I'm a decker and I hacked into a security network and I now have access to the various files and other icons. Assuming I have the proper marks on everything I need, I have the following question:

If I want to edit the feed to create a loop and the feed is composed of dozens of cameras, is it one file that I need to edit or 1 file per camera feed?

I think it could be both depending on how it was setup. A really safe host would probably have a seperate file for each feed and a simple one would probably have one file for every cameras.

Am I understanding this correctly?

  • 3
    \$\begingroup\$ Wouldn't this, as you say, depend entirely on how the GM says the system is set up? \$\endgroup\$ Feb 28, 2014 at 18:34
  • 1
    \$\begingroup\$ I'm not sure, that's the thing. I don't know if it's described anywhere. Logically, if we take this from the real world, a single file to manage a live feed from dozens of camera is unrealistic. But computers don't work exactly the same way in Shadowrun \$\endgroup\$
    – user4000
    Feb 28, 2014 at 18:44

3 Answers 3


Assuming an entirely digital setup, and the multiple cameras are viewed in the same way as modern times (a separate monitor per camera, or you have to select a camera to view its feed) a modern camera system would typically be a separate file per camera feed, written for a certain period of time before creating a new file with a new timestamp.

While a single file per camera feed would be nice, continuous reading and writing to an ever-enlarging file would cause slowdown due to manipulating a huge file, and it would make sense to perform archiving on small separate files periodically rather than archiving a file then trying to archive essentially the same information with a little bit more information added.

A more advanced setup, assuming the future of Shadowrun can allow for it, would be a single file with header information per data packet that identified not only the image being stored and the timestamp, but also the ID of the camera used to take the image.

Essentially, from a coding standpoint multiple files is an easier and more efficient system to make, as you only have to write simple code for one camera at a time, then store the information with a suitable filename to indicate which camera was being recorded; this also means you can perform smaller write operations more frequently asynchronously. A single file system would have to perform a single write action synchronously to a large file, which would produce a resource bottleneck that would make the system eventually crash due to out of memory exceptions.

If I were GMing your loop, I'd have a "multiple file" camera system, and you'd need to access the files for each camera feed for the amount of time you wished to loop; typically this would be a single file per camera giving you about an hour of looped feed, which would then loop until the people watching the feed noticed. In terms of difficulty/time taken though, the number of files wouldn't matter.

The process would involve editing code in the display element of the program to loop the video of each camera feed given a specific timestamp and apply a foreach to iterate through any and all cameras (or a switch-case statement if you had specific cameras you wanted to loop and not loop). If you rolled well, you'd also be able to subvert the recording and archiving element to write the looped feed to the new files, so there was no evidence you were there at all.


There are a couple of things that seem to be important (ignorant of your GM's opinions and understandings of course)

First: The other answers I've seen as of this one imply that each camera is constantly writing the loop to its drive. Are you just trying to fool people actively monitoring the live feed or the archives when they go back to check? if so, I would start by saying multiple files but not necessarily one per camera (see below)

Second: If you want to use the same loop feed on all cameras (although not stealthy it's hilarious when all the cameras show laughing skulls instead of feeds), it could in theory only take one to two files - one master file to broadcast, and one file on their server for a backdoor but it is a fragile structure.

Depending on how similar the hallways are, you don't need one file per camera, you would need one file per loop plus one file to distribute unless you're doing it manually (yay tedium). In response to the HDD sizes in other answers, as long as your loop takes up comparable or less space, you can override the recorded feed by updating it instead with its loop just nanoseconds before it goes live simply by ignoring the camera input and appending the loop every few seconds. Any latency from the camera to storage would cover the overwrite. Or you could simply use the code loop as the only feed with a line to factor the loop to restart (hopefully) seamlessly but they may be suspicious if the file is exponentially smaller than expected.

With technomancy and decking, things can get nice and buttery soft though. I'm assuming the standard rules are in effect to get into the system, and depending on the resilience of the target network's security (especially if their responses are stronger than their passive firewalls) then redundant files and tactics may come into effect.

Bottom line as always is the GM has final say. If the network is complex enough, they could deem (regardless of how logical it seems to you) that it requires more or less files to represent how many layers of security, back doors, and even active deckers/technomancers waiting on the other side.


Did they save money when they installed it or did they splurge for security? Perhaps security was intended from the get go? Or was it added as an afterthought?

As Ardavion says: multiple files are more efficient and easier to code - and also back up. If one file is sent to hell, your entire monitoring just died, because of a HDD failure. Also, if I have a camera for monitoring, it's prudent to have it WITH storage - so whatever it records is stored with the camera. That way it can be standalone component or can be plugged in a larger monitoring system. Of course, that is fine provided the camera can't be accessed and toyed with by potential attackers.

As your decker, I'd do the following:

  • get intel on authors of the system you're hacking, of contract they signed (what is the system supposed to do, what was asked of it)
  • get intel on company you're hacking - are they known for being security-aware? Are they just stating security is crucial for good PR?
  • lay low in your hacking
  • observe which files are growing and with what kind of information (even now every OS gives you commands to see recently modified files)
  • try to occasionally browse the system in a least intrusive way to see where the camera's data are going / being stored. Are they duplicated?
  • If the company you're hacking is paranoid about their system they may have autodetection on some attacks (id data that should be sequential, coded timestamp, checking that new data you receive is identical with already received data - all that can easily spot that cameras are looped)

If you are unsure how to role-play it, just state your intentions, let the GM state difficulty, add ideas (I'm using extra AI for the task, or I'm using this-and-that-surefire-technique) for extra dice / lower step and roll.


You must log in to answer this question.