what the WMI needs to be Options

[mfb]

okay. SR4 says it's going to streamline things, especially Matrix things. let's talk about what it should be, especially from a rules perspective. i've listed my requirements in no particular order below. note that while i've put a lot of thought into these concepts, i've put very little into their execution. this is primarily a list of what needs to change, and only secondarily a list of ways to change it.

Tests
a single roll against a TN with little variance should resolve most actions. a wide array of ridiculously high TNs, which are modified by a ridiculously large number of reductors, is completely ridiculous. personally, i think all decker-hackers should use Channels, a la otaku. at least, the mechanic should be similar: a small number of multi-purpose, very expensive programs.

i also think that the ACIFS should be reduced by at least one. get ride of Slaves completely: this is the realm of rigger-hackers. replace all Slave operations with rigger-hacker actions. i think Index and Files could possibly combined; Access and Control can certainly be combined. jesus, you use Access for one thing: Logon to X! it does not deserve its own stat. either that, or expand Access to include all forms of electronic warfare and cryptography in general (since cracking crypto is basically all Access is anyway). want to listen in on your enemies' comms? Access test. want to gank control of a rigger-hacker's drone network? Access test. you get the idea. a name change would be appropriate: Crypto gets my vote.

even better, give hosts one TN, which is then reduced by a decker's programs. for instance, you could have a rating 10 host. if you're performing an Access action--oops, Crypto action!--and your Crypto program (or whatever) is 5, then your TN is 5.

Hackability
everything on a computer should be hackable, and every computer with a network connection should be hackable. i want to play a Comp Sci major who hacks girls' pocsecs for their numbers, goddammit. i want to be able to gank control of the IC that's attacking me, and send it after the security deckers that just popped in. i want to be able to hack another icon and pervert his actions. this includes cyberdecks, or whatever replaces them: the connection is two-way. i want to hack cyberdecks. i want to be able to hack a computer so completely that i can change the ACIFS ratings, reset my security tally, and delete IC--and i want everything in that last sentence to be hard as hell.

Definitions
i want simple, non-detailed definitions for computer-related statistics. basically, i want to know if ACIFS are hardware, software, or firmware. i want to know how much Mp a Red host has. i want to know how many users an Orange-6 host can handle, and i want to know what happens when you overload it with connections.

Crafting
related to the above, i want to be able to build hosts. i want to be able to pay more for a host that's pocket-sized, or nanowoven into my clothes. i want a server in my goddamn sunglasses.

Similarity
everything needs to use the same damn rules. none of this 100Mp wristtop computer / MPCP-8 cyberdeck / Orange-6/10/12/11/10/11 host crap--everything that can be hacked (see Hackability, above) should use the same stats as everything else that can be hacked.

Security
the entire concept of security sheaves should be reworked. it's interesting, but in the end, it's both limiting. the sheaf system means that computers can't proactively respond to intruders (unless the computer operator can afford IC Constructs, and it's a very reactionary proactivity even then). moreover, the reactivity of computers to intruders is incredibly limited. there's no way to respond intelligently to specific methods of attack.

i think that every host should have a security controller. a single piece of IC (larger/more important systems might have more, but one would be the norm) that handles all security for that host. as hackers gain attention, this controller would spawn countermeasures as required. it should be possible, though difficult, to simply take the controller down in combat, neutralizing all security for as long as it takes the host to rebuild the controller (several minutes for low-end hosts, several rounds for high-end). to keep things simple and streamlined, the mechanics controllers use for spawning countermeasures should be similar to summoning nature spirits.

Countermeasures
countermeasures should do more--and less--than just attack things. after all, hackers appear as glitches in the system. in the real world, it's very rare for guards of any stripe to be given shoot-on-suspicion orders, where anything that moves at all (wind, animals, wandering children) gets an automatic three-round burst. yet that's exactly how IC reacts to everything, in SR.

instead, the first response by the security controller (see above) should be to spawn a countermeasure that raises the TN for the specific action that the controller detected, or which allow the controller to gain more information about any further attempts (a la probe or scout IC).

SIMPLICITY
jesus christ, hacking should be simpler. none of this crap with 500,000 programs that take up 15613.54Mp which you then have to find the square root of in order to discover how much time it takes to get to hell with that. operations should be action-based: write up a definition of what an Index action is, or a File action, or an Access/Crypto action. we don't need four seperate Logon to X actions. we don't need umpteen billion Locate X actions that all use the same goddamn program, subsystem rating, and skill anyway.