anything aside from what's in SR4 will have to be House Ruled.

Yeah it sucks having to wait for source books to come into play, but there are lots of good ideas in this thread.
blakkie
QUOTE (Serbitar @ May 4 2006, 03:15 PM)
But still: Can anybody resolve the "why not load the maximum number of agents at first node hit" problem?

*sigh* First off, as pointed out about the downside of bogging down the system you are suppose to be protecting:

QUOTE (page 228)
Note that nodes are careful not to run so many IC programs
at once that it aff ects their performance (see Response, p. 212).


Second, loading an Agent is a action, and since you apparently don't have the book i'll look it up.....page 219, it is a Complex action. So one IC Agent loaded per IP.

I and others have mentioned this before in this thread. What is the problem? You just don't -like- the answer, or what?
GrinderTheTroll
*snip*
QUOTE (Serbitar)
Is there any reason not to load every IC available at the first net hit of the node and track the hell out of everything?


I'm not sure if I understand this part of your question. Do you mean "why not dump all the Agents into a node once the system's on alert?" If so, then it's a matter of preference on how the target would respond. I imagine most systems wouldn't unleash an army of Agents unless there was good reason or unless the node in question is paranoid like that.

You can dump large amounts of Agents at a time with enough IP, but you'd suffer response/rating degradation if you get too many in the same node.

EDIT - I suppose other nodes could release Agents in response to an alert and they could move into the invaded node.
Serbitar
QUOTE (blakkie @ May 4 2006, 04:28 PM)
I and others have mentioned this before in this thread. What is the problem? You just don't -like- the answer, or what?

Yes, I want a justification for a security sheet like thing, where IC is triggered only after a certain number of node ihts. With increasing security (more IC, alert, shotdown . . .) as the node hits increase.
blakkie
QUOTE (Serbitar @ May 4 2006, 03:50 PM)
QUOTE (blakkie @ May 4 2006, 04:28 PM)
I and others have mentioned this before in this thread. What is the problem? You just don't -like- the answer, or what?

Yes, I want aj ustification for a security sheet like thing, when IC is triggered only after a certain number of node its. With increasing security as the node hits increase.

Congratulations, you just wrote it down! Now go forth and be fruitful!


P.S. As has been mentioned a few times you can wait for Unwired to come out with the port of the relavent info. Or hey, why not just go and port it yourself. Past canon seems to be sort of a justification, yes? Page 228 does mention different flavours of IC based on their program loadout.

EDIT: I'm pretty sure I've seen someone doing a translation of the old Tar Baby and such into program loadouts.

EDIT2: rotfl.gif Ya, it was in that thread i linked for you. The first response post.
blakkie
QUOTE (GrinderTheTroll)
EDIT - I suppose other nodes could release Agents in response to an alert and they could move into the invaded node.

That's a interesting idea. IC are independant Agents that can move around node to node. But that's going to introduce a bit of a delay (request for IC goes out, load occurs, IC moves in). They'll too still bog the crap out of the node they are coming to the aid of when they move in.
damaleon
Here is something I read in the Linking and Subscribing section:
QUOTE
The subscription list may be unlimited in size, but the number of nodes, agents, or drones that a persona may actively subscribe to (access) at any one time is limited to the personaís System x 2.


If you apply that to the node (from someone's commlink to the access gateway to a AAA host) you are limited to at most 12 IC accessing/protecting any one node, less since it is connecting to other nodes.

For someone's commlink, I would count any cyber that is accessing wirelessly, like cybereyes transmitting visuals to another team member or a smartlink to a weapon without a skinlink, against the subscription limit. A drone could be activated, issued instructions and then unsubscribed to act on its own, no longer counting against the limit.

For a rating 5 network gateway, it would likely have its own IC (say 2 for detecting hackers), subscribing to 2-3 nodes and multiple IC actively running on other nodes within the network (up to 5-6 other IC possible, takes at least a CT to show up) that it can call for backup if an alert is triggered. I would have all the programs that the IC can use running on the node it is loaded on, so IC and programs they use would count against response, however you decide to handle major hardware.
Personally I don't see a AAA corp's system slowing down if it is running only 6 programs. I would consider any "node" to be a mesh network of equivalent devices hardwired together with its own gateways, allowing a much higher number of programs, but still limiting active subscriptions to System x 2 (so 12 for any rating 6 system).

But this is all how I would do it, so as someone else already said, just do what seems reasonable.
GrinderTheTroll
QUOTE (blakkie)
QUOTE (GrinderTheTroll @ May 4 2006, 03:32 PM)
EDIT - I suppose other nodes could release Agents in response to an alert and they could move into the invaded node.

That's a interesting idea. IC are independant Agents that can move around node to node. But that's going to introduce a bit of a delay (request for IC goes out, load occurs, IC moves in). They'll too still bog the crap out of the node they are coming to the aid of when they move in.

Reminds me of Matrix 2 when all the Agent Smith's started coming into the area a little at a time... then Whammo!
mdynna
Ok, just stop and think about this whole "everything is a node with the same rules" concept for a moment. As has been pointed out, the cheapest/most efficient way to make a Rating 6 Commlink is to buy a Meta Link and upgrade everything up to 6. Right? If a suped-up Meta Link can do exactly every other device can do, then why not do it? You know corps are always looking for the most cost effecient way to do things. So, they would use the "upgraded Meta Link" method. Think about it. Renraku's AI Research Host is a modified Meta Link Commlink. Aztechnology's Blood Magic Library is a modified Meta Link Commlink. News headline: "THE ZURICH ORBITAL HOST HAS JUST UPGRADED ITSELF TO A META LINK COMMLINK"

Meta is now the most powerful Megacorp in the world because EVERYTHING IN THE WORLD RUNS ON THEIR COMMLINKS. "Everything is a node?" "Corp hosts use the same rules as everyone's Commlink?" No. My SR4 Matrix is not built on legions of Meta Link commlinks. What's yours?
Kanada Ten
Meh, the corps probably would make it cheaper than the Meta since they don't need the roll out keyboards and other interfaces for most things. Now we know why it only took 5 years to upgrade all the tech, though. Just billions and billions of skeletal commlinks interconnected.
James McMurray
In today's world it is cheaper in terms of straight money to buy a junk computer and upgrade it. It doesn't happen in corporations for a few reasons:

1) you incur more costs in terms of the time it takes people to build these computers for you

2) large computer companies offer discounts for large purchases

3) many managers who aren't tech savvy don't think that a hoemade computer can be as good as a brand name one

In SR you also come across situations where many of the bigger corporations either manufacture or have ties to people that manufacture computers. If you own an assembly line, building name brand computers is even cheaper than buying and upgrading.
Glayvin34
My question is still about the Load that an Agent has. Say you've got a Response 6 ICberg running 2 Agents and you don't want to decrease the response, does the program list look like this:
1.Agent1
2.Agent2
3.Armor
4.Attack
5.Stealth
or like this? (with a -1 response)
1.Agent1
2.Armor1
3.Attack1
4.Agent2
5.Armor2
6.Attack2

Does an Agent use programs that the node is running or does it use its own load and can't share programs? Because if they use programs on the node, then one program is taking multiple actions each Turn (which I guess could be possible).
damaleon
QUOTE (Glayvin34)
My question is still about the Load that an Agent has.  Say you've got a Response 6 ICberg running 2 Agents and you don't want to decrease the response, does the program list look like this:
1.Agent1
2.Agent2
3.Armor
4.Attack
5.Stealth
or like this? (with a -1 response)
1.Agent1
2.Armor1
3.Attack1
4.Agent2
5.Armor2
6.Attack2

Does an Agent use programs that the node is running or does it use its own load and can't share programs?  Because if they use programs on the node, then one program is taking multiple actions each Turn (which I guess could be possible).

I would run it as option 1, each agent has access to all 3 programs. I see it as multiple users reading the same part of a cache to queue an instruction.
GrinderTheTroll
QUOTE (Glayvin34)
My question is still about the Load that an Agent has.  Say you've got a Response 6 ICberg running 2 Agents and you don't want to decrease the response, does the program list look like this:
1.Agent1
2.Agent2
3.Armor
4.Attack
5.Stealth
or like this? (with a -1 response)
1.Agent1
2.Armor1
3.Attack1
4.Agent2
5.Armor2
6.Attack2

Does an Agent use programs that the node is running or does it use its own load and can't share programs?  Because if they use programs on the node, then one program is taking multiple actions each Turn (which I guess could be possible).


It's not the Agent's rating that degrades the node (although the Agent is capped at the node Rating), it's the number of Agents running in a node that degrade it. The Agent's Reponse equals that of the node it occupies.

A Rating 3 node can have 6 Agents running and any Agent will function at a MAX Rating of 3 even if >3. If the Agent moved into a Rating 5 node, then that node could handle 10 Agents before degrading and the Agent would function at a MAX Rating of 5 even if >5.

IIRC the limit to how much you can load into an Agent, is its Rating x2 in Program Ratings. So Agent 6 could have Trace-3, Attack-3.

Can anyone verify this?
Shrike30
Did you mean Trace-6, Attack-6?
Glayvin34
QUOTE (GrinderTheTroll)
It's not the Agent's rating that degrades the node (although the Agent is capped at the node Rating), it's the number of Agents running in a node that degrade it.† The Agent's Reponse equals that of the node it occupies.

A Rating 3 node can have 6 Agents running and any Agent will function at a MAX Rating of 3 even if >3.† If the Agent moved into a Rating 5 node, then that node could handle 10 Agents before degrading and the Agent would function at a MAX Rating of 5 even if >5.

IIRC the limit to how much you can load into an Agent, is its Rating x2 in Program Ratings.† So Agent 6 could have Trace-3, Attack-3.

Can anyone verify this?

First, from page 212:
"Response may be affected if you run too many programs. For every x number of programs you have actively running, where x = System rating, your Response is reduced by 1. So if youíre running 10 programs with a System 5, your Response will be reduced by 2."
So at each multiple of your system you take a -1 to response.

It's that second part that I'm wondering about. Does an Agent's Load count against the total number of programs running? It says on page 228 that
"Agents can be loaded up with copies of your programs so that the agent may employ them on its own. If an agent is acting independently, any programs itís carrying must be active, and so may affect its Response."

Now it seems to be that the Agent runs the program on whatever node it's on, affecting response. So if you've got the aforementioned attack IC, you'll need it to run about 3 programs to be effective in cybercombat, plus the fact it's an Agent, so that means one attack IC with Attack, Stealth and Armor takes up 4 programs. Does that not sound right to anyone?
Glayvin34
On 232 there's that "Simultaneous Combat in Multiple Nodes" section, it seems to indicate that you can get attacked through any node you're subscribed to. So if you're in a node with 10 ICbergs subscribed to it, like any high-security "Mainframe" (I use the word lightly) is likely to have, then each of the ICbergs attacks you.
So I guess being subscribed to a node and being in the node is the same thing for personas? Agents can enter a node so if they're tracked the track just goes to whatever node they are running on (I think).
damaleon
Here is how I see it.

System/Response 6 node running the following programs:
1. Agent Attacker (Pilot 4)
2. Agent Detector (Pilot 6)
3. Agent Hunter (Pilot 5)
4. Attack-6
5. Analyze-6
6. Blackout-6 (node loses 1 response, now at 5)
7. Stealth-6
8. Armor-6
9. Track-6
10. Exploit-6
11. Decrypt-6

Node is now effectively System/Response 5 and Programs only get max of 5 dice. Another program or agent would reduce response again, down to 4.

Now the Agents are loaded as such:
Attacker - Defend node
-- Attack, Blackout, Armor, Stealth (-1 response for 4 programs with a 4 pilot)
Equivalent to a 4 system(pilot), 4 firewall, 4 response persona (node response at 5, -1 for programs), programs limited to 4 by Pilot rating

Detector - Detect intrusion and locate source of intruder
-- Analyze, Stealth, Track, Exploit, Decrypt (no response loss due to programs, but pilot is limited to 5 effectiveness by node response of 5)
Equiv. to a 5 sys, 5 firewall, 5 response (limited by 5 response node), programs at 5

Hunter - Defend node or attack intruder's node
-- All programs listed above (8 programs, -1 response loss for 5-9 programs loaded)
Equiv. to a 4 sys, 4 firewall, 4 response persona, programs at 4

Obviously this would not be an ideal loadout, but I was trying to cover as many situations as possible. What do you guys think?
Big D
Backing away from the exact mechanics for a second...

If I'm hacking a server, and see some alert flags pop up, but want to stave off the inevitable for a few rounds while my team gets out, can I do something as simple as uploading a few agents and then ordering them to go out and create havoc in other nodes on the system, then uploading a few more?

It sounds like once I get into a system, I can basically DOS it with agents. The IC and deckers will be so busy tracing down my agents, that they won't be able to get control of the system. I probably won't either, but this is a "run away" tactic, not a take and hold ground tactic.

Also, was there any mention of party ice in the BBB? I don't remember seeing it. I think it would need some kind of special treatment under the "stacking" rules discussed here.
Glayvin34
Yeah, Damaleon's model seems to be right. But about the Hunter Agent, do Agents lose response for the programs they have loaded? They're not running the programs themselves, the node is. All that's in the book is the programs "must be active, and so may affect its Response" line on page 228, I think that's referring to the Response of the node the Agent is running on, not the Agent itself.
Glayvin34
QUOTE (Big D)
Backing away from the exact mechanics for a second...

If I'm hacking a server, and see some alert flags pop up, but want to stave off the inevitable for a few rounds while my team gets out, can I do something as simple as uploading a few agents and then ordering them to go out and create havoc in other nodes on the system, then uploading a few more?

It sounds like once I get into a system, I can basically DOS it with agents. The IC and deckers will be so busy tracing down my agents, that they won't be able to get control of the system. I probably won't either, but this is a "run away" tactic, not a take and hold ground tactic.

Also, was there any mention of party ice in the BBB? I don't remember seeing it. I think it would need some kind of special treatment under the "stacking" rules discussed here.

Ooo, good one. You could have a Comm full of Agents that all run out and slow down whatever node is running the IC that is coming after you. If they were all loaded up with programs, they could go and run them all on the ICberg protecting the system you're infiltrating, and the Mainframe's Response would bottom out. That's kinda cheap, though.
damaleon
QUOTE
Payload
Agents can be loaded up with copies of your programs so that the agent may employ them on its own. If an agent is acting independently, any programs itís carrying must be active, and so may aff ect its Response (see p. 212). Any program run by an agent
is limited by the Pilot rating.


I read the independent agent and payload rules like the agent is a mini system, running on the node and granted access to programs also running on the node. I see the agent being able to handle only so many threads before it loses efficency. Basicly it is only smart enough to handle access to so many programs before its logic routines take longer to decide what to do, like in turn based strategy games when the computer has more and more units to control and more places to attack you from.
damaleon
Yeah, I would agree with Big D, as long as you can hack the proper access needed to load an Agent on their system, it should be possible. I would restrict running Agents to security or admin access, so it takes longer to hack in, but if you manage to do that, you should be able to use their system's resources against them.

I wouldn't allow you to do it immediately though, except for what programs/IC/agents are already available to the node your on. If for some reason there is no agent available to that node, it isn't rated high enough for you, I would make you spend several actions transfering some of your's to the node before you could activate them. Not a problem when the system isn't on alert, but it can take precious IPs away if you're being attacked, which is a suitable penalty if you don't think ahead.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Dumpshock Forums © 2001-2012