1) If you want three Rating 4 Agents running (eg, one out on a search, one helping you hack, and one in your commlink acting as IC) do you need to purchase 3 agents? Or do they act like all other software, and you only have to buy it once and run it many times?

2) Is it required to attach cyberware devices to your PAN for interoperability? I realize it says cyberware devices can talk to each other directly:

SR4 p. 330 Cyberware
In addition to wireless functionality, most cyberware devices are equipped with a direct neural interface (DNI) that allows the user to mentally activate and control their functions. They can also be linked to other cyberware implants.

But what about a weapon's smartlink? Can it communicate directly with the user's imagelink and smartlink without needing to be routed through their PAN by their comlink?

3) And if that is the case, then could skinlink handle the connection? The Pistol's smartlink would be connected via skinlink, which would somehow connect to the imagelink and smartlink in their eye's?

Thanks,

milspec

For 1, considering how much more agents cost than anything else I'd allow it, though I might make you pay for a second agent to act as IC.

I've mentioned this elsewhere, so apologies to those hearing a broken record.

Before I go under the knife, I'd ask for all of my cyber to network using only DNI. I'm tempted to keep the old-fashioned hardwire interface for my gun and commlink, even.

I ain't interested in being a network node.

Agreed. If I want to have foreign signals invading parts of my body--even if they are replacement parts--I'll plug 'em into a datajack.

agents and IC work a little bit differently... i believe you could have an agent active and checking for problems, but i think you need IC to launch itself when there is a problem, agents have to be manually launched (or such is my understanding... that may be off...)

but anyways, you could certainly have an agent with attack, trace, or whatever active, and on patrol. you would just have to manually turn it on/off. or by some rating 1 IC to turn it on/off for you

As I read things, Electronics by default are Wireless, but converting them to wired takes no extra money, just extra effort on your character's part. Say you are turning off the wireless capacity. Tell the Doctor not to include the parts that broadcast. You are done. Now, Skinlinks are another issue-they require special modifications in the cost of 50 . But wires are free. (Just troublesome at times)

Keep in mind though that Average Joe 2070 Probably jsut lets everything be wireless. He doesn't think in terms of security, but in terms of convienence. In fact, he leaves most devices wireless becausze it never occurs to him not too. So yeah, his cyberware has wireless connectivity, even though it doesn't need it. He never thought to turn it off. If you ask for non-wireless goods, they will be available, but the clerk may look at you funny and might contact the cops to report 'Suspicious activity".

So yes, your cyber interacts with other cyber in your system without wireless connectivity. You can link other systems to your brain with physical conductors (Such as wires) through a datajack or trodes. This will keep you safer from hacking.

Also remember that almost everything comes with an RFID tag these days. GET A TAG ERASER!!!!! This is the most vital peice of gear a runner needs. Without it anything they have on them is bradcasting. Granted, it's like a drop in the data bucket, and would be a pain in the ass to locate someone by the RFID of their clothes, but it is possible, so get rid of those tags. And don't think your Cyber comes tag free. I bet those damn corps put RFID, maybe even security grade RFID tags into all the cyber they produce. Twould be a pain if the cyber you tried so hard to make hacker-proof broadcasted on demand when the corps wanted.

besides... many shadowrunners will have gotten their cyberware through, shall we say, "non-standard" channels.

the shadowclinics don't care, and the corp/government clinics probably prefer it to not be wireless, so they wouldn't likely be looking at you funny, IMO.

i think when the book says people look at you funny for wanting non-wireless stuff, it's talking about clothes, and vehicles and such, that you might buy at a normal store.

no, my point is that given the usual place a shadowrunner will be buying cyberware, the people there are payed for not asking you for ID, a license, or any of that other stuff. what makes you think they're gonna raise eyebrows if you ask for non-wireless gear?

Actually I think it's the Commlink's firewall that detects the intruder, then activates the IC in response. So I guess that's just a customizable part of the firewall:

As I see it, the way it works is that the user sets up what happens when the firewall detects an intruder. They can tell the user (who will usually send some kind of backup); they can reboot the machine; they can do all the other nifty things that are mentioned in the active alert section (p. 222-3), and/or they can activate an Agent to fight off the intruder. Agents activated by firewalls are callid IC; that seems to be the only difference.

Now Pilots, those I suspect are different, though they are priced the same. Agents are coded to do Matrix work, while Pilots are coded to do drone work. And so you have to buy them seperately, but an Agent and an IC program is the same thing, differing only in the manner in which they are activated.

That sound right?

"Activated" vs "run" vs "loaded" is causing confusion I think.

The firewall tells the IC/Agent to go after a particular user, but it isn't loading the IC/Agent right there on the spot (I think, please correct me if I'm wrong) because that would normally take time (don't have a page reference on me, but I believe it is a complex action).

Pilot is variant of System...

Always was, always will.

The Pilot software is installed instead of a System, though... not like Agents or IC, that run on a System.

Is there a page reference where it tells you that IC on your commlink don't count against your running programs?

yeah, that sounds about right.

begins to make sense why corps aren't so bothered by the idea of having to defend wireless networks, doesn't it? protecting the network is cheap. much cheaper than having to renovate your building to install a system.

By that same token, it's also pretty cheap to mount an army of Agents attacking a corp host.

Agents from other nodes would end up logging into the node under attack and would slow it down. I think. Sec hackers with agents on their commlinks would be just fine though.

right then. so the node under attack just piles in more and more agents until it's system 0, and then noone can do anything except those relying on another node (ie the corp hackers vs the hackers who organised the assault). considering all the agents would have pilot ratings of 0, at that point (except those loaded on commlinks), they would be insignificant, and the corp hackers could feel free to extract info from them... you know, fun stuff like where they came from, what their target is, what kind of mission (datasteal, crash system, etc) and so forth.

Pretty much.

This is, btw, why you don't pull a DoS attack with your *own* Agents.

Provided that the Agents weren't "clean", containing only a pilot program with no identifying information. About all that you might get would be some information on what account loaded them on to the system. Granted, to do the attack that way would require a large number of complex actions (node's System^2 actions at most depending on it's load before you started). VR would help but only on nodes with a System of 3 or less could you do this quickly.

I would guess that a node under enough load to slow down would set off some sort of alarm so a sec hacker could come in and check things out since they normally would not run that many programs on their own. A node reduced to 0 might even reboot.

Yeah, I'd say that any security hacker worth the job would notice a DOS attack the instant the response started to slow. Additionally, I'd say that most large corporate networks have a distributing processing system between so many nodes that you're not trying to overload a response 6 machine, but thousands of them.

However, if you loaded your agent with another copy of the agent, and instructed it to load copies of itsef...

Vonn Neuman Agents!

Does any one have any ideas as to how the load balancing that PlatonicPimp alluded to would work in SR4? Maybe corp nodes on server grade hardware rather than a commlink would get some sort of a multiplier to their System for load purposes?

Well, in specific terms, I'd say that corporate networks are made up of hundreds if not thousands of individual ondes, each having the response and system ratings we are familiar with. However, they are set up in such a way that programs running are run on a node that has spare processing power before it is run on one that would have it's response reduced. Thus, in order to reduce response times, you would have to fill up every node's processing power first.

In practice during games? You can't do it, the corps have too much processing power for you to effect.

So by your thinking you simply wouldn't restrict corp nodes to the Sytem limit on active programs?

heck no. otherwise, how would they ever get anything done? you'd need one host for every 2-3 people.

Not on any level that would ever effect gameplay.

Basically. Otherwise to simulate Corp hosts would require system ratings so high as to preclude the chance of ever breaking in. I mean could you realitically break into a system running a rating 2000 system with a 1000 firewall in less than a month?

Is there anything in the canon rules to explicitly say that or is this stuck being an errata/houserule?

Edit: Btw, I like the rule.

i think it's just a house rule, atm. hopefully that will change with wireless. when it comes out. which isn't for a while, as i understand it.

I would love to see that book make it to second in the list, right after a book talking about the 'plexes that are now going to be highlighted by SR4.