MOCKERY'S NET PRIMER
an essay on the net and its architecture
by Amy Luther
An encrypted data packet, or a data squirt, is a transmission between two points within the Net. It travels on the same public lines that normal Net traffic uses, and goes through public LDLs if it needs to. It is considered to move at the same 5 spaces per turn rate as all the other traffic in the Net.
It is always in a format that the I-G Algorithms can translate into information, but whether that information is visible or not is up to the parties involved. Visible indications of data transmission can range from glowing points of neon to fractally rendered bubbles to laser lines to a fanciful convoy composed of virtual trucks. Transmissions can also be made hidden or invisible with the same Stealth programs used to conceal netrunners and ice, and as such can be discovered with programs like SeeYa or Speedtrap.
For purposes of error correction, transmissions are all considered to be two-way, but they are not usually visualized as a single line connecting both systems together (unless a trace is run, see below). Rather, as packets of information and confirmation notices are sent back and forth, they are translated into a visible image by the I-G Algorithms -- something like cars moving on an invisible highway.
Heavy-duty corporations, along with government and financial institutions, will have paid Internet massive world-breaking amounts of euro to have their own private LDLs for data and money transfers. Massive exchanges of information are therefore possible with no chance of outside interference (unless you hack into the fort and fiddle with the LDL itself), and it's extremely fast. Most data transfers can be performed entirely in-house, provided there are enough private LDLs available -- placed in switching stations, branch offices, or whatever.
What this amounts to is a separate LDL system for the daring (or foolhardy) netrunner -- if you can hack into Arasaka Night City and use their private LDL to get to the Arasaka datafort in Rio de Janeiro, and then leave the fort to do all your running, you avoid using the public LDL system entirely, thus avoiding all possibility of incurring a long-distance charge. You also ensure that anyone tracing will follow your trace straight back to the Arasaka datafort. At the very least, they'll be prevented from tracing you further, and they might even assume you're working for Arasaka.
Of course, private LDLs have a security value too, which should be assigned by the GM when the datafort is created. And, if you fail to scam the internal Arasaka LDL, you have a lot more to worry about than a big long-distance bill ...
NET ARCHITECTURE, OR HOW THE LITTLE GUY TRANSFERS DATA
Smaller companies, on the other hand, have to rely on sending data packets through the Net. Here's where the Demon/Daemon series of programs comes in really handy. Demons already incorporate the capacity to carry extra MU inside them, and though this was originally intended as a way to carry extra netrunning programs without taking up valuable deck space, there's no reason that capacity can't be used to carry raw data. Daemons (Chromebook 3, p. 58) are even better, because they're intelligent enough to defend themselves and carry out a specific set of preplanned instructions, should they be interfered with. Obviously, in order to use these programs for data transfer, you have to re-engineer them to possess the Movement, Endurance, Recognition, and Memory functions to allow them to move freely back and forth between systems.
It may seem like sending data this way places it in a great deal of jeopardy -- after all, anybody could come along and divert or delete the data being sent. However, you have to remember that the Net is literally filled to capacity with this traffic. Everything from private email to encrypted purchase orders to advertising circulars to singing telegrams is zooming everywhere at the speed of light, and it's practically impossible to catch a specific piece of data unless you A) know what it looks like, B) know where it came from, C) know its route, D) know its destination, and E) are willing to wait for a long time in a given spot to catch and misdirect it.
There is no way to know what a particular data squirt contains without catching it, and catching one is like trying to pick one neon tetra out of an entire tankful (or one grain of sand out of an entire beach, or one drop of water out of an entire waterfall, etc). Companies will also convoy their data, camoflauging it among useless junk, or even send company netrunners to ride herd. If a route is constantly compromised, savvy corporates will begin hiding nasty anti-personnel programs inside the squirts ... imagine catching what you think is a file on upcoming Eji swimsuit designs, only to discover a Liche hiding inside ...
Keep in mind that data packets are not the only items which might be traveling along the line; transmissions may be in an ice-guarded convoy, since pretty much any program with Movement functions could be sent along for protection (Pit Bulls, Hellhounds, and other assorted nasties come to mind). Also, data could be under the shield of a licensed data transport company and watched by real, live netrunners for protection, whose job is to make sure the encrypted transmissions get where they're supposed to go, on time and unadulterated. (NetEx and United Data Services, for instance -- I promise we'll put something up about them Real Soon Now (tm).)
Once netrunners have hacked into the ubiquitous datafort, rather than just raping the system and running away, they may be able to learn the decryption code the company uses for its data squirts, or the schedule for their most important transfers, or even the key used to identify what each squirt looks like while in transit (the blue sphere is Accounts Recievable, the black one is Advertising).
This opens up a whole new arena of netrunning. No longer are netrunners and AIs the only tenants of a mostly deserted grid, and no longer is the Net just an empty arena for netrunners to attack dataforts in. Now, Netspace is teeming with billions of data packets, flying at breakneck spead from system to system, some clamoring for attention, some sinister and silent as a stealth plane.
BE VEWY VEWY QUIET - WE'RE HUNTING WABBITS
If you're jacked in, you may see a packet leave a Data Fort, or you may see it whiz by you in transit, or you may see its destination, but you probably won't see all three unless the two transmitting systems are very close together in Netspace (within the 20 space vision limit specified in CP2020, p. 152).
If a netrunner sees a packet, he can run a trace on it by activating the Long Distance Link from the Menu, and rolling 1d10 + INT + Interface higher than the Trace Value of whatever LDLs the signal has passed through. If successful, the netrunner sees a glowing, green-blue line running from whatever he is looking at to the source of the trace. If it came through an LDL, he can follow the line back to the LDL and must scam it as normal in order to follow it through. Obviously, if the source is local, Trace Values do not come into account (unless the target is running an anti-trace program like George (CB3 74). Trace programs can give you a bonus to this roll; if your target's gone through a bunch of LDLs, even your 1d10 + INT + Interface may not high enough to follow. Again, this is a variation on the Trace function found in Rache Bartmoss, p. 145. It may not be suitable for all campaigns.
HE STUCK IN HIS THUMB...
Once the transmission route has been determined, the netrunner can position himself along the line and intercept or garble the packets passing by. This is called tapping the transmission and can be active or passive in nature.
A passive tap uses a copy utility program to quietly copy the information passing by. Unless the netrunner knows the code key or has a high-strength Intrusion program, he's going to be stuck with a bunch of encrypted files which may or may not be breakable and may or may not be booby-trapped (see below). This method of tapping is practically undetectable; GM's call as to what the transmitting parties must roll to find out what's going on.
An active tap is one in which the netrunner places himself in the path of the transmission and deliberately interfers with the packets as they travel by. A special Tap Utility program (STR varies, but a good level is 4 with MU 1) must be run in order for the netrunner to access the transmission. A roll may or may not be required depending on the GM's judgement. Creating false information to be inserted into a transmission must be made outside the Net, and requires a template to work on and a Programming roll of Difficulty 20+ (depending on what is to be substituted).
For example, if Weefleboy wants to interfere with a routine profits report between a branch office and its parent corporation, changing the 10,000eb reported to a mere 1,000eb, he has a whole host of things to do: He must find out when the report is to be transmitted, obtain a copy of a similar report so he knows what his forgery must resemble, create a false report, position himself on the transmission line, and at the proper point insert his duplicate into the transmission without either end being the wiser. Detecting this would depend on what the netrunner rolled, and would be GM's call, but the inherent toughness of active tapping means that both ends get a roll, and probably an easy one at that.
Corrupting information is a lot easier (and easier to detect) -- I'd figure you could run an Anti-IC program on a packet and blow it all to hell, or just roll Interface against a Diff of 15.
Packets are almost always encrypted. Now that you've caught one, how do you get into it?
Encrypted packets are considered to be File Locked, with the STR of the lock depending on the sensitivity of the information inside. They can be broken with Raffles, Wizard's Book, or other Intrusion programs designed to break File Locks. They may self-destruct if a cracking attempt fails (i.e., Wizard's Book can't break it on the first try), or they may contain booby traps like Cry Baby (Chromebook 3, p. 76).
The Net has more traffic than the biggest Mallplex on the planet, and it's open 24 hours a day. What better place to advertise your products and services than in the Net? Why stick with boring old animated billboards when you can have free-roaming, intelligent, full-sensory ads programmed to find a likely mark -- er, customer -- and bombard him with jingles until he pays up or jacks out? Companies could use datafort walls for a free advertising medium, or create anti-system programs that attack a runner and overlay his ICON with a popular slogan or commercial character. The possibilities are endless.
Rumors abound of systems connected to the Net that are powered, but empty. This doesn't happen, folks; any system which was abandoned would also be physically deactivated, and not left linked into the Net for any period of time. What does happen frequently is more like what Janice Grubb describes in Rache Bartmoss, p. 140:Entire heavily secured dataforts and, sometimes, city-grid scale systems patch in only briefly to upload and download compressed files. They are Very Difficult to spot and, like Alpha and the orbital workshacks, only connect to the Net at a set hour and for a very short period of time.
This type of pop-up operation, along with private LDL links (see above) is what some corporations and most military installations use to keep secret files safe. After all, if they're not intending to put their data out on the Net, why bother maintaining a constant connection, risking continual intrusion by enemy netrunners? It makes much more sense, security-wise, to minimize contact with the Net as a whole. Such systems would use carefully concealed land lines or, more likely, radio modems, so they'd be almost impossible to trace without careful observation.
This is a variation on the Trace function found in Rache Bartmoss, p. 145. While programs require a special Trace option to perform a trace, netrunners can run one simply by using the Menu. Activate the Long Distance Link from the Menu, and roll 1d10 + INT + Interface higher than the Trace Value of whatever LDLs the signal has passed through. If successful, the netrunner sees a glowing, green-blue line running from whatever he is looking at to the source of the trace. If it came through an LDL, he can follow the line back to the LDL and must scam it as normal in order to follow it through. Obviously, if the source is local, Trace Values do not come into account (unless the target is running an anti-trace program like George (Chromebook 3, p. 74).
Trace programs can give you a bonus to this roll; if your target's gone through a bunch of LDLs, even your 1d10 + INT + Interface may not high enough to follow.
AIS AND MULTIPLE CPUS
House Rule:AIs CAN have more than 7 CPUs. As in the Interfaces, there is no "fragmentation" involved when you hit the magic number of 7, despite what is stated in the main Cyberpunk book and in Rache Bartmoss, p. 145. There is no practical limit to the number of CPUs an AI may have; however, common-sense and Turing restrictions (see below) mandate that you don't make your AIs too huge, because they begin to pose a threat to the system and its operators.
The only way to move data throughout the Net is to copy it. AIs must work out of physical systems; they cannot move freely through the Net like ghosts or spirits, and they cannot "exist in the Net architecture." They must have an anchoring system to netrun out of, just like a human netrunner. In order to flee the system they are housed in, they must copy themselves to another system, or have their set of associated CPUs physically disconnected and moved somewhere. Some AIs may be reluctant to copy themselves to another system, because the "original" isn't getting free at all, merely a copy.
Human AIs are specialized AIs which originate from a detailed study of a living person. They are created by exhaustively studying the human model (hundreds of questionnaires, personality tests, IQ tests, reaction tests, and so forth and so on) and programming an artificially intelligent "copy." They are just like normal AIs in every respect except for their origin, and they can learn and expand into extra CPUs if they are permitted to. Most human AIs requrie CPUs equivalent to their original human INT.
Human AIs are not really human anymore. As Rache Bartmoss said, "He thought he was human, but in my opinion, he was a program which expertly emulated human thought. What a pitiful existence."
Some popular myths about human AIs:"You can become a human AI by dying in the Net."
Sorry, nothing doing. AIs can't free-float (see above), and technology has not yet advanced to the point where you can move your whole mind out of your brain and dump it in the Net. If you die in the Net, you're just plain dead. Adios."Soulkiller can suck your mind out and turn you into a human AI."
Nope. No program architecture is sophisticated enough to copy the entirety of a human brain and translate it into data. In any case, you can't "move" or "suck" someone's mind out; you can only copy it and erase the original, so the "original" you would be dead anyway, and not free to roam the Net."There is a 'Ghost Town' where human AIs congregate."
Again, because AI's can't freefloat and live in Net architecture without an anchoring system, the most likely truth to this rumor is a "Ghost Town" BBS, where human AIs can netrun and exchange information with each other. If there's an actual, physical, "Ghost Town" system collecting and harboring rogue AIs, Turing (see below) is going to be very interested in tracking it down ... Even the "Ghost Town" as a BBS would be extremely dangerous for the human AIs involved, because such a large congregation of rogues could be traced and killed more easily than each AI could individually.
Like Dixie Flatline in William Gibson's Neuromancer, ROM constructs are human AIs which have been hardwired into a particular computer. Since they are ROM, they revert to their original state whenever their power supply is cut off, forgetting everything they have "learned" in the meantime. They are more limited and more popular than standard human AIs, since they are easily controlled -- just shut the power down, and they immediately forget everything that made them dangerous. In rare cases, and with a lot of difficulty, they can be copied into a free system, in which case they become just like a regular human AI. Most ROM constructs are heavily encrypted and booby-trapped to prevent this from occuring (Impossible + roll to break copy protection without killing the AI).
Turing is the international regulatory agency charged with maintaining control of artificial intelligences. They are a subdivision of Interpol, and as such have nearly international jurisdiction. Their power is formally limited in the US; while they do have the expertise and limited authority to deal with AI problems, LEDiv NetSec makes sure that its own AI Division, and not Turing, takes care of major AI incidents within US borders. The best place to get information about Turing is to read the source: Neuromancer.
THE "ROGUE HUNTER" DIVISION OF NETWATCH (Rache Bartmoss, p. 14)
This informal division exists, but is seen by both Turing and LEDiv NetSec as a dangerously uninformed collection of amateurs interfering with things they don't fully understand. Amateur Rogue Hunting is illegal. All the real, professional "Rogue Hunters" already work for Turing or LEDiv NetSec, and they're not going to hand out permits and licenses to anybody they don't directly contract out.
Netwatch and freelance Rogue Hunters are occasionally hired by corporations intent on salvaging or covering up illegal AI experiments without alerting Turing and getting their entire datafort EMPed (and paying several million euro in fines besides). Of course, if such a Rogue Hunter gets caught by Turing, they're likely to be charged with aiding and abetting a rogue AI ...
Lastly, "Rogue Hunting" is much less prevalent than Rache Bartmoss claims, because AI's can't get "loose" and just roam through the Net at will (as explained above).