February 2015 - tor-dev - lists.torproject.org

[RESEARCH] Student of University of Verona, Thesis about TOR
by Diego Sempreboni 02 Feb '15

02 Feb '15

Hi, I am a student of Engineering of the University of Verona. For my master's thesis, I decided to analyze TOR network, analyzing various attack scenarios. By consulting several documents have not been able to obtain a pattern of messages exchanged during the handshake protocol nor other details about the packets exchanged later. It would be possible to have some information about this? I write hoping someone can help me. Thank you.

3 2

Estimating censorship lag by obfs4 blocking
by David Fifield 02 Feb '15

02 Feb '15

According to a couple of blog comments on February 1, 2015, the default obfs4 bridges are blocked by GFW. https://blog.torproject.org/blog/tor-browser-45a3-released#comment-86416 https://blog.torproject.org/blog/tor-browser-45a3-released#comment-86907 The obfs4 bridges were introduced on November 17, 2014. https://blog.torproject.org/blog/tor-browser-45-alpha-1-released I have a message from Yawning Angel verifying that the obfs4 bridges were reachable on December 2, 2014. With this information, we can put rough bounds on how long it took the GFW to react to a new circumvention system: between 2 and 10 weeks. Does anyone know information that could tighten up the bounds, a date between December 2 and February 1 when obfs4 bridges were known to be either accessible or inaccessible? David Fifield

1 0

Damian's Status Report - January 2015
by Damian Johnson 01 Feb '15

01 Feb '15

Greetings wonderful world. January has been delightfully focused on a couple substantial feature branches that are now merged! -------------------------------------------------------------------------------- Descriptor Lazy Loading -------------------------------------------------------------------------------- When reading descriptors without validation (which is now the default) Stem is 25-70% faster! https://lists.torproject.org/pipermail/tor-dev/2015-January/008211.html This is because we now parse descriptors fields on-demand. Just need a server descriptor's exit policy? That's all we'll bother reading. This is why the heavier the descriptor type the larger the benefit. -------------------------------------------------------------------------------- Unified Python 2/3 Codebase -------------------------------------------------------------------------------- Thanks to Foxboron Stem now has a unified python 2/3 codebase! https://trac.torproject.org/projects/tor/ticket/14075 This means that Stem now runs directly with any version of python from 2.6 to 3.x. Prior to this Stem did a 2to3 export when you installed it to provide cross-version compatibility. For users, the benefit of Foxboron's overhaul is that python3 installation and test runs are now much, much quicker. -------------------------------------------------------------------------------- Beside these a few other noteworthy tasks were... * Nick had a handful of suggestions for testing improvements, neatest of which was to include the command to re-run just individual tests when we have a failure. (#14113) * Our port_usage() function always returned None. (#14046) * Proc connection resolution could fail on busy systems. (#14048) * Fixed how the 'strict' argument works with exit policies. (#14314) Cheers! -Damian

1 0