I will be submitting this to be made official, but have decided to post it here to get feedback and have at least be a house rule option in case it gets shot down.
Slaved devices = limit is the controlling device's Data Processing, these devices are part of your network but can also be controlled remotely. Primary use would drones, vehicles, and firing platforms, but I'm sure you could get creative with other devices.
Networked devices = limit is infinite as long as your able to be connected. Noise or other signal blocking would be a factor. A networked device is part of your network (PAN) and therefore gains all of the strengths and weaknesses of your security. (Willpower and Firewall)
Unattached/unattended devices = these devices have no network connection of any kind what so ever, and only have their own built in security (often a barely existent firewall if anything) but also can not be used to gain access to a network.
Godspeed to you

I believe that the main reason why the slave limit was added in the first place was to give Data Processing a better purpose. Precisely, a better purpose
in comparison with the Firewall Attribute. Before that, there was next to no reason not to take a setup with a higher Firewall (right now, only VR Initiative comes to mind). And then, someone tried die "fix" this problem with the slave limit, but instead created a much bigger problem due to the overall low stats of commlinks.
So, hereīs another alternative that just came to my mind:
Data Processing Actions.
Just like the way Attack and Sleaze actions work, some (mostly, legal) Actions may profit from higher Data Processing and thus be hampered by a higher Firewall setup, imposing a negative Dice Pool modifier if the Firewall Attribute is higher. That
may seem a bit weird at first, but just think about your everyday computer experiences: The more "secure" systems with proper firewalls, virus scanners, blockers, etc. are often a bit of a nuisance to work with, because you will be constantly pestered with warnings, constant updates, affirmation prompts, checks, small delays.... To me, that seems like a good enough explanation for
some Matrix Actions being penalized if the setup of the PAN is more about security and less about doing stuff:
- Matrix Perception: An overzealous Firewall might block out more "potentially harmfull signals", making it also harder to find hidden stuff or have a closer look at Matrix entities.
- Matrix Search: The Firewall interferes with your searches and tries to hinder you from straying into unsafe territories - which is often were you find the more valuable informations.
- Edit File: "Are you sure you want to run this file?" "For your security, you can only view this file in safe mode, some functions have been disabled" "Due to security risks, this file format is not supported anymore..." You know what I mean. While there are no rules for file-embedded Viruses and such (yet?), that doesnīt mean that this isnīt a phenomenon a that your overzealous firewall wants to protect you from.
Other possible Data Processing Actions may include
Format Device,
Jack Out*,
Reboot Device and
Trace Icon. However, I wouldnīt overdue it here. Also, while it might be fitting for
Encrypt File and
Disarm Databomb, I also wouldnīt dub any of the actions as "Firewall Actions" (yet), since pretty much every defense test against attackers is already profitting from a higher firewall.
*"Please make sure to alway log of from the Matrix in a secure fashion", sayeth the firewall, while the hacker gets fragged by the black IC...
