5 ms·
I honestly just wonder if this was ever discussed in any court, because I've never heard of it. Like how much data has to be sent before it's distribution of co
by pingyong 7y ago
I honestly just wonder if this was ever discussed in any court, because I've never heard of it. Like how much data has to be sent before it's distribution of copyrighted material. You could really go either way with this. Either 0 bytes (just metadata), or 1 byte (that would be reasonably expected to be taken from the copyrighted work, even if the byte is not unique in any way), or a byte sequence that is fairly certainly unique to this specific .mp3 (probably like 32-512? bytes), or enough data that you can reconstruct some sort of sound from it that you can play and that is long enough to be grounds for a copyright infringement.
- Dylan16807 7y agoI'm confident that the threshold is not whether you can reconstruct anything. Because it's easy to have a system where each uploader provides only even or odd bytes and it's impossible to get anything but noise out of a single source.