If you were driving along the public road on a car, you can't object to people taking a picture of you, or taking notes about what you were wearing, number plate, etc. You effectively loose the right of privacy in public. And if someone saw your car on a public road, and decided to go and make an exact duplicate (with the exception of trademarks, and design marks) you couldn't do much about it. I would have thought that once your traffic is on the "public" Internet your right to privacy is somewhat diminished. I say "public", because the Internet is clearly not a private network. So I would have guessed that the capture of summary data (packets headers) should be fine. I would also guess that the exact duplication of that data (full packet capture) should also be okay. I'm more perplexed by copyright though. If I write a book and put it into a public library you don't have the right to make an exact copy. So if I email that same document across the Internet, does someone else have the right to make an exact duplicate with a packet capture? I'm kinda guessing the answer would be no. But then the same old argument explodes into web proxies, which also obviously make a copy of data that is being browsed. With regard to determining which traffic is p2p; I believe this would be futile with only packet headers. You would need at least the start of the payload. -----Original Message----- From: Nathan Ward [mailto:nznog(a)daork.net] ...
I'm sure the issue of ethics has been raised on this topic but I hadn't seen any mention in this thread and am unclear where users stand.
Are users advised that their data is being captured for analysis? ... Is it any different to analysing traffic in order to, for example, detect and limit p2p file sharing?
Generally, this sort of analysis is done with only packet headers. p2p file sharing detection/limiting stuff often looks at full packets. One could argue that that is more invasive. Plenty of other things do deep packet inspection as well - ddos detection, transparent http proxying, etc. Much more invasive than the simple header analysis that has been proposed and discussed here. -- Nathan Ward _______________________________________________ NZNOG mailing list NZNOG(a)list.waikato.ac.nz http://list.waikato.ac.nz/mailman/listinfo/nznog