Changes between Version 12 and Version 13 of TorWeeklyNews/2013/11

Sep 11, 2013, 9:42:27 PM (6 years ago)

draft of entry guards/linkability item


  • TorWeeklyNews/2013/11

    v12 v13  
    5454 [XXX]
     56Entry guards and linkability
     59Leif Ryge pointed out [XXX] an issue with Tor's current 'entry guards'
     60system, whereby connections entering Tor from different points on the
     61same network could potentially be linked to an individual user based on
     62the three entry nodes selected by that user's Tor client, which remain
     63constant for a period of 4-8 weeks [XXX].
     65Leif suggested that "assuming this is an accurate assessment, wouldn't
     66it make sense to maintain separate sets of entry guards for each network
     67that the user connects from?"
     69Nick Mathewson replied [XXX] with an acknowledgement of the problem and
     70a number of reasons why simply generating separate sets of guards might
     71also harm a user's anonymity: "You would *not*, for example, want to
     72maintain a different set of entry guards for every IP that you receive,
     73since if you did, a hostile DHCP server could feed you new IPs until you
     74picked a hostile guard. Similarly, if you are a busy traveller who changes
     75your view of what network you are on hundreds or thousands of times, your
     76chance of picking a hostile guard would rise accordingly." He also pointed
     77out that "having a record in your state file of every network you have
     78visited is not necessarily the best idea either."
     80Nick concluded by mentioning Roger Dingledine's proposal to lower the
     81number of entry guards selected by a client to one only, "to avoid the
     82property of letting guard choices identify Tor clients".
     84 [XXX]
     85 [XXX]
     86 [XXX]
    5688Miscellaneous news