Also, in your scenario, 2GB at 16 captures per second is 32GB per second. One would have to have a VERY expensive disk array to keep up with that kind of load. A standard home desktop would be hard pressed to capture memory once a minute... and that would be pegging the system just to RECORD everything, not even talking about using the system to do anything USEFUL...
To all posters that still think this is about turning in RAM chips: stop thinking you are clever by pointing out RAM is volitile!
OK, so the judge says that RAM is a document. Fair enough, but there are flaws: 1) RAM is transient (changes over time), so it would either have to be captured at a particular instant, or some sort of transactional logging would have to be implemented. The overhead of that is HUGE, not to mention storage requirements. 2) RAM is binary. This might be obvious, but the data in RAM is only meaningful to a program that knows how to interpret the raw 1's and 0's. In theory, TorrentSpy could just send a binary image of the RAM contents to the MPAA and say "good luck".
I suspect the next ruling will be to capture memory dumps and provide debugging symbols of TorrentSpy to the MPAA. This would allow the MPAA to drill down into the memory snapshot and recover the IP addresses they are after. Unfortunately, this potentially exposes other information contained in the memory dump that was not ordered from the subpeona.
The ultimate point is that, as other users have pointed out, data in non-persistent places are now deemed "documents" in whatever form they might be in. Now, with the help of these "memory documents" sensitive information could potentially be recovered: Passwords, banking information, the list of websites you visited recently, even searches that you made on TorrentSpy (even if you download or share nothing!).
At some point I hope a distinction is made between having data, looking for data, using data, sharing data (access), and distributing data. Currently it seems the idea of data is that if you have/want to have/or did have it, you fit into all of the above categories. This makes it easy to find you guilty!
Slashdot Mirror
Also, in your scenario, 2GB at 16 captures per second is 32GB per second. One would have to have a VERY expensive disk array to keep up with that kind of load. A standard home desktop would be hard pressed to capture memory once a minute... and that would be pegging the system just to RECORD everything, not even talking about using the system to do anything USEFUL...
To all posters that still think this is about turning in RAM chips: stop thinking you are clever by pointing out RAM is volitile!
OK, so the judge says that RAM is a document. Fair enough, but there are flaws:
1) RAM is transient (changes over time), so it would either have to be captured at a particular instant, or some sort of transactional logging would have to be implemented. The overhead of that is HUGE, not to mention storage requirements.
2) RAM is binary. This might be obvious, but the data in RAM is only meaningful to a program that knows how to interpret the raw 1's and 0's. In theory, TorrentSpy could just send a binary image of the RAM contents to the MPAA and say "good luck".
I suspect the next ruling will be to capture memory dumps and provide debugging symbols of TorrentSpy to the MPAA. This would allow the MPAA to drill down into the memory snapshot and recover the IP addresses they are after. Unfortunately, this potentially exposes other information contained in the memory dump that was not ordered from the subpeona.
The ultimate point is that, as other users have pointed out, data in non-persistent places are now deemed "documents" in whatever form they might be in. Now, with the help of these "memory documents" sensitive information could potentially be recovered: Passwords, banking information, the list of websites you visited recently, even searches that you made on TorrentSpy (even if you download or share nothing!).
At some point I hope a distinction is made between having data, looking for data, using data, sharing data (access), and distributing data. Currently it seems the idea of data is that if you have/want to have/or did have it, you fit into all of the above categories. This makes it easy to find you guilty!