4 ms·
Might work for piracy but unfortunately it looks like they punted on issues that would help use by activists. Design assumption #1 (white paper section 3.1): "T
by hedgehog 15y ago
Might work for piracy but unfortunately it looks like they punted on issues that would help use by activists. Design assumption #1 (white paper section 3.1): "The traffic of every node in the network is assumed to be eavesdropped on (individually, but not globally in a fully correlated fashion) by an external party."
A state-level adversary can instrument their major ISPs, put a bunch of nodes on the network, push traffic through the network and do analyze the resulting traffic. Once they figure out who they want to question it's pretty simple, the mere presence of the software on your computer would be incriminating.
I think right now social networking sites are the best game in town because they have legitimate non-activist uses so they won't automatically get you in trouble. To improve on them resistance to traffic analysis and on-client footprint both need to be dealt with or the tool needs to be really popular for some innocuous application.
It is a really interesting set of problems though.
- robert_foss 15y agoLike the whitepaper states, the current specification isn't aimed at defeating state level adversaries. However relatively small changes will improve that situation. Changes like peer selection algorithms, two-way tunnels, random padding and a bunch of others.
- kragen 15y agoThe reason they punted on defeating the global-traffic-analysis adversary is that that's, well, really hard. The projects that are requiring themselves to solve that problem before releasing their first version are still in stealth mode.
- antihero 15y agoI'm working on a site that allows really simple encrypted messaging, anonymous event organisation, etc. If I made it available via phantom (or someone used the source code to set it up themselves), that would solve this issue.