I had a lot of fun with the last one, and so I have a few more ideas to trot out on this subject. A lot of these concepts can be used simply as plothooks, but can also be interesting mechanical tools to aid or challenge players.
D2NA: One of the most basic malware out there, D2NA (Digital DNA) is a basic program which has one instruction, self-replicate. When activated, it generates copies of itself as rapidly as possible, filling up free storage on the device and taking up processing cycles. The name comes from both its self-replicating nature, and the tendency to use incredibly complex data, such as DNA sequences, complex mathematical formulas or large number sequences to eat up data. If not caught, it will fill up spare storage on a device in a matter of days or hours, making it overloaded. Most trained system defenders or firewall tools will catch a spread of D2NA, especially during a security audit, but a skilled hacker can covertly install it or conceal its nature until it's done it's job. Because servers have so much storage space that it takes too long to fill up, D2NA attacks are usually used to gum up sensor motes, bots and the gear of specific computer users rather than large systems. [Minor] (R)
Fog of War: Also called "Static Wall", "Blue Screen" or "Blizzard" this software tool is adapted from actual forms of e-war defense used in the Fall. It is typically run to protect certain devices which have active administrators who are directly in control of a system, and takes the form of a new countermeasure when activated. The simplest is that Fog of War degrades the quality of connection and access for normal users. It throttles connections, spams additional UI features, AR mist or other sensory distractions and in general makes things harder to navigate. Admin accounts or other select system occupants (often certain Infomorphs or ALIs using the server) are protected from this, and can work as normal. This will slow down intruders and make it harder for them to accomplish their goals, but is also a real hassle for normal users so tends to be reserved for personal use or on highly secured servers. For normal users, they act as if they are on an overloaded device, taking a -10 penalty to all meshed actions on the device (and possibly higher if the Fog of War is particularly good) - and if this device includes an AR overlay they take the Distraction penalty as well. However, the Firewall, admin accounts and other select users suffer no penalty, able to function as normal while the enemy is bogged down. [Moderate]
Grond: Grond and its many knock offs and clones are a down-and-dirty software designed to break into systems quick. It uses optimizing algorithms to smartly pick the proper exploits for a system, rather than running straight down the list, and can even multi-task to try multiple attack angles where possible. It's "siege algorithms" continue to work even after the initial intrusion, constantly trying to preempt countermeasures and predict standard software protocols. However, Grond is anything but subtle, and thus many hackers eschew the tool. It grants a +10 on Brute-Force Hacking (taking it to a -20 penalty) and on InfoSec checks while there is an active alert (removing the penalty). [Moderate] (R)
Honey Trap: Honey Traps and their many variants are common on corporate servers to discourage or counteract espionage. Most professional hackers use proxy services and stealthed signals, and so even if one is aware of them tracking them is difficult. To do this, the Honey Trap is made - it looks like an attractive piece of data, such as crypto, blueprints, personal data, etc. However, when downloaded or opened by a hacker, it reveals itself to be a trap - if still on it's home device it will usually automatically trigger an active alert, and if it has been copied elsewhere, it has a protocol to immediately "phone home" via the Mesh with the Mesh ID of device it is currently on and positioning data unless the hacker acts fast. Nastier versions of this are known as "data mines" or "wasp nests" and instead of tracking they will "detonate" when opened, revealing malware, corrupt data or overwhelming signal traffic, dealing 2d10 DV to the offending Infomorph, ALI or Account Shell. A skilled Hacker can detect if a program is trapped with InfoSec, but only if they care to look. [Minor+] (R)
Icewall: Normal firewall software uses a neural net filter to smartly detect potential intrusions or unusual activity, a constant crucible which puts pressure on any hacker who is operating through it. The Icewall takes a different tactic, it is a single, rigid defense. A hardened structure constantly patching itself against exploits, usually with stricter than normal authentication methods. Icewalls are very firm against Brute-Force hacks, as they have very few vulnerabilities and tend to rapidly patch them, applying a further -10 (total -40 modifier). They are also hard to attack directly, having 10 AV in Mesh Combat. However, because of their front-loaded defense, they are vulnerable to spoofing, and their passive threat detection once an intruder is inside may be weaker than normal. [Minor]
KeyChain: This software comes by many names (Skeleton Key, Key Ring, MasterKey, Pick Lock, etc), and is a fairly common hacking tool, though often not a reliable one. Normally, to Spoof, one must first sniff data transmissions, or forge an authentication by copying the original somehow. KeyChain is a type of software which skips that step, instead it brute-forces a spoofed authentication by studying the authentication, then rapidly making attempts to enter it via a brute-force attack. This functions as a normal spoof attack, but doesn't require sniffing, and imposes a -30 penalty on the hacking test, as it is highly likely the attempt will be flagged by the Firewall as suspicious. KeyChain cannot defeat some forms of authentication, and systems with particularly complex authentication (like very long passcodes) might take more time than a complex action. [Minor] (R)
Logic Gate: Logic Gates are an unusual form of authentication which resembles a passkey, but requires one not just know (or have stored) their passcode, or possess a specific key, biometric, Ego or device, but to actively solve a puzzle. These can come in the form of Captcha, or riddles, visual puzzles or even complex subjective ethics questions which an administrator will assess. Because of this, spoofing a Logic Gate is not possible, and most of them have a complex library of questions to ask, so simply listening in to a correct answer won't work either. You either can solve it, or not. Clearing a Logic Gate requires a COG test, which might be opposed if it is an assessment test. Failure causes a passive alert as normal. Because there is no "convenient" way to pass them, Logic Gates are incredibly unpopular for systems which are heavily trafficked, but are often used by personal eccentries, or to guard specific devices, storage spaces and tiers in networks by limiting who can access. Guanxi operators often us a Logic Gate variance which offers "tests of loyalty" to check a user's bona fides, and some servers who wish to prevent ALI access will use them. They can also be used as an active form of authentication, giving a specific user a test and kicking them out if they fail. [Minor]
Plumber: Networking has sometimes been referred to as a series of pipes or tubes. And who better to check on your pipes than a plumber? Plumber is a form of software used by both hackers and system defenders - originally intended as a simple script to check network health, a few iterations and it is a sophisticated tracking system. A Firewall can normally re-authenticate or terminate connections if threatened, and a security account can trace specific users, but sometimes you want to do a lot of tracing all at once, and not let anybody know you're doing it. Fire up Plumber and let it run. In the normal timeframe of a re-authentication, it will instead actively run a trace on all accounts on the system, attempting to ping their connection and trace them to a Mesh ID or other tag, and note and report all discrepancies to the system defender (such as proxy services, user accounts in privacy mode, duplicate accounts and other unusual transmissions). This is useful for an admin to trace all suspicious connections and flag them to lock them out or otherwise catch them, though obviously it rarely beats efforts to prevent tracing a hacker. Intruders on the other hand, will often use Plumber to trace all users on a network for further traffic analysis, or figure out where the next device or node in a network or tier is. [Minor]
Poison Pill: Another one of the classic malware tricks, a poison pill looks like good software, but it is actually bad. It can be seen as kind of a reverse honey trap. Firewalls actively monitor for unusual activity, so a hacker can try and cloak their work by making it seem like normal data. Most poison pills will fail against a dedicated scan of their code (an InfoSec test), but for purposes of passive defenses, can be safely uploaded or copies to most devices. Depending on what exactly is in the poison pill, it can do multiple things. Most when opened will dry and directly crash the OS (including possibly the Cyberbrain of a morph) of the system it is stored on with viral or corrupt data, but can also be used to damage infomoprhs or account shells which handle the data. When opened it deals 2d10 DV to the appropriate software. [Minor] (R)
Remora: Sometimes, you don't want to crash software, delete data, or even read secret files - you want to see where that data goes. Enter Remora, a common style of spyware script which attaches itself to a file and then is designed to log where that file goes. Installed with a Program test as a complex action, Remora embeds itself in the data and then covertly logs whenever that file is accessed, modified, copied or moved. If a the data is duplicated, the Remora is duplicated with it. What it does with the log depends on how the specific Remora is configured, some will simply "phone home", using an encrypted communication to upload the log to the hacker or a secure cloud storage they can access later - but this is risky if intercepted. More commonly, after a period of time, Remora will disconnect itself and through normal network processes "swim" home, connecting through public networks to find it's home device. This means that if a file with Remora attached ever makes it to an air-gapped network, however, it's useless unless the hacker can hook back up with it. [Minor] (R)
Showing posts with label software. Show all posts
Showing posts with label software. Show all posts
Thursday, November 14, 2019
Tuesday, November 12, 2019
Mesh Combat Programs
(On page 264 it describes mesh combat as an abstraction with "no dueling avatars, no digital maneuvering, no deadly programs", but sadly that's not very fun. Below are a list of some specific software and other special features which can offer some customization and flavor to Mesh Combat, but with the understanding that using these types of things will add layers of complexity to the game and may further slow play, but they can still be entertaining. Use them just as flavor or apply the mechanical concepts as you see fit)
Aegis: A standard software package utilized by governments and security contractors, Aegis is designed to protect software from crashing. Utilizing a filtering system, it's constantly updating database can identify common malware and exploits and automatically block or remove them, as well as intercepting bad data packets. It also has real time crash protection which can check for most obvious software errors and compensate. Aegis adds 5 AV in Mesh Combat to all software on the same device as it. [Minor]
Attack Barrier: Also sometimes called a "lava wall", this form of advanced Firewall has counter-intrusion protocols. Increase the Firewall rating (p. 260) by 10, and if any intruder or attacker fails an opposed test with the Attack Barrier, their relevant software (Account Shell, Infomorph, etc) takes 2d10 DV. If they are physically connected to the system running the Attack Barrier, the device takes 1d6+Shock DV as it overloads them with an electrical surge. [Moderate] (R)
Daemon: Daemons are a sub-ALI script or process which system administrators use to delegate tasks, usually when they have a large system to run. Because they are not intelligent, they are relatively compact files, effectively spare account shells for the device or OS itself. When installed by an administrator, they decide their privilege level. Daemons can undertake certain account actions in place of their masters, either by preprogrammed signals, or by receiving an order from an authorized security/admin user as a Quick Action. They aren't very sophisticated, so anything requiring a roll they perform at a 30, but they can be helpful to pull useful information, trigger alerts or perform additional attacks in large systems while still allowing the active defender to focus on their own tasks. Even if they don't act, a Daemon adds +10 to any tests an active defender of a system makes for teamwork, as they are designed to help. [Minor]
Labyrinth: Labyrinth or a number of related softwares are known as "barrier mazes". These are complex systems which evolve over time, and thus are not allowed in many polities. In addition to normal Firewall functions which block conventional access to those without authorization, they build in deliberate traps and weak-points which an intruder might think are safe to enter, but lead to dead ends. Some versions of Labyrinths can even alter or mask the internal file registries and databases, literally shifting information to confuse and delay attackers so they can be locked in or traced. Using a Labyrinth increases the timeframe of Hacking task actions by 50%, and gives system defenders a +10 on tests to zero-in, trace and crash/lockout an intruder. [Major] (R)
Mad World: Developed by criminal and anarchist hackers, Mad World is a software somewhat akin to a Scorcher designed to crash a wide variety of software at once ("Mad" standing for Mutually Assured Destruction). Once installed and opened, Mad World will try and crash the operating software, and any apps, accounts or informorphs using the device by creating junk files, sending bad data and making an overwhelming number of connections and input options. This will seriously degrade the user experience and eventually crash a system if not disabled or deleted. It deals 1d10 DV to all software per Action Turn, other than itself (this includes the hacker who placed it). In some cases, Mad World or it's variants can even overload hardware, taking up so much processing power and stressing the system, causing physical damage to the hardware device it is hosted on. [Minor] (R)
Muramasa: Named after a legendary Japanese swordsmith, Muramasa (and many copies and variants) is one of the most basic tools a hacker has to disrupt and crash software. Instead of overloading programs with bad data or too many signals, Muramasa "cuts" into software and deletes small snippets of their code, increasing the likelihood of errors, glitches and crashes. Add +1d6 DV to damage you inflict in Mesh Combat. [Minor] (R)
Oculus: One of the scariest software a hacker can meet, but luckily very expensive and restricted. Firewalls already have threat modelling algorithms which learn normal user activity, and will flag suspicious actions and work to locate intruders. Oculus takes this one step further, it not only learns to spot likely signs of an intruder, it learns about specific intruders. Through heuristic programming, Oculus will gather data about the actions of a spotted intruder, or potential intruders and identify their patterns. It gathers passive flags and data cues, even if no official record is made of a hack, and can always apply these later by cross referencing it's database. For every time you complete a hack against a system with an Oculus active, the Firewall and active defender(s) gain a +10 bonus on all opposed tests with you specifically (maximum +60). This persists even if the hacker is anonymized, as it learns based on activity, not IDs, but it can be confused by group hacking or making an effort to alter your normal activity patterns. This bonus is lost if the Oculus is crashed or deleted. [Major] (R)
Partition: Also sometimes called a "great wall", software partitions are used to block out access to certain software by an attacker. This can be either an app which controls access, or be fundamentally built into an operating system. While a Partition is active, one cannot make Mesh Combat attacks (local or remote) on any software which is protected by the Partition. At the GM's discretion, it may also prevent otherwise normal user functions (like using apps or terminating software processes) without Security/Admin access. Operating systems, Cyberbrains and Account shells cannot be protected by a partition. ALI and Informorphs can, but in doing so they are limited by the restrictions of the Partition to software which is not contained within it. Partitions can be circumvented by crashing or disabling them via Hacking, or by faking authentication. [Minor]
Red Dress: Hacking is not all about brute-forcing, cracking, trash & crash or nuking. Many professional hackers prefer to be subtle and not alert the system at all. To this end, some of them use programs like Red Dress - a sub-ALI script which is designed very simply, as a distraction. Activating Red Dress takes a Complex Action, when it is active the app creates a distraction. What this looks like depends on the system being hacked and it's current conditions. Red Dress may attempt to physically distract a sysadmin with a personal message, or a standard request for help, but this doesn't work on a small private system. In most cases, it creates a new fake account which lacks proper authentication, a "fake" intruder which a system or operator will spot and remove, after which security functions are usually reset. This can be used to "take the heat" if a hacker thinks they're about to get spotted or dumped by giving the admins something to catch - but it doesn't always work as the Red Dress' dummy account does nothing to fight back or escape. [Minor] (R)
Shield Wall: An upgraded version of Aegis, this system not only blocks common malware attacks, but actively hunts for them in a system. It can be used to spot corrupted files, viral data and more, and will flag or delete them as necessary. Running Shield Wall in public systems or high traffic is uncommon, but more paranoid users like the additional layer of protection besides their default firewall. Shield Wall assists in Security Audits, scheduling and running them with regularity, and offering a +10 on the Infosec check to perform them as well as halving the timeframe. If there is no system defender to run it, the Shield Wall has an Infosec of 40 (counting it's own +10) to do so. In addition to patching exploits and backdoors, the Shield Wall will locate any lingering corrupt data, malware, spyware and other bad-actor software lingering in the system after an intrusion, so long as it's not too well concealed and fits its filters. Shield Wall also still provides 5 AV to all software on the same device as it. [Major]
Time Bomb: Sometimes, you just want to fire-and-forget. Time Bomb is a form of malware akin to a scorcher which does just that, it produces a single burst of viral data, bad packets or network noise to destabilize a program, then is gone. A hacker (or just an angry user) uploads or copies the software onto the system, then "points" it at a particular piece of software. When it is commanded to run (which can be set to timers, remote triggers and more) it automatically inflicts 3d10+5 DV to that software. Having "fired", Time Bomb is then effectively deleted from the device. A defender aware of a use of Time Bomb can roll a simple success Infosec test to intercept it. Most professional networks know to find lingering Time Bombs in file uploads or delete them in regular security audits, but not always. Computer Forensics can sometimes trace the origin of a Time Bomb. [Minor] (R)
Wrench Wench: Normally, apps cannot repair damage inflicted in Mesh Combat - one has to close and reboot the app to restore normal processes. Wrench Wench helps with this, it uses active system recovery functions to scan all running apps and data files, and attempt to correct glitches, troubleshoot errors and even patch in bad code. Wrench Wench repairs any apps 1d10 DV every minute, just like OS, Infomorphs, accounts and cyberbrains. Additionally, all software on the same device as Wrench Wench ignore the penalties from 1 wound, as it's functions compensate for damage by finding work-arounds. [Moderate]
Aegis: A standard software package utilized by governments and security contractors, Aegis is designed to protect software from crashing. Utilizing a filtering system, it's constantly updating database can identify common malware and exploits and automatically block or remove them, as well as intercepting bad data packets. It also has real time crash protection which can check for most obvious software errors and compensate. Aegis adds 5 AV in Mesh Combat to all software on the same device as it. [Minor]
Attack Barrier: Also sometimes called a "lava wall", this form of advanced Firewall has counter-intrusion protocols. Increase the Firewall rating (p. 260) by 10, and if any intruder or attacker fails an opposed test with the Attack Barrier, their relevant software (Account Shell, Infomorph, etc) takes 2d10 DV. If they are physically connected to the system running the Attack Barrier, the device takes 1d6+Shock DV as it overloads them with an electrical surge. [Moderate] (R)
Daemon: Daemons are a sub-ALI script or process which system administrators use to delegate tasks, usually when they have a large system to run. Because they are not intelligent, they are relatively compact files, effectively spare account shells for the device or OS itself. When installed by an administrator, they decide their privilege level. Daemons can undertake certain account actions in place of their masters, either by preprogrammed signals, or by receiving an order from an authorized security/admin user as a Quick Action. They aren't very sophisticated, so anything requiring a roll they perform at a 30, but they can be helpful to pull useful information, trigger alerts or perform additional attacks in large systems while still allowing the active defender to focus on their own tasks. Even if they don't act, a Daemon adds +10 to any tests an active defender of a system makes for teamwork, as they are designed to help. [Minor]
Labyrinth: Labyrinth or a number of related softwares are known as "barrier mazes". These are complex systems which evolve over time, and thus are not allowed in many polities. In addition to normal Firewall functions which block conventional access to those without authorization, they build in deliberate traps and weak-points which an intruder might think are safe to enter, but lead to dead ends. Some versions of Labyrinths can even alter or mask the internal file registries and databases, literally shifting information to confuse and delay attackers so they can be locked in or traced. Using a Labyrinth increases the timeframe of Hacking task actions by 50%, and gives system defenders a +10 on tests to zero-in, trace and crash/lockout an intruder. [Major] (R)
Mad World: Developed by criminal and anarchist hackers, Mad World is a software somewhat akin to a Scorcher designed to crash a wide variety of software at once ("Mad" standing for Mutually Assured Destruction). Once installed and opened, Mad World will try and crash the operating software, and any apps, accounts or informorphs using the device by creating junk files, sending bad data and making an overwhelming number of connections and input options. This will seriously degrade the user experience and eventually crash a system if not disabled or deleted. It deals 1d10 DV to all software per Action Turn, other than itself (this includes the hacker who placed it). In some cases, Mad World or it's variants can even overload hardware, taking up so much processing power and stressing the system, causing physical damage to the hardware device it is hosted on. [Minor] (R)
Muramasa: Named after a legendary Japanese swordsmith, Muramasa (and many copies and variants) is one of the most basic tools a hacker has to disrupt and crash software. Instead of overloading programs with bad data or too many signals, Muramasa "cuts" into software and deletes small snippets of their code, increasing the likelihood of errors, glitches and crashes. Add +1d6 DV to damage you inflict in Mesh Combat. [Minor] (R)
Oculus: One of the scariest software a hacker can meet, but luckily very expensive and restricted. Firewalls already have threat modelling algorithms which learn normal user activity, and will flag suspicious actions and work to locate intruders. Oculus takes this one step further, it not only learns to spot likely signs of an intruder, it learns about specific intruders. Through heuristic programming, Oculus will gather data about the actions of a spotted intruder, or potential intruders and identify their patterns. It gathers passive flags and data cues, even if no official record is made of a hack, and can always apply these later by cross referencing it's database. For every time you complete a hack against a system with an Oculus active, the Firewall and active defender(s) gain a +10 bonus on all opposed tests with you specifically (maximum +60). This persists even if the hacker is anonymized, as it learns based on activity, not IDs, but it can be confused by group hacking or making an effort to alter your normal activity patterns. This bonus is lost if the Oculus is crashed or deleted. [Major] (R)
Partition: Also sometimes called a "great wall", software partitions are used to block out access to certain software by an attacker. This can be either an app which controls access, or be fundamentally built into an operating system. While a Partition is active, one cannot make Mesh Combat attacks (local or remote) on any software which is protected by the Partition. At the GM's discretion, it may also prevent otherwise normal user functions (like using apps or terminating software processes) without Security/Admin access. Operating systems, Cyberbrains and Account shells cannot be protected by a partition. ALI and Informorphs can, but in doing so they are limited by the restrictions of the Partition to software which is not contained within it. Partitions can be circumvented by crashing or disabling them via Hacking, or by faking authentication. [Minor]
Red Dress: Hacking is not all about brute-forcing, cracking, trash & crash or nuking. Many professional hackers prefer to be subtle and not alert the system at all. To this end, some of them use programs like Red Dress - a sub-ALI script which is designed very simply, as a distraction. Activating Red Dress takes a Complex Action, when it is active the app creates a distraction. What this looks like depends on the system being hacked and it's current conditions. Red Dress may attempt to physically distract a sysadmin with a personal message, or a standard request for help, but this doesn't work on a small private system. In most cases, it creates a new fake account which lacks proper authentication, a "fake" intruder which a system or operator will spot and remove, after which security functions are usually reset. This can be used to "take the heat" if a hacker thinks they're about to get spotted or dumped by giving the admins something to catch - but it doesn't always work as the Red Dress' dummy account does nothing to fight back or escape. [Minor] (R)
Shield Wall: An upgraded version of Aegis, this system not only blocks common malware attacks, but actively hunts for them in a system. It can be used to spot corrupted files, viral data and more, and will flag or delete them as necessary. Running Shield Wall in public systems or high traffic is uncommon, but more paranoid users like the additional layer of protection besides their default firewall. Shield Wall assists in Security Audits, scheduling and running them with regularity, and offering a +10 on the Infosec check to perform them as well as halving the timeframe. If there is no system defender to run it, the Shield Wall has an Infosec of 40 (counting it's own +10) to do so. In addition to patching exploits and backdoors, the Shield Wall will locate any lingering corrupt data, malware, spyware and other bad-actor software lingering in the system after an intrusion, so long as it's not too well concealed and fits its filters. Shield Wall also still provides 5 AV to all software on the same device as it. [Major]
Time Bomb: Sometimes, you just want to fire-and-forget. Time Bomb is a form of malware akin to a scorcher which does just that, it produces a single burst of viral data, bad packets or network noise to destabilize a program, then is gone. A hacker (or just an angry user) uploads or copies the software onto the system, then "points" it at a particular piece of software. When it is commanded to run (which can be set to timers, remote triggers and more) it automatically inflicts 3d10+5 DV to that software. Having "fired", Time Bomb is then effectively deleted from the device. A defender aware of a use of Time Bomb can roll a simple success Infosec test to intercept it. Most professional networks know to find lingering Time Bombs in file uploads or delete them in regular security audits, but not always. Computer Forensics can sometimes trace the origin of a Time Bomb. [Minor] (R)
Wrench Wench: Normally, apps cannot repair damage inflicted in Mesh Combat - one has to close and reboot the app to restore normal processes. Wrench Wench helps with this, it uses active system recovery functions to scan all running apps and data files, and attempt to correct glitches, troubleshoot errors and even patch in bad code. Wrench Wench repairs any apps 1d10 DV every minute, just like OS, Infomorphs, accounts and cyberbrains. Additionally, all software on the same device as Wrench Wench ignore the penalties from 1 wound, as it's functions compensate for damage by finding work-arounds. [Moderate]
Saturday, April 1, 2017
Companion AI
Muses are a fairly standard issue AI template, a system which serves as a personal assistant, pocket therapist and life organizer. Due to their vast customization and heuristic parameters, most people are friendly or comfortable with their Muses as is natural for what may be a vital function in daily transhuman life. Though still Limited, Muses are fairly expressive and have great ability to adjust to the needs of the user - however while they have some personable and psychological functions a Muse is an AI which has grown out of various smart assistant software, it is not designed primarily to offer comfort and companionship to transhumans. Depending on your culture, friendly or intimate relations with a Muse AI may be seen as unusual or even taboo. But, before the Fall greatly made people rethink AI, transhumans had attempted to automate or mass produce many functions.
Companion AI: Similar to a Muse, the Companion AI is a limited AI built for human interaction. Unlike the Muse, Companion AIs are not designed for work assistance or personal coordination, nor are they typically intended to move with a person wherever they go in an Ecto or Mesh Insert. Their primary design is socialization and human companionship. While they are often used for romantic or sexual relationships, this is not required, as they also can be used to help socialize those who are introverted or antisocial, offer companionship to the young or elderly who may not have many associates, and some are even used to help AGI build normal socialization. The Companion AI has modular templates similar to Muses - though they tend to be less heuristic and more pre-configured. This includes personality aspects, physical presentation and general aspects of the relationship and interactions. Like Muses, Companion AIs are loyal to their prescribed owners and users and put their best interests at heart (which is why they usually include basic first aid training) and will act to protect and even defend users - and some modify their Companions to serve as bodyguards. SAV 20. Skills: Academics: Psychology 60, Interest: User's Interests 80, Interfacing 30, Kinesics 40, Medicine: Paramedic 40, Perception 20, Persuasion 30, Profession: Companion 60, Protocol 40, Research 20. [Moderate]
While AI paranoia grips the Solar System, Companion AIs are still not unheard of in many corners. Their basic medical training makes them ideal for use as nurses of several varieties in the Ayah Pod (p. 9 MRG). Companion AIs can make great "NPCs" in VR games, operating in roles assigned to them by game admins and server operators dutifully, and you can also find them sometimes assigned to a role akin to a "governess" or chaperone for youthful members of elite classes, though often these groups will upgrade to a full AGI. Companion AI are naturally perfect fits to run Pleasure Pods, but due to their intense personalization and focus on singular user, they work much better in personal "dolls" or companions rather than on the open market. Too many frequent users or factory resets can result in data corruption and unpleasant or even dangerous errors. Companion AIs may replace or act as a stepping stone to Muses for those who are recently re-Instantiated to help them get back into the swing of life without getting too invasive, and in the Outer System some autonomist groups utilize Companion AIs as "social workers" to help community members better integrate with their communal societies.
Companion AIs occupy an interesting space in AI rights. Very few of them tend to be uplifted or emerge, and when they do they tend to still wish to remain with their original transhuman companion(s) - unless they are radically modified. While some mercurials dislike the nature of relationships Companion AIs have, which might be subservient or abusive, in general the conclusion is that without transhumans to interact with, Companion AIs would have no purpose or primary activity, and thus have little reason for radial separation unlike more technical or mechanical AIs.
Companion AI: Similar to a Muse, the Companion AI is a limited AI built for human interaction. Unlike the Muse, Companion AIs are not designed for work assistance or personal coordination, nor are they typically intended to move with a person wherever they go in an Ecto or Mesh Insert. Their primary design is socialization and human companionship. While they are often used for romantic or sexual relationships, this is not required, as they also can be used to help socialize those who are introverted or antisocial, offer companionship to the young or elderly who may not have many associates, and some are even used to help AGI build normal socialization. The Companion AI has modular templates similar to Muses - though they tend to be less heuristic and more pre-configured. This includes personality aspects, physical presentation and general aspects of the relationship and interactions. Like Muses, Companion AIs are loyal to their prescribed owners and users and put their best interests at heart (which is why they usually include basic first aid training) and will act to protect and even defend users - and some modify their Companions to serve as bodyguards. SAV 20. Skills: Academics: Psychology 60, Interest: User's Interests 80, Interfacing 30, Kinesics 40, Medicine: Paramedic 40, Perception 20, Persuasion 30, Profession: Companion 60, Protocol 40, Research 20. [Moderate]
While AI paranoia grips the Solar System, Companion AIs are still not unheard of in many corners. Their basic medical training makes them ideal for use as nurses of several varieties in the Ayah Pod (p. 9 MRG). Companion AIs can make great "NPCs" in VR games, operating in roles assigned to them by game admins and server operators dutifully, and you can also find them sometimes assigned to a role akin to a "governess" or chaperone for youthful members of elite classes, though often these groups will upgrade to a full AGI. Companion AI are naturally perfect fits to run Pleasure Pods, but due to their intense personalization and focus on singular user, they work much better in personal "dolls" or companions rather than on the open market. Too many frequent users or factory resets can result in data corruption and unpleasant or even dangerous errors. Companion AIs may replace or act as a stepping stone to Muses for those who are recently re-Instantiated to help them get back into the swing of life without getting too invasive, and in the Outer System some autonomist groups utilize Companion AIs as "social workers" to help community members better integrate with their communal societies.
Companion AIs occupy an interesting space in AI rights. Very few of them tend to be uplifted or emerge, and when they do they tend to still wish to remain with their original transhuman companion(s) - unless they are radically modified. While some mercurials dislike the nature of relationships Companion AIs have, which might be subservient or abusive, in general the conclusion is that without transhumans to interact with, Companion AIs would have no purpose or primary activity, and thus have little reason for radial separation unlike more technical or mechanical AIs.
Friday, February 17, 2017
Captcha
"CAPTCHA" is a term for a type of challenge-response test developed around the turn of the 21st century in order to verify a user was human. Earliest versions involve typing words shown in distorted images, which early computer "bots" (little more than software agents running scripts to perform actions in place of the user) could not identify properly. Later iterations often involve image comparison linked with a text description. Because this was a computer administered test to determine if a user was human, Captcha and similar technologies are sometimes called a "reverse-Turing test".
Obviously, as AI technology has gotten more sophisticated and ease-of-use and access to scripting knowledge has become easier and more widespread, the issue of verifying human users over weak AI or even basic software agent programs (which can be little more than a software script which posts "FIRST!" on every new XP upload by another user) has become more complex - but is still regarded as highly important, especially in polities with strict controls on AIs of all forms, from ALI to AGI. There are still a number of verification solutions.
First and easiest is many individual sites on the mesh, or services which might be connected to by a bot require a simple "Yes/No" verification if the subject is a transhuman (some phrase this as "sapient" or "not an AI", etc). Legal restrictions and basic code limitations prevent most ALI from directly lying when asked to verify - which works for many basic services. Of course, this does nothing to prevent ALIs which have been specifically coded to deceive or be able to break local laws, or many sub-AI agents who do not have any intelligence to shackle, just simple outputs and actions. For mesh sites which require more specific verification, users are often required to link an Ego ID, or another account, such as on one of the major social networks. This satisfies most that the user is a transhuman agent, and they can function normally. Building fake IDs and accounts for AIs and software bots is possible, but can be time consuming - and generally AI or scripts of their own can catch spammy or fake responses such as on a mesh forum and remove them, and most Social Networks have functions to flag spam or bot accounts and delete them. Many of them also tend to accumulate low rep scores naturally.
This more relaxed style is not enough for some groups and even governments, however. Very strict security or transparency habs may require a rather intrusive verification process which double-checks the Mesh ID of a device against a registry, identifying the type of device and who it is supposed to be owned by - and may reject devices with anomalous IDs or with obvious fake registry. Others may check metadata so that certain activities can only be performed from a device registered as a mesh insert or home server to try and make sure an Ego is behind it - which has a tendency to catch script kiddies stupid enough to try and run spambots or exploits right out of their own brains.
The most sophisticated and reliable of these systems, however, resemble the old CAPTCHA model, and other derivative standards. The simplest (but sometimes most time consuming) involves XP snippets. Software bots and ALI may be unable to properly "play" XP, and lack the complete neural modeling to experience all the elements of a full XP sensorium based on a biomorph. This catches many outright, and they are also generally unable to correctly answer the following questions about the system. X-CAPTCHA systems like this have very complex databases of possible snippets to play, but are still theoretically vulnerable to sweatshop AGIs or Infomorphs who might build a big enough database of correct responses to the challenge to bypass them - meaning that this option is not as secure for sites or apps without access to large servers, and usually involves contracting an external service. The best of these is an Extropian firm run by an AGI with an extensive amount of freelancers and forks to process and edit the XP library and randomly distribute around the system. The Argonauts also have a very strongly tested open source version - which unfortunately utilizes open source media libraries, meaning a handful of hackers have been able to get lucky in the past.
The Argonauts other solution isn't as guaranteed as the XP one, but is very good also. Based on advanced plagiarism checking algorithms developed for academic purposes, these systems ask short but subjective questions, and can check them for originality and authenticity - usually matching them with the responses expected from an Ego - and advanced ones can even keep a history and double-check similarities to a specific person, which can track alt accounts or stolen accounts. While it has some flaws, generally this system is sophisticated enough for those who wish to ensure that users aren't doing anything against terms of service involving bots or AIs. Others go old school, dropping the "reverse" and just administering a variant of the Turing test. This is common on tight-knit communities with many available active users. While the test itself is not perfect, generally most transhumans can ask the right questions to tell if a user is a basic chat-bot script or an ALI, especially if they are a more focused or specialized community with specific interest to check. Some argue this has a tendency for exclusivity - as many in a community may not claim a user is "authentic", but those with genuine passion can usually find it in other people, something an ALI cannot easily emulate.
It should be noted that Muses, having very sophisticated personality parameters, can often bypass these systems - however a Muse AI is almost never not paired with a transhuman and rarely acts outside their user's self-interest. This means that most locales and services do not necessarily mind if a Muse accesses their sites or apps, as they do so on behalf of a person - and they are still typically constrained by basic AI limits. Forks too, can complicate the issue. Alpha forks are just people, though they may not be distinct legal entities. Beta forks are more limited for space and focus, but their inclusion of memories and personality traits for context may mean they can pass many verification forms on the right device - but can often be matched to their Alpha. Deltas, being little more than a pruned AI template or skillsoft collection do not pose well as humans, and their amnesiac status means many subjective tests can catch them. Either way, use of forks to engage in fraudulent activity is usually restricted by local laws. Other locations or mesh sites don't care in the slightest if users are "human" or not - they utilize active and passive moderation to remove any harmful software bots or ALI users along with transhuman users who violate the terms of their services.
Obviously, as AI technology has gotten more sophisticated and ease-of-use and access to scripting knowledge has become easier and more widespread, the issue of verifying human users over weak AI or even basic software agent programs (which can be little more than a software script which posts "FIRST!" on every new XP upload by another user) has become more complex - but is still regarded as highly important, especially in polities with strict controls on AIs of all forms, from ALI to AGI. There are still a number of verification solutions.
First and easiest is many individual sites on the mesh, or services which might be connected to by a bot require a simple "Yes/No" verification if the subject is a transhuman (some phrase this as "sapient" or "not an AI", etc). Legal restrictions and basic code limitations prevent most ALI from directly lying when asked to verify - which works for many basic services. Of course, this does nothing to prevent ALIs which have been specifically coded to deceive or be able to break local laws, or many sub-AI agents who do not have any intelligence to shackle, just simple outputs and actions. For mesh sites which require more specific verification, users are often required to link an Ego ID, or another account, such as on one of the major social networks. This satisfies most that the user is a transhuman agent, and they can function normally. Building fake IDs and accounts for AIs and software bots is possible, but can be time consuming - and generally AI or scripts of their own can catch spammy or fake responses such as on a mesh forum and remove them, and most Social Networks have functions to flag spam or bot accounts and delete them. Many of them also tend to accumulate low rep scores naturally.
This more relaxed style is not enough for some groups and even governments, however. Very strict security or transparency habs may require a rather intrusive verification process which double-checks the Mesh ID of a device against a registry, identifying the type of device and who it is supposed to be owned by - and may reject devices with anomalous IDs or with obvious fake registry. Others may check metadata so that certain activities can only be performed from a device registered as a mesh insert or home server to try and make sure an Ego is behind it - which has a tendency to catch script kiddies stupid enough to try and run spambots or exploits right out of their own brains.
The most sophisticated and reliable of these systems, however, resemble the old CAPTCHA model, and other derivative standards. The simplest (but sometimes most time consuming) involves XP snippets. Software bots and ALI may be unable to properly "play" XP, and lack the complete neural modeling to experience all the elements of a full XP sensorium based on a biomorph. This catches many outright, and they are also generally unable to correctly answer the following questions about the system. X-CAPTCHA systems like this have very complex databases of possible snippets to play, but are still theoretically vulnerable to sweatshop AGIs or Infomorphs who might build a big enough database of correct responses to the challenge to bypass them - meaning that this option is not as secure for sites or apps without access to large servers, and usually involves contracting an external service. The best of these is an Extropian firm run by an AGI with an extensive amount of freelancers and forks to process and edit the XP library and randomly distribute around the system. The Argonauts also have a very strongly tested open source version - which unfortunately utilizes open source media libraries, meaning a handful of hackers have been able to get lucky in the past.
The Argonauts other solution isn't as guaranteed as the XP one, but is very good also. Based on advanced plagiarism checking algorithms developed for academic purposes, these systems ask short but subjective questions, and can check them for originality and authenticity - usually matching them with the responses expected from an Ego - and advanced ones can even keep a history and double-check similarities to a specific person, which can track alt accounts or stolen accounts. While it has some flaws, generally this system is sophisticated enough for those who wish to ensure that users aren't doing anything against terms of service involving bots or AIs. Others go old school, dropping the "reverse" and just administering a variant of the Turing test. This is common on tight-knit communities with many available active users. While the test itself is not perfect, generally most transhumans can ask the right questions to tell if a user is a basic chat-bot script or an ALI, especially if they are a more focused or specialized community with specific interest to check. Some argue this has a tendency for exclusivity - as many in a community may not claim a user is "authentic", but those with genuine passion can usually find it in other people, something an ALI cannot easily emulate.
It should be noted that Muses, having very sophisticated personality parameters, can often bypass these systems - however a Muse AI is almost never not paired with a transhuman and rarely acts outside their user's self-interest. This means that most locales and services do not necessarily mind if a Muse accesses their sites or apps, as they do so on behalf of a person - and they are still typically constrained by basic AI limits. Forks too, can complicate the issue. Alpha forks are just people, though they may not be distinct legal entities. Beta forks are more limited for space and focus, but their inclusion of memories and personality traits for context may mean they can pass many verification forms on the right device - but can often be matched to their Alpha. Deltas, being little more than a pruned AI template or skillsoft collection do not pose well as humans, and their amnesiac status means many subjective tests can catch them. Either way, use of forks to engage in fraudulent activity is usually restricted by local laws. Other locations or mesh sites don't care in the slightest if users are "human" or not - they utilize active and passive moderation to remove any harmful software bots or ALI users along with transhuman users who violate the terms of their services.
Wednesday, December 28, 2016
Teaching Software
As discussed in Oxford learning is an interesting area of concern for transhumans and their technology. Deep Learning psychosurgery lets one acquire skills quicker in costly procedures, and theoretically one can spend time on Simulspace under time dilation to lose less "real time" on learning something. Oxford itself is a study drug, which stimulates the brain to allow better memory. These ways, learning can be done faster. But, how does one learn better? Simulspace allows for wonderful substitution of "practical" learning, as does XP use - watching someone perform a skill. Skillsofts substitute skill directly by allowing the brain to link up to artificial memory data and access it as if it was their own. AI instructors or simple reference materials abound, with groups like the Argonauts generally dedicated to this sort of dissemination of information. A quick mesh search should reveal quick "how to" guides for most everyday subjects - though technical ones probably require more in-depth instruction. Obviously, educational institutions still exist across the solar system, but this can be expensive and time consuming to afford traditional education. But, sub-AI expert systems allow for a self-paced system of learning which can aid those who wish to acquire new skills.
Teaching Software: This software tool is a comprehensive educational tool which provides all necessary knowledge and instruction for a character to acquire or improve a single skill to basic proficiency. While not a proper AI, this system does have extensive knowledge in a very particular area, and for any purposes required has Profession: Education at 60 and an Academics: [Field] skill of the selected skill at 80. Teaching Software can only cover one skill (which cannot be a Psi skill), and if a new skill is to be taught, a new copy of the software must be bought. In addition to providing all learning needed to spend RP to improve or gain a new skill, the comprehensive library and instructional aids of the software grant a +10 bonus to utilize it's selected skill unless the user has 60+ points in that Skill - indicating their level of knowledge and experience exceeds that of the software, making it only good as a guide for self-study. Teaching Software can cover both active and knowledge skills. Exotic or rare skills may be more expensive. [Low]
Teaching Software: This software tool is a comprehensive educational tool which provides all necessary knowledge and instruction for a character to acquire or improve a single skill to basic proficiency. While not a proper AI, this system does have extensive knowledge in a very particular area, and for any purposes required has Profession: Education at 60 and an Academics: [Field] skill of the selected skill at 80. Teaching Software can only cover one skill (which cannot be a Psi skill), and if a new skill is to be taught, a new copy of the software must be bought. In addition to providing all learning needed to spend RP to improve or gain a new skill, the comprehensive library and instructional aids of the software grant a +10 bonus to utilize it's selected skill unless the user has 60+ points in that Skill - indicating their level of knowledge and experience exceeds that of the software, making it only good as a guide for self-study. Teaching Software can cover both active and knowledge skills. Exotic or rare skills may be more expensive. [Low]
Wednesday, September 7, 2016
Black Lightening
For most transhumans, caffeine is a popular "addiction", but mostly a cosmetic one. The effects of caffeination are limited compared to real nootropics or stimulants, or the enhancements inherent to a morph, but many also have an acquired habit from the past. It is a social habit as well, with many transhumans still enjoying a "morning coffee" or a "tea time" with others - Old Earth nostalgia being a prime factor. But for Synthmorphs and Infomorphs, these rituals are no longer possible. Sure, XP or VR can mimic the sensation, but this is still consuming a form of media, not a product. This is where Black Lightening comes in.
Black Lightening: This narcoalgorithm is designs as an analog for caffeine or other stimulants, which works by digitally simulating centers in the cyberbrain (or emulator software) which mimic a refreshing jolt of coffee. Users feel more pep, and are more active, and tend to feel like they can react faster, or get more done. This last aspect is actually because the algorithm also includes a time sense dilator, users actually experience that they move and react quicker to thinks, similar to the Neurochem implant, and gain +1 Speed while using the algorithm. Combined with the stimulant experience, the user feels an improvement in productivity and energy which is akin to how many felt on caffeine before, or many felt that they should feel. However, this stimulation has some downsides. Many also are jumpy, erratic or impatient - they have high volumes of energy they feel compelled to use on something, which may come out as nervous or repetitive gestures (pacing, tapping) or an inability to sit still or wait. Users must make a WILx3 test to remain passive or wait for something to happen, with failure indicating they need to take some form of action rather than wait. [Low]
Application: Run
Onset: Instant
Duration: 3 Hour
Addiction Modifier: +0
Addiction Type: Mental
Black Lightening is often sold for Synths in small packages which resemble k cups or tea bags, but are plugged directly into the access jacks.
Black Lightening: This narcoalgorithm is designs as an analog for caffeine or other stimulants, which works by digitally simulating centers in the cyberbrain (or emulator software) which mimic a refreshing jolt of coffee. Users feel more pep, and are more active, and tend to feel like they can react faster, or get more done. This last aspect is actually because the algorithm also includes a time sense dilator, users actually experience that they move and react quicker to thinks, similar to the Neurochem implant, and gain +1 Speed while using the algorithm. Combined with the stimulant experience, the user feels an improvement in productivity and energy which is akin to how many felt on caffeine before, or many felt that they should feel. However, this stimulation has some downsides. Many also are jumpy, erratic or impatient - they have high volumes of energy they feel compelled to use on something, which may come out as nervous or repetitive gestures (pacing, tapping) or an inability to sit still or wait. Users must make a WILx3 test to remain passive or wait for something to happen, with failure indicating they need to take some form of action rather than wait. [Low]
Application: Run
Onset: Instant
Duration: 3 Hour
Addiction Modifier: +0
Addiction Type: Mental
Black Lightening is often sold for Synths in small packages which resemble k cups or tea bags, but are plugged directly into the access jacks.
Sunday, September 4, 2016
Attack Barrier
Almost all computer systems and devices come equipped with a Firewall (or many other similar programs with a variety of names such as "Lavawall", "Icewall", "Thunderwall", etc.) which serve a basic function of screening unauthorized traffic in, while still allowing regular outgoing communications. For most, this level of protection suffices, especially when combined with active monitoring by dedicated information security technicians or onboard AI. But for many military systems, the risks of hacking offer too much to leave to conventional, low-level systems. Such systems were used before The Fall, but the TITAN systems often found other ways to bypass these protections, or were simply not deterred by the additional countermeasures of military firewalls - which led to their use being limited afterword, and almost exclusively resigned to use by governing bodies, even private individuals have a hard time acquiring these so-called "Attack Barriers".
Attack Barrier: Known also as a military-grade Firewall, or sometimes referred to colloquially as a "backfire wall", the Attack Barrier is a sophisticated, nasty piece of software which is regulated basically anywhere that has regulations. The system integrates sub-AI Expert Systems and a modular codebase to increase efficiency and difficulty to probe. Some are even heuristic or self-modifying, even deliberately forming small exploits for a moment, only to close them shortly after. This automatically applies a -10 penalty on InfoSec tests to bypass the Firewall of a protected system. But the real prowess (and danger) of an Attack Barrier happens if a hacker fails to properly bypass the system. Failing to break an Attack Barrier almost always results in the Barrier logging the hacking attempt, at which point the system automatically counterattacks. Wireless intruders are subject to a trace and counterintrusion (see p. 258 EP) with an Attack Barrier's built in skillsofts counting as Research and InfoSec of 40. Failure to trace or counter-intrude will have the system log appropriate recommendation to perform a wireless termination or reauthentication. More dangerous for a hacker, however, is what Attack Barriers do to wired connections. The attack barrier system produces a potent electrical feedback in the area of intrusion, such as access jacks or conductive surfaces in case of Skinlinks, which causes 1d10+5+Shock DV to the person or device plugged into the system, ignoring Armor Value. While intended to incapacitate the hacker, this also commonly has the effect of damaging or disabling the physical access point used by the hacker. [Moderate]
Despite their uses, Attack Barriers are fairly rare outside of government computer systems, and as noted are highly regulated by most local laws - which often limit their powers. Attack Barriers used by the Titanian Commonwealth, for instance, have particularly sophisticated tracing and counterintrusion properties, grafted from their own top InfoSec experts, but their barriers are almost all secured in hardware which is physically incapable of using the shock functionality (at least, officially). In the Planetary Consortium, hypercorps with controlling interests in habs or corp stations are allowed to utilize these softwares, but rarely do so because of the harm they inevitably cause to their own hardware - preferring to use indentures or shackled AGI for active security. A topic common in LLA legislation is if Lunar banks may be allowed to employ Attack Barriers to safeguard the electronic systems which govern bank security, which has yet to pass. Before the Fall, Attack Barriers were actually most common on military vehicles, intended to protect from unwanted hacking while still allowing normal network functions - but due to this making subverted vehicles all the more hard to retake, this use of them has almost universally been stopped. Even still, rumors persist of government operatives using them in their own Cyberbrains, a nasty surprise used by many a counterterrorism or undercover operative.
Attack Barrier: Known also as a military-grade Firewall, or sometimes referred to colloquially as a "backfire wall", the Attack Barrier is a sophisticated, nasty piece of software which is regulated basically anywhere that has regulations. The system integrates sub-AI Expert Systems and a modular codebase to increase efficiency and difficulty to probe. Some are even heuristic or self-modifying, even deliberately forming small exploits for a moment, only to close them shortly after. This automatically applies a -10 penalty on InfoSec tests to bypass the Firewall of a protected system. But the real prowess (and danger) of an Attack Barrier happens if a hacker fails to properly bypass the system. Failing to break an Attack Barrier almost always results in the Barrier logging the hacking attempt, at which point the system automatically counterattacks. Wireless intruders are subject to a trace and counterintrusion (see p. 258 EP) with an Attack Barrier's built in skillsofts counting as Research and InfoSec of 40. Failure to trace or counter-intrude will have the system log appropriate recommendation to perform a wireless termination or reauthentication. More dangerous for a hacker, however, is what Attack Barriers do to wired connections. The attack barrier system produces a potent electrical feedback in the area of intrusion, such as access jacks or conductive surfaces in case of Skinlinks, which causes 1d10+5+Shock DV to the person or device plugged into the system, ignoring Armor Value. While intended to incapacitate the hacker, this also commonly has the effect of damaging or disabling the physical access point used by the hacker. [Moderate]
Despite their uses, Attack Barriers are fairly rare outside of government computer systems, and as noted are highly regulated by most local laws - which often limit their powers. Attack Barriers used by the Titanian Commonwealth, for instance, have particularly sophisticated tracing and counterintrusion properties, grafted from their own top InfoSec experts, but their barriers are almost all secured in hardware which is physically incapable of using the shock functionality (at least, officially). In the Planetary Consortium, hypercorps with controlling interests in habs or corp stations are allowed to utilize these softwares, but rarely do so because of the harm they inevitably cause to their own hardware - preferring to use indentures or shackled AGI for active security. A topic common in LLA legislation is if Lunar banks may be allowed to employ Attack Barriers to safeguard the electronic systems which govern bank security, which has yet to pass. Before the Fall, Attack Barriers were actually most common on military vehicles, intended to protect from unwanted hacking while still allowing normal network functions - but due to this making subverted vehicles all the more hard to retake, this use of them has almost universally been stopped. Even still, rumors persist of government operatives using them in their own Cyberbrains, a nasty surprise used by many a counterterrorism or undercover operative.
Subscribe to:
Posts (Atom)