Of course security is always relative, and you can never have enough to make anything 100% secure. But there are different sorts of risk. Matrix risk is generally pretty low, but over an extended period. Courier risk is very high, but only for a short period.

Remember, a well-funded team of corp hackers usually isn't employed to crack into a matrix site for the explicit reason that the corp funding the operation could potentially be traced and open them up to huge risks in the corporate court. The same reason that corp special forces don't go traipsing into office buildings fishing for files. The entire reason Shadowrunners exist is because the corps want SINless noncitizen disposable deniable assets that never have enough information to get anyone in trouble but themselves ("who do you work for?" "Mr Johnson, that's all I know"). In that world, the number of actual people actually well-funded and competent enough to crack a secured Matrix site is relatively low -- remember that a good deck/commlink and all the assorted programs and goodies (not to mention the necessary cyberware) are all godawfully expensive, and a high-rating agent/IC with high rating programs (all written in-house by wageslaves) are relatively cheap.

Meanwhile, the courier is at risk of falling under simple surveillance (drones/satellites/eyes), corporate espionage, loose lips and social engineering, bribery or blackmail, and ultimately bullets. Those are different sorts of risks, and there are surely corporate executives weighing those risks against each other every day in the Shadowrun universe.

What can't possibly be the case is that someone's successfully hacking into every secure matrix site every day, just because your PC hacker can do it. Otherwise matrix security and corporate practices would logically have to be so tight there'd be no game there. This gets back to the idea being that all of these matrix connections must therefore be necessary for some Shadowrun universe reason. So here are some reasons:

1) Physical security on the move is very difficult/dangerous. Due to the way that information is treated as currency, any scrap of paydata, any kidnapped wageslave family member, any vehicle tracking number bribed out of the GridGuide system monitors, and so on, might lead back to a covert courier's identity or whereabouts.

2) Corps are rightfully paranoid about espionage. Breaking into a facility is illegal and the corporate court will smack you for it. Coercing or bribing a scientist or a programmer or a wagemage into leaving a door cracked, or dropping an access card in a fountain, or revealing some little helpful detail no matter how small could knock over an entire physical security operation or compromise 'cold storage' data with no foolproof way to track what happened. Data on corporate matrix sites can be audited and monitored.

3) Matrix security is generally good enough to keep out all but the best and most well-equipped on a day to day basis, weighed by risk-vs-cost.

4) A distributed network is safer for data storage than a concentrated physical storage site, and remember data = money. Data isn't just stuff that people like to steal, the corps are actually making money from it, developing it, using it, selling it. Making everything off-grid data stores moved by courier would essentially make a corp the kooky old guy who stores his cash in coffee cans buried in his backyard, and occasionally digs one up and secretly moves it in the dead of night to his brother's farm (again, to bury it in the cow pasture), and back again. No, data has to be backed up, stored, protected, and yet remain easy to distribute.

5) Most data is remarkably useless to the Shadowrunner. A corporation might develop a software tweak to help an assembly line produce aglets 1% more efficiently, and need to move that to one of their automated factory facilities. Would anyone really care? No. Will that save them millions of nuyen? Yes. So the company needs a more efficient and cheaper way of moving that information than a physical courier with physical security possibly needing to move around the globe (process developed in Uzbekistan, needed in Bogota). Which means they'll need to open up a matrix connection, because hey they might as well pick up the thousands of megapulses of equipment and production logs, stock logs, personnel timesheets, and reports while they're at it. The data that exists just isn't steal-worthy stuff. The vast majority of it, in fact, has no street value.

6) Destroying data is almost as good as stealing it. That worthless data above? It's very valuable to the corp that uses it, but not so much to other people. If you can hire a team to go trash all the data storage you can cost that company tons of money while they try to recover what they can and re-ship those aglet manufacturing upgrades. This could set them back on aglet production at a critical time and allow your new Aglecorp subsidiary to win the contract producing aglets for Lone Star uniforms throughout the UCAS. And there goes a huge profit (and probably some executive suicides) all because the corp was too paranoid to drop a matrix line out to a factory of virtually no other consequence.

Simple surveillance is not going to bring you anything of use. Bribery is not going to be cheap. Surveillance satellite really aren't cheap either, and aren't going to bring you anything of use ("Are you telling me we spent a billion dollar on a satellite so you can brief us that there are cars with unindentified people inside driving from MCT lab in Bellevue to MCT downtown on a daily basis?").

I consider the number of actual people actually well-funded and competent enough to stop a car carrying one or two armed guards, seize the chip they're carrying before any of them has the chance to activate the erase function, and escape, is relatively low as well. Such team is actually quite likely to require a good hacker among them, and each street sam to pack as much nuyen in ware than the cost of a decent hacking suit of programs.
Meanwhile, a hacker can get into a node anywhere in the world, and doesn't need more than a few hours to do it.

Sure, we all know how to walk and how to grab, while most don't know what stack overflow and sequence prediction attack mean. Even Navy SEALs plan can be as simple as "open the door and look into the room". It may be easier to conceive how a physical security can be defeated. It doesn't mean it's easier to do.

Oddly enough, you used the plural practices but reads like there's only one. I am for one not saying there should be no Matrix connection and al data transfer should rely on courier. There are facilities who are going to be online because benefit vs. risk tends toward this solution. And I think there are a minority of facilities who are going to have offline network because benefit vs. risk goes the other way.

MCT computer security lab is likely to have an online node because they need a wide access to the Matrix, to communicate with the rest of the company, download bug and security reports from customers, upload security update as fast as possible, outsource icon designs to Indian coders, and order pizzas.

The Shiawase MIFD would have an online node, and a separate offline node because they need to maintain a list of their clandestine agents and their cover, and the slightest breach of security would have tremendous consequences. The management is unlikely to consider "only a few dozen of people in the world can hack it and it's unlikely to happen more than once every two or three years" a satisfying answer.

And I don't think that having some site with a Matrix security so tight is going to ruin the game. I would rather this could actually comes in hand quite often for the GM to tell the story the way he wants.

I think you guys are hitting on one of the biggest flaws of the SR4.5 matrix, which is that it's heavily based on an imperfect understanding of modern computing. All editions of SR have this flaw, but SR4.x is the worst, since they tried harder to mimic the real world while not knowing a whole lot about it.

Wireless decking is not a bad idea. I was doing it as far back as SR2. Having everything be interconnected isn't a bad idea either, although it leads to serious problems; SR1-2 dealt with this via system maps and one-way datalines, but Sr4.5 has no such thing. In short, I think they went for realism over playability, but accidentally failed at both.

This is totally not really important, but I felt that this item wasn't exactly true. I think most any actual corp data that isn't "publicly available" is valuable to those others who can get their hands on it. In your example, getting the "equipment and production logs, stock logs, personnel timesheets, and reports" for a rival company could prove invaluable. Gives a slight edge to the Aglecorp subsidary to know what the rival company is paying employees, which employees they might want to poach, trends in equipment or production logs (ramping up for a new product?), reports listing testing results, etc.

Even the software tweak you mention could be really valuable to competitors who steal it. They can do several things, including using the tweak themselves, holding it for ransom, or modifying the tweak and inserting it back so that the company loses millions instead of saving millions. If something saves millions, I would think it would be worth 10s (100s?) of thousands to have, at least.

I don't have a degree in business or finance or anything, so of course I could be way off base.

Wireless decking isn't an awful idea, but I think it creates a lot of problems for the game, especially with the system that's in place to handle it. Also if you have an unconnected network you shoot security in the foot -- security hackers would have to be on-site (also opening up the danger of access to their own commlinks just by physical proximity, meaning potential spoofing of security-level users), meaning a well-placed grenade is all you need to take down the most meaningful security response. And if a site is physically lost there's no way for additional security to remotely move in to retake it.

Just as physical security tends to have a backup (ARTs on standby to assault and retake a facility), matrix locations would have off-site backups ready to hack back in and secure the site, so even if a hacker totally destroyed all the local matrix security, they'd still be on a clock to get their job done and get out fast before the cavalry arrived, which would conceivably arrive in a matter of seconds.

Ultimately I find SR4 matrix sites too easy to hack and too difficult to design effectively, since there are a ton of decisions you have to make about each node (which leads to designing sites with as few nodes as possible), and it's almost like designing a system that's effective in one way makes it virtually undefendable from other angles.

This is why I kind of think they should backtrack a little and draw a hard distinction between a fixed site and a commlink -- hacking a commlink should be fast and furious, achievable in combat, and better integrated, something that's in the sole realm of the dedicated hacker. Hacking a fixed site should be a little more complicated and should honestly be something that an entire team could potentially get involved in, even if all they have is an off-the-rack commlink and a couple decent utility programs. Then you could really get back to these sites as 'virtual dungeons' full of their own traps, challenges, and monsters, where the hacker is the strongest character but no one is completely useless. That would solve the pizza problem.

You could accomplish this by giving the hacker an option to break open a hole in the system that allows other people to connect undetected (or rather detected based on the hacker's ability and not their own) and then mask the others' activity sort of like a mage provides Counterspelling. While the hacker in meatspace might have a few points in Pistols and a decent gun, the street samurai could likewise have a few points in hacking or cybercombat and a decent attack program, standing by to assist the hacker. The mage could even have a couple points in computer and a decent Analyze program and stand by as a sentry. The infiltrator might have a decent system and dedicate time to defeating traces to buy the group time so the hacker can ultimately disarm the data bombs and decrypt the files before covering the team's escape. If you block off hot-sim from non-cybered connections (illegally modified datajacks) you could get a mix of cold-sim, hot-sim, and AR members all operating at various levels, and the matrix could become an analogue for meatworld activities and combat, just with the roles all flipped around. Given an added amount of complexity to the treatment of the matrix, you could even make more room for individual non-hackers to specialize in some useful task while keeping the hacker as king of the hill -- he'd be the street samurai or mage of the matrix, while the street samurai gets to be street samurai of the street and the mage gets to be the mage of magic.

What I'm getting at is that most information has value but virtually no utility to anyone but the owner -- you could spend a great deal of money hiring a team to steal the data and then have to throw it through so much R&D to apply it yourself that you'd have been better off just ignoring it. Runner teams might grab it as an opportunity target, but most of the time they'd never know if they were getting something juicy that will sell, or the formula for New Coke. Hence why random paydata ('hey while we're in here..') usually has low payouts -- the corp has to take time to sift through it and see if it's actually useful, which all costs time and money. And why runs tend to target specific things (prototypes, schematics, etc) that corps have found out about through their various means (espionage, bribery, threats, data mining, etc) that runners don't tend to worry much about. A corp isn't going to pay a runner team a whole bunch of nuyen to steal something that might be vaguely useful if they're lucky.

Pretty much. They tried to mix their limited understanding of how computer networks and computer security work with their interpretations of what Mr. Mechanical Typewriter wrote and what they remember of TRON and ended up producing such a contradictory mess that the developer and the team that wrote the rules can't make it work even for a single painstakingly written example.

TRON was 31 years ago. Neuromancer was 28 years ago. It's time to move on. At this point every person who might possibly play 5th edition SR has a computer on the Internet. Why not start writing rules that make sense and appeal to these people instead of writing yet another game that is pitched at people whose formative moment was watching TRON in a movie theater in 1982?

I see it from a different angle - a corp might not specifically hire a team to steal something like payroll data, but they probably would be willing to pay some amount for it. It would be a matter of shopping it around, having your fixer / fence find the right buyer, etc. In your R&D example above, the stolen data doesn't have to actually produce anything, it can just be information on where the company is headed, what they are researching, where they are spending their money. If corp A is producing a wiz new widget that will dominate the market that needs parts X,Y, and Z, then competing corp B can work to corner the market on parts Y and Z to keep corp A from succeeding. Or something.

Any data that a company doesn't have to show publicly, will be behind some kind of matrix defenses. Maybe not very strong defenses, but something. I'm reminded of Snow Crash, where Hiro finds a few seconds of some video clip in a company's virtual trashbin, and lives off the proceeds for a month by selling it to a pop-culture website (ref needed).

I bet the New Coke formula was worth a lot, at one point. Maybe not now...

Anyway, I'm just arguing to argue, and feeling pedantic about it, so I'll stop

I actually disagree here. I think they should divorce it from modern computing entirely, and create a system based on sheer wow factor. If they went pure TRON/Neuromancer, they could come up with something very cool and playable, and not have to worry about if it seems realistic to a modern computer user. Sure, it'd be a lot of style over substance, but if the rules work and things look cool, I think a lot of players would overlook technical accuracy.

Part of the reason Shadowrun endures so was is that it was iconic. Shadowrun's setting influenced the entire cyberpunk movement, and the mechanics influenced the entire World of Darkness. (I'm not exaggerating in the slightest-- if you look at the original books, they have some of the same playtesters, and I believe Mark Rein Hagen was one of them.) Being iconic, people can overlook the fact that it's not 100% real-- no one complains about the lack of internet in Star Wars, for example. A balls-to-the-wall, neon cowboy matrix could be fun, easy to play, and avoid all concerns about being unrealistic. Shadowrun has never quite achieved this, but it's what it originally tried to do.

I agree with you, but I suspect only because I also grew up with the cool stories of SR and Gibson and so on. The idea of the "neon cowboy matrix" is cool to me, a (somewhat) older gamer. Perhaps younger generations won't find it such a cool concept, the same way that (I think) younger generations didn't have as many problems with all of the changes in Star Wars. It seems a difficult tightrope to walk, attracting new blood while still somewhat catering to old-timers.

It's not about technical accuracy, it's about intuitive understanding and creative usage. It's easier to understand how classical medieval equipment can be creatively used in a stereotypical dungeon than how vaguely defined superscience devices can in a physical reality hardly resembling our own.
So completely divorcing Matrix from RL computing is making it much poorer as a system, imo, on par with Astral (without the metaplane shenanigans).

Like I said, if they came up with something cool and playable, I think that'd go over better than realistic and believable. People can suspend disbelief more readily if the wow factor is high enough. Nobody seriously questions technology in Star Trek or Star Wars, because they're cool. There's no reason why Shadowrun can't do the same thing.

...but a bad interpretation of the rules, even as far back as the original 4th ed BBB

Like I said, the concept itself isn't bad. The execution left a lot to be desired, though. They really should have thought things through better.

But there is nothing stopping you from doing that in SR4A. You can indeed set up Hacks exactly like you describe, but, from what you indicate, you do not have the patience for this kind of scenario creation. If you design them effectively, they will provide the challenge you desire, but you indicated that that is too dificult and thus is a waste of time.

Or am I misunderstanding what you are saying here. I am truly curious and not trying to make any waves here.

I could get behind that entirely. Since I already completely ignore the disconnect between Game and Computer Knowledge anyways, this would probably be really, really cool. Unfortunately, I do not see that actually happening. I agree that it was their originating intent, but it has gone far by the wayside.

Real computers are awfully boring and don't lend themselves well to Shadowrun.




I really can't tell what you're responding to here.

Sorry... The post I quoted of yours started as such.



You then went on to describe what you wanted to see as a system Hack. My point is that you can do that in SR4A, if you want to do so, and I was confused by the above quote. You say that you find that matrix hacks are too easy, and you would prefer a more complex Network as a target to attack (to hack into) but you are not wanting to take the time and effort to provide such a network, because thay are too difficult to design effectively, because they require too many decisions that need to be made, and thus, the design principle is to use as few nodes as possible to design the network.

It seems like you want more complexity in your target hacks, but you are not willing to actually take the time to actually put it in place because it is difficult and requires too many decisions? I do not understand this logic, and would like to better understand it.

SR 4 hacking adds in the nontrivial steps of determining if nodes are wireless (wireless and connected or wireless and standalone) as well as determining security levels and associated system rights for each node. Furthermore you have to design nodes with agents and programs (as opposed to just IC), and keep in mind that there's a program limit on each node (and for quasi-realism's sake you have to leave some space for actual workers). That means each node becomes a lot of decisions for the GM, and since hacking tasks are aimed generally at one person in any group it seems like an awful lot of effort for not a huge payoff in terms of session utility.

To put it another way, designing a "good" matrix site would be a lot like designing a traps-and-puzzles dungeon for the party rogue in Fantasy RPG, and leaving everyone else standing around outside waiting.

Hmmmmm... Strange... I have not had the issues you talk about. As I have indicated... Our Networks are complex and they flow in tandem with the combat around us. Probably helps that our Primary Shadowrun GM has been running his Matrix the same way for at least the last 3 Editions, so only the mechanics have actually changed. I don't know.

I do know he has never been completely happy with the Matrix overall, but then, who has. But it does work well, if you put the time in.

In your previous post you talk about teh many tasks and design steps that are required, and you are right, it does require some thought. But I do think that the thought processes are not all that difficult. In my experience Matrix design is pretty easy, and is quite fun. Though I really like that kind of stuff, so I do spend a lot of time just fiddling with the little things. Maybe it is just me, then. Would not be the first time that my OCD for setting details blinds me to the issues that others have.

Thanks.

In theory. But the problem has been there is no there there in SR. Decking/hacking has not really worked well in any edition of SR. And based on those universes that I care about, people into SW or ST do in fact argue deeply over the tech. Midichlorians anyone?

It's a real problem if the supporting fluff isn't firmly supported by the rules. Players don't like it when they go to do something "cool" that makes sense from the fluff and find out that in the game mechanics it sucks, it isn't allowed or requires a two hour pizza run.

You need a strong model that the writers understand and can use, and then a line developer who can control the fluff and the crunch and make them mesh. It's a lot easier to understand computers when you are using the real world as a basis, just like it is easier for players to understand a car racing game where horsepower, tires and weight are important elements rather then one where the car color, day of the month and whether Virgo is rising are important elements.

The problem with using the real world as a basis is exactly what we have in SR4.5. They tried to model things off the WinXP user's manual, and look what we got. The system is not simple, nor intuitive, and not especially cool or playable.

The key is, as you say, consistency and playability. Somebody needs to design a matrix physics engine and stick to it. If reality disagrees with the system, ignore reality so long as it's playable, and screw reality if it's fun.

Not to bring this post back from the dead.

But with the mutual signal range rule (SR4a p 224), the hacker does have to be on site in order to hack if there is no public account and/or if its off the Matrix. Because that's the only way to probe the firewall in order to find an account to hack in to. And if a building is built as a Faraday cage (UW p63) there is no way to get access from the outside. And it'd make sense that highly sensitive data would be kept in a network that is only inside a Faraday cage building/room.

I don't think stay at home hacking is viable.

No.. he doesn't need mutual signal range to hack. The page you reference to access/login you need to subscribe whether it's direct connect (mutual range) or going through the matrix.

He can hack from farther than that, he just needs to use up a subscription to do so... (and I'd say this would also leave more in the access log to track him if noticed in time). Notice that big 'or' immediately afterwards about simply going through the network.

p235 hacking....
Must have mutual signal range or direct connect... *OR* spend a subscription through the matrix.

Well, you can put drills on your drones... Drill a small hole from the outside, extend the antenna of a repeater inside, now you have wireless inside. Or use a hole saw and send a series of microdrones inside. However anyplace nuts enough to set up a faraday cage will almost certainly also have RF sensors, and you'll trigger an alarm. And their network connection runs on a fiber backbone through the cage.

At which point you just tap the Fiber Backbone from outside, via old-style decking.

Sure, it's a point-to-point connection to the KE monitoring office, which is inside the KE regional operations center. That complex of huge massive built windowless reinforced concrete buildings where their regional matrix security team works out of, where their armed drones live and all the emergency response guys hang out waiting for something exciting to happen. What could possibly go wrong?

...nvm the fact that your teammates might not be so keen on taking all of the bullets, dodging corpsec, and trying not to get filled with hole by the numerous drone, while you, the decker, while you sit idly at home and chuckle over their struggle over a nice cup of soykaf; all just so they could drill a hole in the wall to poke a wire tap through?

I'd drag him out with us just so that he could suffer and directly fight against death like the rest of us. If I have to dodge Powerbolts from wagemages and try to tango with Wired up Red Samurai, EVERYONE has to.

If it's a point-to-point connection from somewhere to somewhere, what sits at the ends matters little because you're tapping in at the middle.

Tapping a secure fiber run that has continuously monitored traffic is not at all trivial.

You know, I'm not sure why every system needs to have an external connection. You just need a place to store your data while working on some experiment and can always take the data that needs to be reported, to a room with an external link, or if you're working on AI you definitely don't want them to escape, so wouldn't it make sense to have an entirely closed off system.

Or if the building has a telematics infrastructure (UW p62) this will cause a race against the clock where you are trying to hack the system as quickly as possible before the TI finds your commlinks in hidden mode. The hacker will have to make fake authorization for the team's commlinks or turn off the system while setting up some kind of false data so that the spider's don't notice anything wrong. So if a TI is set up inside a Faraday cage, I don't know if there would be enough time to drill a hole in the wall to get a wireless signal to the hacker before the runners are discovered.

It is as trivial as it always was in Previous Editions. *shrug*

Well, since they don't include any rules for it, sure.

Tapper drones in Unwired.

Trivial to tap a single fiber optic cable and splice a new fiber into it... or to use the drone as a wireless repeater for the traffic moving through the cable.

I may be drafted to run a SR game soon. If I do and have to support a decker/hacker, I'm running it like how it was done in the TV show Leverage. Hardison's job was to handle the cyber threats so the mundane threats could be addressed by the rest of the team. Sometimes he had to do the hack from on-site and others from his office.

He would take the time to disable security, disable traces on his connection, fight off security actions, etc. The thing is, everything he did stayed in the realm of realism instead of going into a virtual world with iconography, virtual rooms, or other forms of electronic dungeons.

If hacking/decking is going to be realistic and fun, then perhaps we should ditch the Tron style of the Matrix and go with something more mundane but so much easier to integrate and adjudicate (ala: Leverage).

Every time I've tried it the "Shadowrun" way, it's taken the focus from everyone else to just the hacker/decker.

If it's a point-to-point connection from somewhere to somewhere and one of the ends has wireless access, then the other end has wireless access too. Faraday cage be damned.

I think part of the problem here is how GM's run the matrix. Some of us <me> saw it as almost a throwback to 1st edition in a way. What you do is start with the various components of the Target area. What is the system trying to do? In one system for a marketing firm I had an 8 node system. The first node was a public node that did connect to the rest of the matrix, it's security was limited and was really designed for the public/customer to interact with the firm. Data that is stored there was basiscally correspondence and orders. It was also the out way point for the rest of the nodes which were a wired system. That node connected to the Orders processing node that was more secure and tracked orders placed and accounting for bills paid. That node connected to a couple of other nodes, which also connected to some others (including a UV host being run by the CEO to monitor some feral AI's).

Point being that as a GM, when doing a matrix challenge for the hacker/technomancer you need to decide the structure of the system they need to hack. This includes the purpose of the nodes they need to hack through, how the topography works, and the sculpting of the host (is it a medieval castle, sci-fi, a park, etc). Sure the GM could run it as one big node, if their lazy about it or don't want to focus on it too much.

Back OT:
1st issue I have with decking: Challenging the decker, it is really easy to make a system either too tough or a cake walk. Making a system that is enough of a challenge and is plausible within the confines of the game is hard. Yes you want to challenge the player, but really, the security at a stuffer shack would at best be a rating 3 (so why bother rolling for a decker/techno with ratings that are 5+ ?).

2nd Issue: Too many program types, I think there are a total of 50 differrent programs. About a dozen of which are some what useful most of the time , the rest are situational at best.

3rd Issue: Unwired tried to introduce RL computer hacking techniques (virus's and worm's and such as optional rules), they should abstract this and consider part of an attack program's rating.

4th Issue: Psychatropic IC, great concept, needs more crunch to deal with duration and getting rid of it. All it says in the book is go see a shrink.

Except that 4th explicitly got rid of that kind of mapping.

Not really, No. It is no longer Required, but it is not Gone.

For 90% of nodes, yeah you're right.

For stuff that has any bit of security (where the paydata/corporate security systems are) and no need to be out there in the open, you better believe it is still there and most likely a wired system. Now you no longer have slave nodes, data stores, CPU, and what not (they have all been subsumed into the generic nodes). Mapping out a node network is necessary--ok not necessary but advisable. What nodes are slaved to other nodes. What data should be comming and going between nodes on an automated basis. This gets back to the point about making decking challenging, but not a cake walk or impossible.


Again YMMV depending on GM. I'm just giving my opinion on it as a long time GM.

Also, in touching up on this thread: I agree, get rid of access levels. According to the rules, the GM can adjust what each level does in a specific nodes. I could adjust the Admin account so when someone is logged in as admin, a notification is sent to the security decker that this account is being used. It's 2:00 AM, and the legit admin is asleep, the security hacker hits the security protocols and tripps the alarm and reset the account logs immediataly (dropping the invading decker to a non legit/public user with no access to anything but the virtual lobby).

Another point is the agent smith affect. Once an alert is tripped, the security decker can show up, with a train of agents + his co-workers and their agents. This gets back to it being a challenge, but not impossible.

Having worked with fiber optics in building construction before, you can at least read the light pulses by something as simple as gently bending the fiber a bit. The coating on the fibers is thin enough that a slight bend will show light through it.

It will probably impact the signal strength, however, so it might be detectable that way.

Injecting your own signal into the fiber is more complex. However, even today fiber isn't a continuous unbroken line - there are junctions, signal boosters, etc. Tapping into the line at those points is likely a lot easier than attempting a mid-fiber splice.



-k

Almost surely. A splice is notoriously difficult and expensive (at least it is today).

Reminds me that when Verizon called my parents and offered them FiOs with free installation, I told them "make sure that includes laying the cable" because I knew that cable laying is expensive (on the order of $10 per foot) and that our house was a good 150 feet from the nearest anything (that's just to our FENCE, not counting how far they'd have to go to connect up with the main line).

Turned out that "free installation" did in fact cover laying the cable.

They laid 550 feet of cable.

Cable comes in rolls of 500 feet.

They had to splice it.

No idea how much it cost Verizon to install the FiOs for us, but we got the long end of the stick, whatever it was.
(Obviously yes, they got infrastructure out of it, but on a purely monetary cost thing, connecting our house was a steal for us)

The point is, though, unless you're using a system map of some sort, you face the same security going in no matter where you do it from. It's certainly possible to have every room linked to a chokepoint.

Yes, it is indeed possible. So?

So, without a system map, what's the advantage of jacking in from the inside?

System map is not necessary. As for my comment, I was actually commenting on the Chokepoint issue. Each node with a Chokepoint works for me.
Having said that, if it is an important system, I have a System map, along with notes.
Semi-Important Systems may or may not have a Map, dependent upon exactly how important it is.
If it is not important, there are generally no rolls involved.

See, we're back to RL logic for splicing. That's exactly what I was talking about.

Now, splicing's not something you can easily do in RL; but neither is decryption. As for the rules, a tapping drone's in Unwired, and it taps the fibre each time every time easily.

But only because these things still require the optical signal to be transduced into an electronic signal and back again. In SR they should have mastered optical switching, which makes tapping into the signal significantly harder.

That really doesn't answer the question. What's the advantage to jacking in from inside, if you don't have a system map? For that matter, since each node is technically independent of each other, is there any advantage to jacking in elsewhere at all?