Hacking a Node: Diary of Failure

[Username17]

So more than one person has asked me to describe how the RAW matrix rules work and also why I say that they don't work. So in the interests of getting that all out in the open, I decided to write up a description of point by point how you would go about hacking into a Renraku node according to the RAW rules in SR4A and Unwired. To add to the excitement, the ways in which the matrix rules allow one side or the other to simply bypass the system and go into infinity crazy town will be Footnoted[1] rather than breaking into the narrative and confusing the issue. So here we go: one Hacker, one Renraku node, go:

[1]: Like this.

---

You[1] decide that you want to hack a node because you suspect it of having a data file that you want.

You decide that you need to create a false and temporary Access ID. This is completely required, since otherwise when you sign out, Renraku can just plain call you[2]. The dicepool is Hacking + Spoof, but you only need one hit (it's unopposed unless you are currently being traced) and you can do this in downtime, so whatever. You can also do this step by routing your hacking through any electronic item for the duration of your current hacking job and then ditch the item. If you need a specific Access ID, you need to have analyzed that access ID's Icon with Matrix Perception[3] by being in the same node as the Icon and winning an opposed Analyze + Computer vs. Stealth + Firewall test.

Then you have to find the node. This is ridiculously hard. It's an Electronic Warfare + Scan test, and as written it's an extended test (with a threshold of 15 or higher) to find a Hidden Node and mysteriously a non-extended test (threshold 4) to find a node that is not hidden. So as written, with 6 dice it's an average of 10 rolls to find a node that is not hiding and only 8 to find one that is. May be a typo, who can tell?

Armed with your fake and temporary Access ID, you attempt to hack yourself an account into the node you found. This is done by making opposed extended tests against the node. You roll Hacking + Exploit and need to get a number of hits equal to the Firewall of the node (or 3 or 6 higher if the node requires a "security" or "admin" account, which is just a preference toggle for whether you want hacking your systems to be easier or harder). Every time you roll dice, the node gets to roll Analyze + Firewall, trying to add up to a number of hits equal to your Stealth. When you get your total of hits, you get your account. If the computer gets ts number of hits, it sets off an Alert[4]. It is entirely possible for both of these to happen simultaneously. You can also choose to take hours instead of Complex Actions to make these tests, and then the node only gets one chance to spot you instead of one per test you make[5].

Now that you are in the node, you need to render all the icons. This takes a Simple Action to find out how many icons there are (making an opposed Computer + Analyze test vs. Firewall + Stealth test to count any of the icons that are Stealthed), and then an additional Simple Action to actually identify each of the icons that you have counted (which are also an opposed Analyze + Computer test against Stealth + Firewall to identify any of the stealthed icons you have counted)[6].

Whether an Alert has been triggered or not, the node may well have an active IC protecting it. While you are rendering the node, it is rendering you. If it notices you (Analyze + Agent Rating vs. Hacking + Stealth), it will compare you to its list of who is supposed to be on the node (separate from the Node's list, so you haven't had a chance to hack it yet - so you're not on this list). It can then take any actions it want against you until you successfully render it. Since you're the new kid on the block, it can start attacking you as soon as it renders the programs you came in with[7]. Also, it can call for backup before it even renders you[8]. Or, since it presumably already has a Security Account on its own node, it can attempt to kick you out[9].

If the IC doesn't spot you and you didn't set off an alarm, skip this step altogether. You can get rid of the IC in several ways. They seem to think you want to engage it in Cyber Combat, which involves rolling initiative and then following the rules for casting low force manabolts at people with spell defense who also get to soak damage and have armor. What you'd actually want to do is to spoof a command to the IC to make it go away. This is just Hacking + Spoof vs. Firewall + Pilot. One net hit and the IC stops all harassing of you forever. If for some reason you are being bothered by one Agent, it's pretty much over in one action.

When you have rendered the target file, you can try to access it. It has a Data Bomb on it, which means that you need to Disarm it before you open the file or you'll get hit by very weird pile of damage. For no reason, Data Bombs have a unique damage mechanic where they inflict matrix boxes equal to Rating multiplied by the score shown on a single d6. Since Data Bombs normally run in ratings of 3-6, this means that you usually have a 2/3 chance of just crashing: no save. Also, Data Bombs can be set to nuke the file you were after. So having one go off is absolutely not an option. You roll Hacking + Defuse (incorrectly written "Disarm" in the book, but we know what they mean) and the Data Bomb rolls twice its rating. If you get more hits than the bomb, you get the data. Otherwise the data is lost and you probably crash. Note that if you're a Technomancer, that taking 30 boxes of matrix damage actually means that you die. So always have a Sprite or Agent open a Data Bomb if you are a Technomancer.

Now you can transfer the file to your commlink. This requires a Computer + Edit test, but you only need 1 hit, there is no consequence for failure, and retries are allowed. So um... whatever.

The file is encrypted. As written, this is a case of rolling dice a bunch of times but frankly essentially automatically getting the data in a few passes. Watch out, because there is an optional rule in Unwired where you automatically fail to crack the encryption instead.

[1]: Hackastack The idea of "you" in the matrix is pretty weird. It's "your" persona, but it doesn't have to send any information to your brain or even your physical location, nor does it have to receive any actual commands from "you" ever to be "your persona." You can therefore, decide to run agents on several different commlinks and simply have all visual input recorded or even just trashed. Technically "you" are AR-Hacking out of each commlink simultaneously because they all stay on the matrix and continue to get and take actions without them receiving any input from you or you receiving any input from them. There is actually no limit to how many "yous" there can be on the Matrix simultaneously by this method.
[2]:The action isn't even mentioned in SR4A despite the fact that it is well established to be something you absolutely need to do every time you begin hacking anything at all (Unwired, p. 99).
[3]: Dropout The failure point should be obvious. You only make Matrix Perception tests to analyze icons that share a node with you. So if you need the Access ID of something that is for example in the node to get in the node, the node is literally unhackable. And yes, nodes can refuse you without rolling dice for not having an Access ID (Unwired p. 99) or for having an Access ID they don't like (Unwired, p. 101).
[4]: Denial of Service: As written, enough Alerts will force any system to reboot. Since you can generate as many alerts as you want with a hackastack running no stealth, you can effectively crash any node in one combat turn. If your goal is merely to shut the system down, you don't need skills or even dice rolls - it just automatically happens. It takes an average of 12 rolls to resolve, but there literally can be no other outcome if that's what you want.
[5]: Note that the highest threshold you could possibly have is 14 (Firewall 8, Admin account), so if you have five hours to kill, you can crack into the Pentagon or the Red Pagoda. You probably won't even set off an alert.
[6] Infinity Mirror: Yes. Seriously one Simple Action per icon (SR4A, page 228). There is no actual limit to the number of icons that Renraku can put into a node. So since it takes a minimum of 1 SA + the number of icons to render the node, they can literally keep you searching for days just by leaving thousands of folders open. Anyone who legitimately needs the information can just go to the data file without rendering the whole node. You'd think that Browse and/or Data Search could help you find the info you need, but since you're already on the right node, it doesn't.
[7]: Infinity Mirror: Just as the node can have unlimited numbers of icons, so can you. Unrated programs don't take up any space but they are still programs and still have icons. So you can have limitless numbers of Virtual Pets that happen to have the same icon as your persona, and any potential attacker will need to go through them simple action by simple action before they can attack "you."
[8]: Agent Smith The backup can be in the form of a security spider, but it can also be other Agents that belong to Renraku. Since there is no limit to the number of targets in the CC of an email, Renraku can essentially oppose you with one Agent for every computer - in the world - that is in Renraku's hackastack. Which is arbitrarily large. So you could be opposed by a hundred, or a thousand Agents next turn. Seriously.
[9]: While it takes several rounds to kick you out during which you can hijack the node and prevent it from doing so, changing the Access ID automatically works and boots you and every other foreign user off the system without rolling dice in a single action. I'm not sure this counts as an exploit of the rules, since it's actually given as a helpful tip in Unwired.

---

There you have it. A single node that happens to have a single file on it that you want to download. I'm not including subjective failure points like the fact that you may have noticed that actually having any logic or skills is pretty much meaningless in this equation.

-Username17

[RandomCasualty2]

This was altogether a pretty good explanation. Though I'd liked to have seen how it works in a multi-node system.

I'd also like to see how you'd go about hacking a piece of cyberware that's set to active mode, instead of slaved. Obviously you could hack it at close range, but what if you wanted to hack it from a distance? If you knew it's access ID, you could connect via the Matrix. of course, you don't, and you could only find it out presumably by being within its signal range (which is very small). So it seems its actually more secure being an unsubscribed device in active mode than a device set slaved to the commlink.

Anyway, couple questions...

Then you have to find the node. This is ridiculously hard. It's an Electronic Warfare + Scan test, and as written it's an extended test (with a threshold of 15 or higher) to find a Hidden Node and mysteriously a non-extended test (threshold 4) to find a node that is not hidden. So as written, with 6 dice it's an average of 10 rolls to find a node that is not hiding and only 8 to find one that is. May be a typo, who can tell?

Where does it say that it's 4 to find a non-hidden node? I always though it was 4 to find a node that was hidden that you had some idea of what it was (like that Johnson's commlink), where the 15+ was used if you were scanning for all hidden nodes.

What you'd actually want to do is to spoof a command to the IC to make it go away. This is just Hacking + Spoof vs. Firewall + Pilot. One net hit and the IC stops all harassing of you forever. If for some reason you are being bothered by one Agent, it's pretty much over in one action.

But you'd need an access ID of a spider or administrator to do that right?

[2]:The action isn't even mentioned in SR4A despite the fact that it is well established to be something you absolutely need to do every time you begin hacking anything at all (Unwired, p. 99).

Actually this is mentioned under "Access ID" in SR4A, though I don't blame you for missing it, because it's just one or two obscure sentences and it's never mentioned as an actual action in the list of stuff you can do. They also left out terminating a connection or manually triggering an alarm as actions, though according to unwired IC can do that. Disconnecting someone is only listed as an ARC, and manually initiating an alert isn't mentioned at all if I remember right.

Also, side question about Unwired. How the fuck does the Juhseung Saja IC work? (Unwired p.71) After feeling like I've got a decent grasp of the Matrix rules after multiple read-throughs of Unwired and core, I still cannot understand how that IC works. How the hell can it use spoof to transfer itself to the hacker's node?

[mean_liar]

I've lost my handle on SR rules after 3e. The layering complexity and die-rolling of Rigging and Decking at the end of 3e (ECM! ECCM! more DICEDICEDICEDICE!) made me just have my Decker characters self-construct/self-program unbeatable dice pools so that I could just skip all that shit and concentrate on how long it took rather than if it happened at all, and that worked out pretty well.

According to this 4e Hacking is a shitbomb and I can feel warm and fuzzy that 3e is still a decent game.

[RandomCasualty2]

I've lost my handle on SR rules after 3e. The layering complexity and die-rolling of Rigging and Decking at the end of 3e (ECM! ECCM! more DICEDICEDICEDICE!) made me just have my Decker characters self-construct/self-program unbeatable dice pools so that I could just skip all that shit and concentrate on how long it took rather than if it happened at all, and that worked out pretty well.

According to this 4e Hacking is a shitbomb and I can feel warm and fuzzy that 3e is still a decent game.

In most 3E SR games, you'd get a "No deckers" restriction, and for good reason. 3E hacking really sucked, mainly because it was the minigame that you played while everyone else just went and played video games.

4E hacking is sort of a mess, but the basic concept is definitely worth saving. It's much closer to what SR hackers are designed to do, which is provide support to the team during a run, as opposed to just sit at home and hack things in a solo quest like you did in 3E.

Hacking is considered to be one of the major flaws of prior editions because it just didn't mesh with the concept of a group game at all.

[Username17]

First, the easy part:

According to this 4e Hacking is a shitbomb and I can feel warm and fuzzy that 3e is still a decent game.

Yes and no. 4e Hacking is crap, but it's also Neapolitan: 3 flavors of crap. It is changed unrecognizably (mostly in how it fails to contain Agents, but also in basic topology and core action resolution) in Unwired and again in SR4A from the original SR4 book. Part of this is that apparently Aaron can't write Matrix Rules, but going back to SR4 and even Eclipse Phase it is clear that Rob can't write Matrix rules either. And that's hardly new to SR4. In SR3 the BBB's rules don't look much like the ones described in Matrix 3 or Target: Matrix. SR3 also had a few stealth overwrites in the form of some wildly incompatible changes made in the State of the Art series and even Shadows of North America (which brought in the Wireless Matrix that we've been struggling to find a workable topology for ever since).

But let's be really clear: none of those systems of the Matrix actually worked. None of the wildly incompatible versions for SR1 or SR2 did (not the BBB versions nor the total overhauls in the Virtual Realities series). In every case, the only really functional way to handle the Matrix involved either having the Decker player invest huge amounts of time in making sure his hacking was the bomb and figuring out how he thought the hacking rules worked and then fucking ignoring all of that shit while you either had the Hacker roll a computer test to simulate whole branches of decking or just handwaved the whole thing or not even doing that and just having the Decker be an NPC that acted as a complete plot device.

SR4's Matrix is tantalizing close to being playable, and some people streamline it down to about 1/10th scale of die rolling and voluntarily decide to not use Drop Out, Agent Smith, Hackastacks, or Infinity Mirrors. Heck, many people don't actually understand the rules well enough to even understand how those problems work. Many people will swear on a stack of old Dragon Magazines that SR4's Matrix is playable "as is" because they don't even understand how many steps they are leaving out. Almost everyone thinks you can use Browse to find data you want, and almost no one understands the implications of having to spend a Simple Action to determine whether "an icon" is an Arrow, a Program, an Agent, or a Persona. The Infinity Mirror is an exploit that most people have deleted in their house rules by mistake.

So even though the as-written SR4 Matrix is completely unsalvageable and requires a top-down rewrite, it's the best Matrix system the game has ever had. It's the one that is so close to being a playable system in enough ways that people who having working games think they are playing by the rules.

This was altogether a pretty good explanation. Though I'd liked to have seen how it works in a multi-node system.

Multi-noding it adds two failure points:

• Rabbit Hole There is literally no limit to the number of nodes that you could be required to tunnel through, and it can be set to cycle the Access ID of intermediary nodes if any alerts ever come up, meaning that the character basically can't succeed.
• Drop Out If you multi-node it, you can actually make it so that your node communicates to the outside world through a slave node that only accepts log-ins from the main (which itself is wired) such that they can still make accounts on other systems but people have to spoof an Access ID that they literally cannot know to hack the system. Meaning that basically nothing is hackable.

Where does it say that it's 4 to find a non-hidden node?

That was poorly written on my part, sorry. The regular scan rules are a linker to the Find Hidden Node action in SR4A. That's apparently what every node counts as according to the latest ruleset. The cutoff is that looking for a specific node you think exists is a one-off "very difficult" test (threshold 4), while finding every node is a merely time consuming Extended Test (roll until you get 15 hits with no consequence of failure). At medium dicepools it is literally more time consuming on average to find a network you are looking for than it is to find all the networks and then home in on the one you want.

But you'd need an access ID of a spider or administrator to do that right?

Yeah. Although since you're in the node that is running that IC you can actually spoof a command to the node to unload the IC from active memory. Once the other literal millions of IC show up, there is basically nothing you can do, since you have no idea whatsoever what Admin accounts each one listens to.

I don't blame you for missing it, because it's just one or two obscure sentences and it's never mentioned as an actual action in the list of stuff you can do. They also left out terminating a connection or manually triggering an alarm as actions, though according to unwired IC can do that. Disconnecting someone is only listed as an ARC, and manually initiating an alert isn't mentioned at all if I remember right.

The SR4A Matrix section is spotty. It's basically unusable without Unwired in the sense that it doesn't actually give complete descriptions of sufficient amounts of the topology and basic interactions for a first time reader to understand how any of it holds together. The section only makes sense as a changelog of the SR4 -> Unwired hierarchy of rules interactions. But I seriously have no idea whether statements in SR4A are supposed to overrule Unwired rules where they contradict each other or not. The act of medium-term spoofing your Access ID with a program is only indirectly mentioned in SR4A in that it specifies that Unwired has "other options" for anonymity of your data trail.

Also, side question about Unwired. How the fuck does the Juhseung Saja IC work? (Unwired p.71) After feeling like I've got a decent grasp of the Matrix rules after multiple read-throughs of Unwired and core, I still cannot understand how that IC works. How the hell can it use spoof to transfer itself to the hacker's node?

Uh... it doesn't work.

It would need to Exploit onto your commlink (which it doesn't have), and it would then need to spoof an Access ID your comm would accept, which would require Analyzing a trusted source (which it also doesn't have). Also it is then supposed to use Browse on your Hacking commlink to find your personal information despite the fact that Browse does not work that way and no Hacker is going to keep personal information on their hacking combat commlink.

Those suites are basically just wasted space because you can't pass step one of the Infinity Mirror even if your opponent isn't using the exploit unless you have Analyze. And only two have Analyze and neither of them do anything directly (the Bloodhound is just an automated Trace and the Kitsune's only trick is to trigger an alert).

-Username17

[RandomCasualty2]

• Drop Out If you multi-node it, you can actually make it so that your node communicates to the outside world through a slave node that only accepts log-ins from the main (which itself is wired) such that they can still make accounts on other systems but people have to spoof an Access ID that they literally cannot know to hack the system. Meaning that basically nothing is hackable.

This one is pretty interesting. How would you set a slave node to accept incoming communications from the Matrix? I got the impression that a slave blocked all incoming traffic except from its master (or someone spoofing the master).

Yeah. Although since you're in the node that is running that IC you can actually spoof a command to the node to unload the IC from active memory. Once the other literal millions of IC show up, there is basically nothing you can do, since you have no idea whatsoever what Admin accounts each one listens to.

What access ID would you be spoofing here?

The section only makes sense as a changelog of the SR4 -> Unwired hierarchy of rules interactions. But I seriously have no idea whether statements in SR4A are supposed to overrule Unwired rules where they contradict each other or not. The act of medium-term spoofing your Access ID with a program is only indirectly mentioned in SR4A in that it specifies that Unwired has "other options" for anonymity of your data trail.

What contradictions with Unwired are there in SR4A? I can think of a lot of omissions, but not any straight up contradictions.

Uh... it doesn't work.

It would need to Exploit onto your commlink (which it doesn't have), and it would then need to spoof an Access ID your comm would accept, which would require Analyzing a trusted source (which it also doesn't have). Also it is then supposed to use Browse on your Hacking commlink to find your personal information despite the fact that Browse does not work that way and no Hacker is going to keep personal information on their hacking combat commlink.

Ok that's basically what I thought, since it's description totally contradicted what I know about the Matrix. I wasn't sure if I missed some special use of spoof or what, given that reading the matrix rules was one of the most complex set of RPG rules I think I've ever had to go through.

[mean_liar]

Discussion by proxy! Crossposting for fun!

EDITED for clarity.


From another forum:


...


I dunno. I've played several technomancers, dealing with both hacking and rigging. The rules have always worked for me, without the need for house rules. You just have to take the time to read and understand the rules, and it helps if you've got a basic idea of how computer systems and networks work. But then, my DOS-fu is still strong. Once you start thinking about the Matrix in terms of Johnny Mnemonic, things become simpler.

As far as that thread goes, speaking from the standpoint of someone who knows a thing or two about computer systems, it isn't getting in that's the hard part. Anyone with enough knowhow and time can crack any system, even the Pentagon. The trick is getting in, finding the information you need, and getting out, without getting caught, or leaving a trail that can be followed.

To be honest, the poster of the thread apparently either does not have a clear grasp of the rules, or is one of those people that goes and looks for ways to exploit the rules, just so he can hold them up and say they're broken.

To go through his objections:

[1]: Hackastack The idea of "you" in the matrix is pretty weird. It's "your" persona, but it doesn't have to send any information to your brain or even your physical location, nor does it have to receive any actual commands from "you" ever to be "your persona." You can therefore, decide to run agents on several different commlinks and simply have all visual input recorded or even just trashed. Technically "you" are AR-Hacking out of each commlink simultaneously because they all stay on the matrix and continue to get and take actions without them receiving any input from you or you receiving any input from them. There is actually no limit to how many "yous" there can be on the Matrix simultaneously by this method.

The poster is slightly mistaken about the fact that you can be in many nodes at once. While this is correct, he neglects to mention that your persona can be attacked in each and every one of these nodes, and that you must split your dice pool between all of them to defend or fight back, and that all damage done to the persona applies to ALL the nodes you're in, not just the one. Moreover, you can only concentrate on one node at a time, so you can't hack from multiple points at once. So you can be in as many nodes as you want simultaneously. Its just stupid, and ineffective.

Moreover, you can only actively subscribe (access) System x 2 nodes/drones/agents/whatever at a time.

[2]:The action isn't even mentioned in SR4A despite the fact that it is well established to be something you absolutely need to do every time you begin hacking anything at all (Unwired, p. 99).

The poster is incorrect in how spoofing the datatrail works. The hits generated on the test to create a spoofed access ID increase the threshold a track program needs in order to trace you. You can get by with a single success, if you're extremely lucky, but typically you're going to need more. Unless, of course, you like having Lone Star knocking on your door.

The action to generate a false ID is in fact, mentioned in the core book, on page 225, under the heading "Spoofing the Datatrail", which the poster apparently never read. It requires a Hacking+Spoof(2) test to generate a false ID, or a Hardware+Logic(2) test to make a device generate a false ID.

[3]: Dropout The failure point should be obvious. You only make Matrix Perception tests to analyze icons that share a node with you. So if you need the Access ID of something that is for example in the node to get in the node, the node is literally unhackable. And yes, nodes can refuse you without rolling dice for not having an Access ID (Unwired p. 99) or for having an Access ID they don't like (Unwired, p. 101).

Nodes can drop you for not having an access ID, or for changing IDs while in the node. However, the only thing close to what this person is talking about would be the Limitation feature for some programs, mentioned on page 144 of Unwired, that prevents software from being used in certain ways. Unless he's talking about trying to log on to a system with a bad ID and password. And that's the same trying to log on to your email account with the password satrorwaduy when your password's really 12345. That's not hacking, that's being an idiot.

[4]: Denial of Service: As written, enough Alerts will force any system to reboot. Since you can generate as many alerts as you want with a hackastack running no stealth, you can effectively crash any node in one combat turn. If your goal is merely to shut the system down, you don't need skills or even dice rolls - it just automatically happens. It takes an average of 12 rolls to resolve, but there literally can be no other outcome if that's what you want.

Ignoring the references to his earlier flawed notion of hackastacking, Denial of Service attacks do work. The tactic is nothing new, and hackers use it today, typically using specialized bots to do the attacks for them, usually using a worm or virus to plant the bots in a number of systems, to maximize effectiveness. However, for a System 5, Response 5 node (the least you'd expect for a megacorp) it would take at least 100 bots to freeze the node with a DDOS attack. The only way to do this is to either have a group of hackers attacking the same target, or to have a botnet. Either way can be effective, but has its dangers. It only takes one weak link in the hacker network to expose everyone to a heap of trouble. Botnets can be hacked, and turned against you. Both take time and effort to set up, and neither one is subtle. Since most of the time, the idea is to get in, get out, and not have them know you were there, DDOS attacks are not something you want to do on a regular basis.

[5]: Note that the highest threshold you could possibly have is 14 (Firewall 8, Admin account), so if you have five hours to kill, you can crack into the Pentagon or the Red Pagoda. You probably won't even set off an alert.

Given enough time, anyone can hack anything, unless it is on a stand-alone system, cut off from the matrix. However, getting into the node isn't the problem. You have to dodge IC, and Spiders. You have to check for alerts that get tripped when certain files are accessed. You may have subnodes that require their own access codes to get in. And the list goes on. Now the more sensitive the information on the node, or subnode, the more likely that there will be additional security.

You don't think getting past the front door of corp HQ means you're home free when stealing something from their vault, do you? Then why should it be that way in the matrix?

Better yet, just read the section on page 222 of the fourth ed book titled "Hacked! - Once Inside". If you have the 20th anniversary book, check page 238 for Security Responses.

[6] Infinity Mirror: Yes. Seriously one Simple Action per icon (SR4A, page 228). There is no actual limit to the number of icons that Renraku can put into a node. So since it takes a minimum of 1 SA + the number of icons to render the node, they can literally keep you searching for days just by leaving thousands of folders open. Anyone who legitimately needs the information can just go to the data file without rendering the whole node. You'd think that Browse and/or Data Search could help you find the info you need, but since you're already on the right node, it doesn't.

First off, the page reference he gives is bogus. Second, the whole point is complete crap. You don't need to render the icons. They area already rendered in some form. Now, you may have to take time to figure out how the node is set up (what icons mean what, and so forth) but that is only likely in someone's personal node, where they can go wild, not in a corp node where many people have to use this interface. The only thing that comes close to what he's talking about is the use of the Reality Filter program, which would affect your perception of the entire node at once.

[7]: Infinity Mirror: Just as the node can have unlimited numbers of icons, so can you. Unrated programs don't take up any space but they are still programs and still have icons. So you can have limitless numbers of Virtual Pets that happen to have the same icon as your persona, and any potential attacker will need to go through them simple action by simple action before they can attack "you."

Complete and utter crap, just like the the last point. In addition to all the inherent wrongness of his points, add in another point here. For every (System) number of programs you run simultaneously, your Response is reduced by 1. Try this trick, and you'll freeze your own system, to be nice easy pickings for spiders, rival hackers, or even the punk ganger who wants to buy some drugs with your SIN.

[8]: Agent Smith The backup can be in the form of a security spider, but it can also be other Agents that belong to Renraku. Since there is no limit to the number of targets in the CC of an email, Renraku can essentially oppose you with one Agent for every computer - in the world - that is in Renraku's hackastack. Which is arbitrarily large. So you could be opposed by a hundred, or a thousand Agents next turn. Seriously.

Seriously, does he ever get tired of being wrong? Yes, Renraku could swarm you with thousands of agents. No, it isn't going to happen. You want to know why? Because you're not the only joker trying to get into Renraku systems around the world. Some of those other agents are running defense against other hackers.

Second, you're seriously not that important. Even if you hacked the CEO of Renraku's personal account, they wouldn't send their entire matrix security force after you. Their best tracker would quietly track you down, and then a few nice men in Renraku Red Samurai armor would ask you to come in and answer some questions. Devoting all the matrix security to follow one person is a use of resources that will not go unnoticed, and that would be a sign of weakness that the other megacorps would pounce on.

[9]: While it takes several rounds to kick you out during which you can hijack the node and prevent it from doing so, changing the Access ID automatically works and boots you and every other foreign user off the system without rolling dice in a single action. I'm not sure this counts as an exploit of the rules, since it's actually given as a helpful tip in Unwired.

The poster is mistaken. What it says in Unwired is that you can change your own Access ID, and that will cause the node to drop you. Someone else trying to do that to you would have to hack you, as normal.

If the IC doesn't spot you and you didn't set off an alarm, skip this step altogether. You can get rid of the IC in several ways. They seem to think you want to engage it in Cyber Combat, which involves rolling initiative and then following the rules for casting low force manabolts at people with spell defense who also get to soak damage and have armor. What you'd actually want to do is to spoof a command to the IC to make it go away. This is just Hacking + Spoof vs. Firewall + Pilot. One net hit and the IC stops all harassing of you forever. If for some reason you are being bothered by one Agent, it's pretty much over in one action.

Clever, but only partially correct. If you manage to spoof the command, it will leave you alone, until something tells it to do otherwise, such as a Spider, or another alert, or any number of other things that could happen. Spoofing against IC is the equivalent of pointing over a guard's shoulder and yelling "Look out" and then running around the corner when he turns. It is NOT a permanent solution. Moreover, if the designer is halfway intelligent, a log will be kept of all commands given to the IC, especially during alerts, and spiders maybe forwarded a copy of all commands.

For no reason, Data Bombs have a unique damage mechanic where they inflict matrix boxes equal to Rating multiplied by the score shown on a single d6. Since Data Bombs normally run in ratings of 3-6, this means that you usually have a 2/3 chance of just crashing: no save.

Don't know what book he was reading, but it isn't SR4. Data Bombs inflict a number of boxes of Matrix Damage equal to their rating. Matrix Damage would be resisted as normal.


So seriously, don't believe everything you see on forums. Read the rules for yourself.

[schpeelah]

... Has somebody hijacked mean_liar's account or what? I'm just not seeing anybody having 500 posts here yet referring to Frank as 'this person'.

BRB, reading the rules.

[Quantumboost]

... Has somebody hijacked mean_liar's account or what? I'm just not seeing anybody having 500 posts here yet referring to Frank as 'this person'.

BRB, reading the rules.

Context looks like mean_liar's quoting someone else's comments on Frank's post from another forum. So not his words, but not hijacked.

[Username17]

Wonderful. Another "you got to believe in God to understand the proof of God" arguments. I think that everyone who suggests that looking for a way to break the rules invalidates your statements on how the rules can be broken should be beaten to death. That out of the way, let's bother responding to the "points."

---

The poster is slightly mistaken about the fact that you can be in many nodes at once. While this is correct, he neglects to mention that your persona can be attacked in each and every one of these nodes, and that you must split your dice pool between all of them to defend or fight back, and that all damage done to the persona applies to ALL the nodes you're in, not just the one. Moreover, you can only concentrate on one node at a time, so you can't hack from multiple points at once. So you can be in as many nodes as you want simultaneously. Its just stupid, and ineffective.

This is all completely fucking irrelevant. This was supposedly in reference to the Hackastack, but it seems to actually be about subscription limits. Or something. It's a complete nonsequitur. It's like it I made a point about how awesome it is to hire a fleet of truck drivers and you responded with a condescending put down about how difficult it is to get around on two skate boards.

Hackastack involves you having lots of commlinks and sending an agent from each one to attack one node. You aren't splitting or subscribing or doing fuck all. You log in to a commlink, you activate the agent, then you physically put that shit down on the fucking table, and you do it again. The first commlink doesn't go offline just because you aren't pushing buttons this round, nor does it go offline just because you can't push buttons this round because you left the house. That agent persists and continues to follow instructions. What that has to do with how many nodes a single persona can be in I have no idea.

The action to generate a false ID is in fact, mentioned in the core book, on page 225, under the heading "Spoofing the Datatrail", which the poster apparently never read. It requires a Hacking+Spoof(2) test to generate a false ID, or a Hardware+Logic(2) test to make a device generate a false ID.

Sigh. The poster is being a moron. Some more. He's talking about the action to hack the firmware to permanently change the Access ID of the commlink. That's on page 224, not 225. And the listed test is "Hacking + Software (2)" but that's not even important, because what we're actually talking about is running your Access ID through a third party (such as a heat sensor on a pull-tab burrito) or having your Persona ID mimic a new ID in software on the fly. Those actions aren't listed in the SR4A book, which was my original fucking point. There's just a pointer to Unwired, which is in the exact paragraph that dumbshit here was referencing (badly).

Nodes can drop you for not having an access ID, or for changing IDs while in the node. However, the only thing close to what this person is talking about would be the Limitation feature for some programs, mentioned on page 144 of Unwired, that prevents software from being used in certain ways. Unless he's talking about trying to log on to a system with a bad ID and password. And that's the same trying to log on to your email account with the password satrorwaduy when your password's really 12345. That's not hacking, that's being an idiot.

I can't parse this gibberish. At all. Nodes can allow only a set of Access IDs to run on them. In order to spoof yourself to have one of those Access IDs, you need to share a node with one of the icons that already has one of the accepted Access IDs. So if your system doesn't put any of its icons that are allowed into the restricted areas into non-restricted areas, you can't hack it.

I have no idea what the fuck plot-device simple passwords has to do with Dropout. It's completely meaningless. He doesn't understand dropout and likes to hear himself talk. That's the only explanation I can think of.

Denial of Service attacks do work. The tactic is nothing new, and hackers use it today, typically using specialized bots to do the attacks for them, usually using a worm or virus to plant the bots in a number of systems, to maximize effectiveness. However, for a System 5, Response 5 node (the least you'd expect for a megacorp) it would take at least 100 bots to freeze the node with a DDOS attack.

Gosh, I'd love to see the math on that one. Especially because I specifically laid out the hackastack DDOS plan and it doesn't make a fuckwhit of difference what the stats of the node are. You're deliberately failing to hack the system in order to drive up the alerts. Rules as printed, if you generate enough alerts, the node shuts down. If that's your goal, you win. By losing.

Given enough time, anyone can hack anything, unless it is on a stand-alone system, cut off from the matrix. However, getting into the node isn't the problem. You have to dodge IC, and Spiders. You have to check for alerts that get tripped when certain files are accessed.

The poster apparently has no idea what an admin account with no alerts tripped means. The rest of his stuff was just dribble about how the GM can make shit up to make things hard if he wants to. Not really worth going through since he admits first sentence that my point (that you can hack any conceivable node in ~5 hours without setting off an alert) is 100% true. Moving on.

First off, the page reference he gives is bogus. Second, the whole point is complete crap. You don't need to render the icons. They area already rendered in some form. Now, you may have to take time to figure out how the node is set up (what icons mean what, and so forth) but that is only likely in someone's personal node, where they can go wild, not in a corp node where many people have to use this interface. The only thing that comes close to what he's talking about is the use of the Reality Filter program, which would affect your perception of the entire node at once.

Poster does not understand the Infinity Mirror at all.

The rules say that you need to take a simple action to determine whether one icon is a Program, an Agent, a Persona, an Arrow, a Sprite, or whatever else. There is no limit to the number of icons you can have in a node. Therefore, there is no limit to the number of Simple Actions you may be called upon to spend figuring out what various icons are before you can interact with any particular thing in the node. That's the Infinity Mirror. You can add limitless numbers of icons and force opponents to go through them one by one at a Simple Action each before they know which one is what kind of thing. There is absolutely nothing stopping you from doing this.

Indeed, if you go by the actual rules, players are pretty much doing this by default, because technically they come in with a half dozen icons or more.

Complete and utter crap, just like the the last point. In addition to all the inherent wrongness of his points, add in another point here. For every (System) number of programs you run simultaneously, your Response is reduced by 1. Try this trick, and you'll freeze your own system, to be nice easy pickings for spiders, rival hackers, or even the punk ganger who wants to buy some drugs with your SIN.

Sigh. Unrated Programs do not count against program limits for purposes of slowdown. But they do have icons. This was actually expressly stated in the point that is apparently "utter crap" but again the poster doesn't even understand the Infinity Mirror, so the fact that he also did not read the description of Infinity Mirror in practice should not surprise.

Seriously, does he ever get tired of being wrong? Yes, Renraku could swarm you with thousands of agents. No, it isn't going to happen. You want to know why? Because you're not the only joker trying to get into Renraku systems around the world. Some of those other agents are running defense against other hackers.

Buh? Poster conceded the point in sentence two. Then he wandered off into a diatribe about how Renraku wouldn't use a hackastack. I should remind you, that the same poster just a few points back was suggesting that Renraku was going to send multiple agents and actual spiders on actual company time to go harass someone who logged into their system without setting off an Alert, so apparently the real reason that Renraku can't spare a few dozen of their hundreds of thousands of attack dogs to maul a reported intruder is that they have all of them quadruple teaming their actual employees.

What-fucking-ever.

Don't know what book he was reading, but it isn't SR4. Data Bombs inflict a number of boxes of Matrix Damage equal to their rating. Matrix Damage would be resisted as normal.

Huh? SR4A, page 233. Data Bombs inflict Rating x d6 boxes. Seriously, this is even in the changelog file for SR4A. Actually taking time out of his day to post that a clear page citation is "wrong" without having apparently ever read the page citation in question is frankly mystifying.

---

In conclusion, the poster you cross posted me to is a mouth breather who has no idea how the matrix rules work, has no idea how they don't work, and basically ad hocs everything in real play. Not that there is anything wrong per se about replacing the Matrix system with just using your own "intuition" about how computers work in the "real world" and "Johnny Mnemonic" but it's sort of weird to act like those are the actual rules. And it's downright insulting to be a total dick about it when other people don't use the magical teaparty solutions you made up yourself and never wrote down anywhere.

There is basic failures in reading comprehension as well as fundamental Theory of Mind problems with the poster you are cross posting. He has no good points at all. Which is stunning considering that you can find a number of true facts in the speeches of the most vapid and wicked ideologues in history. The poster fails to understand RAW, fails to understand basic descriptions of actions, and fails to understand that other people can't read his mind and don't care what he has to say on this or any other subject.

The linked poster is a wastrel. I only hope that he is thirteen, because if that is the case it is possible that he will eventually grow up and read a book through, address a direct point made by another person, or make a cogent argument for a reasoned position. If it turns out that he's an adult, I'm afraid it's already too late and he makes a damning case against the educational system of whatever nation or region he spent his formative years in.

-Username17

[schpeelah]

OK mean_liar, can you give us a link to the forum this was originally posted on?

[RandomCasualty2]

Second, you're seriously not that important. Even if you hacked the CEO of Renraku's personal account, they wouldn't send their entire matrix security force after you. Their best tracker would quietly track you down, and then a few nice men in Renraku Red Samurai armor would ask you to come in and answer some questions. Devoting all the matrix security to follow one person is a use of resources that will not go unnoticed, and that would be a sign of weakness that the other megacorps would pounce on.

I always wondered how tracking ever even did anything at all. I mean unless you KO the guy with blackout or kill him with black hammer, you really have no means of easily identifying where he is. The best you can do is triangulate his position within 50 meters, but that might as well be fucking useless. That's roughly half a football field, and when you consider 50 meters in a multistory building, the guy could be anywhere. You also don't know what the hacker looks like, so I mean... he could walk right past your tracker and the guy would have absolutely no idea.

Seriously, track is totally useless. It tells you a vague as fuck location, no data to find the guy and an access ID that's made up anyway. I mean seriously, there's no point in trying to block a trace, because a trace is meaningless.

About the only remotely useful thing I could find for tracing people is if it can tell you if the hacker is on-site.

[mean_liar]

Uhoh.

Yeah, it was off of RPOL.net.

http://www.rpol.net/display.cgi?gi=134&ti=4453


Before cross-posting I just responded to him with what I saw as the biggest problem with RAW Matrix rules, which is the essentially infinite icons you can toss into a node and how that means you're screwed. The rest, well... yeah. I caught his Hackastack DDOS problem right out but wasn't interested in making a point-by-point rebuttal and just thought I'd throw some meat to the wolves.

[Username17]

Seriously, track is totally useless.

Yeah. That's why I devoted absolutely no time, space, or words at all to spoofing traces. Because even if they trace you, who the fuck cares? Unless they do it after your hack so that they could conceivably track you to your home neighborhood and blow up your house or something, it just doesn't make any difference.

"That guy who is hacking our system with a handshake range of 50m is within 50m of our system!"
"You needed a piece of IC to tell you that?"

You need to generate a new fake Access ID for each and every hack attempt you do, because the Access IDs are going to be logged by literally hundreds of devices and you can't possibly hack all the records away. So the Access ID that you are leaving behind has to be fake and disposable. But people tracing you during the hack? Who gives a shit?

-Username17

[RandomCasualty2]

Yeah. That's why I devoted absolutely no time, space, or words at all to spoofing traces. Because even if they trace you, who the fuck cares? Unless they do it after your hack so that they could conceivably track you to your home neighborhood and blow up your house or something, it just doesn't make any difference.

Even then, unless you live in the suburbs or on a farm, there's no way you can even trace which person is involved in something like an apartment complex or even if you were hacking out of a wireless node at Starbucks. I mean even if they knew it was coming from Starbucks, there's just no way for them to tell which of the 12 guys sitting in there performed the hack.

I mean really, your only chance of detecting the identity of the guy is to hack the enemy commlink (which you can do just by analyze to get his access ID), and try to hack into some camera or cybereyes that you can use to get a look at him or his friends.

But you don't even need trace to do that.

[spasheridan]

Doesn't the observe in detail action allow you to determine what icons are in a system with a single action? Then you just need to render hidden icons or something?

[Username17]

Doesn't the observe in detail action allow you to determine what icons are in a system with a single action? Then you just need to render hidden icons or something?

Getting the list of icons is actually not that useful. It´s just a list. Finding out whether Purple Pony Number 12 is a data file or a Persona is still a Simple Action for just Purple Pony Number 12.

-Username17

[mean_liar]

Unwired actually talks about this a little more - it's a trivial change but it's not a Node's default state to hide these ID tags.

However, doing so is pretty damn easy.

The train wreck continues. On the upside, I know a lot more about Hacking:

http://www.rpol.net/display.cgi?gi=134&ti=4453

[Username17]

Yeah. Gotta love that guy jumping in with a simple false equivalence argument. The only way to make that train wreck die is to shoot lazers at the guy who thinks that reality filters bypass the need for perception tests.



We may need many lazers.

-Username17

[RandomCasualty2]

Yeah. Gotta love that guy jumping in with a simple false equivalence argument. The only way to make that train wreck die is to shoot lazers at the guy who thinks that reality filters bypass the need for perception tests.

Well, in a way, they kind of can. The reality filter lets you basically substitute your own sculpting for something. Now, normally that may not seem great, until you realize that you can totally sculpt shit to stand out (so long as it's not running a stealth program). So you make attack programs look like guns, any kind of agent look like agent smith and other persona icons look like normal people. Data files can actually look like file folders or whatever. That cuts down the rendering time significantly.

It could really save you a lot of trouble. Now it still doesn't beat a botnet army, but it works nicely to filter icons into types, which is well enough to really cut down on the infinity mirror, since you can instantly filter out program icons from personas, quite possibly even filtered by type of program.

[Username17]

But that's exactly the problem. At no time does the description of Reality Filter say that it gives you any information that would take an entire Simple Action for Analyze to give you. Reality Filter can sort the icons you've identified (for example: if in the original sculpting they are all uniform purple spheres or some shit), but there's no rule that it will actually tell you the type of an unidentified icon. To identify an icon's type, you need to drop at least one Simple Action into the Infinity Mirror.

-Username17

[RandomCasualty2]

But that's exactly the problem. At no time does the description of Reality Filter say that it gives you any information that would take an entire Simple Action for Analyze to give you. Reality Filter can sort the icons you've identified (for example: if in the original sculpting they are all uniform purple spheres or some shit), but there's no rule that it will actually tell you the type of an unidentified icon. To identify an icon's type, you need to drop at least one Simple Action into the Infinity Mirror.

Now keep in mind that an icon's type can totally be given away by the sculpting. Most agents and personas are probably going to look like creatures and most file icons are going to look like inanimate objects. And that's okay, I mean that's the whole purpose of sculpting in the first place. It's not cheating to look at the sculpting and say "this looks like a file and it probably is."

Reality filter just replaces that with your own sculpting. So if you wanted to turn your matrix sculpting into cops and robbers, where you were the robber, the data was bags of money and the IC were the cops, you could. That's entirely the point.

If the filter can't determine the basic type of an icon, then it's basically useless as far as modeling the sculpting since it doesn't know if any given icon should be a bag of money, a cop, a robber, a gun or whatever. Now I mean, if we're assuming the filter works at all, it has to be able to get a basic type information of an icon. Otherwise you can't set up a working metaphor for any node you might visit.

Presumably this doesn't work against things running stealth programs, but it should work for almost every other icon.

Now granted, it doesn't explicitly say anywhere that reality filter does this, but the fact that reality filter is creating a metaphor means that it pretty much needs to know that shit. Otherwise all reality filter can do is fill a node with the same icon. If it's cops and robbers, then you're a robber and every other single icon is a cop, which actually makes you worse off using a reality filter, because you can't get any information at all from the icons listed. You have to assume it can learn icon type, otherwise it's a total hindrance, because it turns even a well sculpted node into something totally confusing and unrecognizeable.

We really have to assume that reality filter can get basic information, otherwise it cannot even do what it's supposed to do.

[Username17]

We really have to assume that reality filter can get basic information, otherwise it cannot even do what it's supposed to do.

What the fuck? No. We really don't. As Mean_Liar already pointed out, a node has the option of giving that basic information away. It can also choose to not give that information away, or to lie. Reality Filter only gets that information free and correctly if the node has been sculpted to give that information away free and correctly.

In any "security node" we can safely assume that is not the case. Getting that basic information for a single icon in an uncooperative node is a Simple Action that requires a successful opposed test. Full. Fucking. Stop.

-Username17

[RandomCasualty2]

We really have to assume that reality filter can get basic information, otherwise it cannot even do what it's supposed to do.

What the fuck? No. We really don't. As Mean_Liar already pointed out, a node has the option of giving that basic information away. It can also choose to not give that information away, or to lie. Reality Filter only gets that information free and correctly if the node has been sculpted to give that information away free and correctly.

Well no, it's not really lying. It's just giving you a different picture than the one you'd expect. Sure I may assign Microsoft Word the icon for Doom 3, and it's confusing to the eyes but on a computer level I haven't really done any kind of deception. It's purely visual. The only way to really lie in the Matrix is using stealth programs. And there's no reason to believe that stealth doesn't work versus reality filter normally.

But for nonstealthed icons, there's no reason to believe that reality filter wouldn't work. Reality filter is like setting a desktop theme on windows. You're replacing the default one with your own. So it's saying that windows have red borders instead of blue or that your folders look like toasters instead of traditional folders. To even begin to do that stuff, it has to know the icon type. Presumably, this is the opposed roll that reality filter has to make against a system to work correctly. It has to figure out all the icon types from the node, so it makes a roll to gather that information.

Now I guess you could go and say that reality filter can't do what it's supposed to do, and it has to represent every icon with the same icon, but there's no reason to believe that. Nowhere does it say that the only way to get an icon's type is via analyze. That's just one method to do so, but it doesn't mean that analyze is explicitly role protected as the only program that can determine icon type. Analyze also determines access ID, but so does trace.

I don't see it as a big stretch to assume that reality filter can get icon types, given that it would need them to do what its supposed to do. You just can't substitute your own system sculpting without knowing icon types. To know that I'm supposed to paint message windows red and make them semi-transparent, I need to know what's a message window and what isn't. Otherwise the program can't do anything.

There's nothing in the matrix rules that expressly forbid reality filter from finding out that information on its own. You just have to assume that it's better at it than analyze is, which is a bit odd, but then analyze is better than trace for finding someone's access ID. So if you want to counter hack them, you're better off analyzing them and then going to hack their machine through the access ID. In much the same way, reality filter is just better at getting icon types than analyze is.

[schpeelah]

There's nothing in the rules that expressly forbid

This phrase is universally considered a bad argument in rules debates.

You are supposed to use Analyse to find out an Icons type. This info may be given away for free in publicly available nodes, but not in secure ones. Under your interpretation the Analyse Icon action has no right to exist, and if there is only thing sure about RAI, it's that rules presented are meant to be used.