You can't pull information from the stack. It's possible to stop the update by hacking the cyberbrain directly, but the stack can't be interfaced at all. It isn't possible to record an ego from a hacked cyberbrain either.

Remember that cortical stacks are the Eclipse Phase version of read-only magnetic tapes stored in a back room vault. They are, [u]by design[/u], safe, offline storage. You [u]cannot[/u] get a read off a cortical stack without extracting the stack, full stop. If the ego in question is linked to a biomorph, you [u]cannot[/u] stop the cortical stack nanites from writing to the stack, unless the user has some other headware which allows for that, like a memory lock, which you can subvert, or you infiltrate their body with hostile nanotech and flush the cortical stack nannies. If the ego in question is running on computer hardware, however, and you do successfully subvert it, you [i]can[/i] fuck with it; hack it directly, lift a copy of the ego running process, etc. But you still won't be able to wipe the stack.

Well, if you had an unlimited data access to the morph, you could make the firmware/diagnostics systems to make false alarms and tell the ego to turn in to the nearest clinic for system check up. Then in the "clinic" you just copy the ego and after that u do some psycho surgery so that it "never happened."

godmoney wrote:

ShadowDragon8685 wrote:

But you still won't be able to wipe the stack.

I don't have the books here with me atm. But how does a stack get wiped (i.e. when you switch morphs)?

If you have hardware access - IE, the body is laying down in an ego bridge - you can have the ego bridge nanites wipe the stack. If the stack has been pulled, you can do it with nanites, too. But you can't do it just by hacking their headware, not even if they're running on a cyberbrain.

MrWigggles wrote:

No no, its not pulling information from the Stack. Its getting the information before it goes to the Stack, or making a copy of the information as it goes to the stack.

The problem with [u]that[/u] is the [i]how[/i] of how an ego gets onto a cortical stack. Think of the cortical stack as a video game emulator's save state file. Only the saved state in question is the entirety of a person's mind. Now consider that there is no computer hardware involved; it's a flood of nanomachines in the subject's brain, watching and recording as neurons fire and such, writing those changes to the stack on about a five-minute loop. The person's mesh inserts [i]might[/i] be involved, but [u]only[/u] as a strict state-write to a partition of the stack which is reserved only for their mesh inserts' files, including their muses, because holy shit it would suck to be resurrected without your muse or your porn stash or your working files (in that order of importance for most people). The mesh inserts would never have any sort of control over the read/write function of the ego-reading nanites, which are strictly a part of the cortical stack augmentation. So you can't execute a Man in the Middle attack by subverting a person's mesh inserts... [u]Unless[/u] they're running on a cyberbrain. A cyberbrain still uses a cortical stack, but the cyberbrain feeds the data to the stack's nanites, who write for it. This is much faster, but the cyberbrain can be itself hacked, albiet not directly. First you have to hack their mesh inserts, and hope they don't have [u]or get[/u] any clue that you're hacking their mind, because the ego running on a cyberbrain can sever their cyberbrain's connection to the mesh inserts at-will, as well as hard shutdown the mesh inserts' wireless connectivity. But if you successfully "tunnel" through to the cyberbrain through the mesh inserts - bearing in mind you'll need Badmin access on both, and there's additional penalties for hacking cyberbrains - then yes, you can execute a Man in the Middle attack on the write-to-cortical-stack function. At that point, though, you could do pretty much anything you wanted, like grabbing a copy of the target's ego wholesale, performing involuntary psychosurgery, locking them out of their own body and remote-controlling them to you, deleting them... Whatever you felt like.

I don't think cyberbrains use anything like standard computer architecture. They can't execute infomorphs or run programs. I suspect that is why cyberbrain hacking is so limited, as when hacking one it's only possible to pull memories from the mnemonic aug, seize control of the body via a puppet sock, control AR I/O feeds (scorchers), entrap the ego, disable a cortical stack feed, or shut it down. All of those work with cyberbrain peripherals, except for shutting the cyberbrain down; which supports the idea that cyberbrains themselves are not standard computers and cannot be normally hacked. I suspect they're kind of cybernetic hardware emulations of bio-brains. It isn't possible to pull an ego from a running cyberbrain, or execute psychosurgery (that typically requires an infomorph). I'm guessing that cortical stacks actually interface with the cyberbrain on a hardware level, as that should be the most secure system (digitizing the ego state and saving to the stack through a peripheral would be a cyberbrain hacking vulnerability). It's probable that some EP cyberbrains do use more vulnerable methods though, as the technology of the setting isn't as monolithic as the rules present. How the hack convinces the stack to stop updating? I'm not sure of that, and that's the hole in this theory. My guess would be triggering some kind of error-prevention part of the stack so it treats the ego as compromised, but there are a lot of assumptions there.

Because there is no other systems between the brain and the stack? That would be the safest design, so i would imagine it would be widely used. The way i see it, the brain is the scene, the stack is the film and the nanites are the camera shutter. Data flows only one way, and any control u might have in all of this would be with the shutter. I would imagine there was a way to deactivate the stack. This would probably be a mechanical switch on the hardware buried under the skin (for maintenance purposes). Anyway, about cyber brains; the only way i can see the whole thing working is to load the stack-image into ego emulator software. the image of the ego would run in a simulated environment and would send control commands through the emulator interface to the morph, if there was any. The digital, multi-layered nature of such a solution opens it up for a number of different attack vectors.