Thoughts on an Anti-Theft Feature

mdo · 09-20-2007, 01:04 AM

I had a thought about a possible anti-theft feature that would make it impossible to upgrade the software in a stolen camera while being transparent to users of legitimate cameras.

1. Give each camera a unique identifier burned onto a chip on the board, if it doesn't already have one.
2. Add a list of the unique identifiers of all stolen cameras to every software load. At four or eight bytes each, the space would be negligible. Hash the install with a shared secret to prevent tampering.
3. If the installer notes that the ID of the camera matches a stolen ID, it aborts the install and sets a flag in persistent memory in the camera which prevents the camera from being rolled back to an earlier version and keeps it from starting.
4. The camera then will remain unusable until a higher build-number software load is used that does not have that ID on the stolen list.

Downsides include:
a. Additional code will be required to execute the above. While the space required would be very small, more code means more chances for bugs or problems. If the chances for serious malfunctions with the feature couldn't be reduced to zero, the risk to production users of the camera would make the approach unacceptable.
b. The solution is predicated on there being a regular flow of upgrades, in the spirit of Red as a tool that will grow rather than becoming obsolete. There are regular upgrades now, during the development period, but once the full feature set has been enabled and the code has stabilized out, it is possible that a single build could serve the camera for a long time. The camera would still be usable as long as installers that were compiled after the camera was stolen are never used on the camera.
c. System upgrades can be expected to be restricted to licensed Red owners, meaning that the thief probably would have difficulty getting an upgrade anyway.
d. There is always a way to hack code -- although there probably would not be much incentive for a qualified coder to bother with it. However, since the camera is such a high-ticket item, the possibility for hack of the code exists.

Responses to some of the downsides:
i. As more and more Reds are sold, availability of upgrade software will become increasingly widespread, so that a thief will find it easier to download or steal the latest software. As availability of upgrades increases, the value of this solution would increase.
ii. Even though a stolen camera may still be usable, the mere fact that it cannot be upgraded reduces its value, and therefore its attractiveness to a thief.
iii. A person buying a Red would be able to verify that it is not stolen (unless very recently) simply by loading the latest build and noting that the display has not returned a message that the camera is on a stolen list.
iv. The longer a camera goes without an upgrade, the more suspect it would become -- unless Red users fall in love with earlier builds and stick with them. A old build in a camera would not be a reason for concern, but combined with other oddities, it could raise a flag.
v. There is never any complete solution to theft. Each anti-theft approach has to be evaluated in terms of the value of the deterrent vs. its cost.

These thoughts are presented in a vacuum of info on how much of an issue theft of Red cameras is likely to be. It may be that the risk is not significant enough to bother trying to implement something like this.

But it's an interesting diversion.

ZzzZZz... · 09-20-2007, 05:16 AM

I dont think that it will help that much, and could cause a lot of uneccesary trouble. If you are worried about security get a device like this one

http://www.pimall.com/nais/gpssnitchit.html

paint it black and glue it somewhere aruond the rods or battery plate. A gps chip inside the camera would be a cool optional feature in the future, especially if you are shooting somewhere in mumbo-jumbo land

jbeale · 09-20-2007, 10:40 AM

I think code hackers are pretty cheap; the value of a red is high enough to be a target. I mean, it costs more than an iPhone... :-)

I was thinking of something simpler, like a registered owner database run by a trusted neutral party (other people have suggested this also). It would have other uses, like finding rentals / DPs (and other problems like privacy or security risks?) Also, a bargain buyer may not care if the goods are hot. I don't know if a registry is practical but it's interesting to think about how it might work.

09-20-2007, 01:04 AM	#1
mdo Banned Join Date: Jun 2007 Posts: 178	Thoughts on an Anti-Theft Feature I had a thought about a possible anti-theft feature that would make it impossible to upgrade the software in a stolen camera while being transparent to users of legitimate cameras. 1. Give each camera a unique identifier burned onto a chip on the board, if it doesn't already have one. 2. Add a list of the unique identifiers of all stolen cameras to every software load. At four or eight bytes each, the space would be negligible. Hash the install with a shared secret to prevent tampering. 3. If the installer notes that the ID of the camera matches a stolen ID, it aborts the install and sets a flag in persistent memory in the camera which prevents the camera from being rolled back to an earlier version and keeps it from starting. 4. The camera then will remain unusable until a higher build-number software load is used that does not have that ID on the stolen list. Downsides include: a. Additional code will be required to execute the above. While the space required would be very small, more code means more chances for bugs or problems. If the chances for serious malfunctions with the feature couldn't be reduced to zero, the risk to production users of the camera would make the approach unacceptable. b. The solution is predicated on there being a regular flow of upgrades, in the spirit of Red as a tool that will grow rather than becoming obsolete. There are regular upgrades now, during the development period, but once the full feature set has been enabled and the code has stabilized out, it is possible that a single build could serve the camera for a long time. The camera would still be usable as long as installers that were compiled after the camera was stolen are never used on the camera. c. System upgrades can be expected to be restricted to licensed Red owners, meaning that the thief probably would have difficulty getting an upgrade anyway. d. There is always a way to hack code -- although there probably would not be much incentive for a qualified coder to bother with it. However, since the camera is such a high-ticket item, the possibility for hack of the code exists. Responses to some of the downsides: i. As more and more Reds are sold, availability of upgrade software will become increasingly widespread, so that a thief will find it easier to download or steal the latest software. As availability of upgrades increases, the value of this solution would increase. ii. Even though a stolen camera may still be usable, the mere fact that it cannot be upgraded reduces its value, and therefore its attractiveness to a thief. iii. A person buying a Red would be able to verify that it is not stolen (unless very recently) simply by loading the latest build and noting that the display has not returned a message that the camera is on a stolen list. iv. The longer a camera goes without an upgrade, the more suspect it would become -- unless Red users fall in love with earlier builds and stick with them. A old build in a camera would not be a reason for concern, but combined with other oddities, it could raise a flag. v. There is never any complete solution to theft. Each anti-theft approach has to be evaluated in terms of the value of the deterrent vs. its cost. These thoughts are presented in a vacuum of info on how much of an issue theft of Red cameras is likely to be. It may be that the risk is not significant enough to bother trying to implement something like this. But it's an interesting diversion.

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

09-20-2007, 05:16 AM	#2
ZzzZZz... Banned Join Date: May 2007 Posts: 36	I dont think that it will help that much, and could cause a lot of uneccesary trouble. If you are worried about security get a device like this one http://www.pimall.com/nais/gpssnitchit.html paint it black and glue it somewhere aruond the rods or battery plate. A gps chip inside the camera would be a cool optional feature in the future, especially if you are shooting somewhere in mumbo-jumbo land

09-20-2007, 10:40 AM	#3
jbeale Senior Member Join Date: Jan 2007 Posts: 1,657	I think code hackers are pretty cheap; the value of a red is high enough to be a target. I mean, it costs more than an iPhone... :-) I was thinking of something simpler, like a registered owner database run by a trusted neutral party (other people have suggested this also). It would have other uses, like finding rentals / DPs (and other problems like privacy or security risks?) Also, a bargain buyer may not care if the goods are hot. I don't know if a registry is practical but it's interesting to think about how it might work.