[tor-dev] Status Report - Torsocks 2.x
dgoulet at ev0ke.net
Wed Feb 5 19:38:36 UTC 2014
After Fosdem 2014, I had the chance to talk with a lot of people about
Torsocks and the current status. It appears it's actually quite used!
(more than I thought). I've already received multiple contributions
after a quick announcement of the project during my talk.
This email is to provide a quick status report on what's going on with
this effort .
In a nutshell, unfortunately, it has been almost two months since the
last commit but free time is coming back to me thus expect *MORE*
Nick did a *very* nice review of the code base  that I finally
answered recently  but yet to fix stuff upstream. Feel free to jump
in the discussion to help or even contribute.
Yawning and I had a discussion on #tor-dev about the use of a bitmap
instead of an hash table to store connection pointers indexed by file
descriptor. After looking in the BSD and Linux code base, we ended up
seeing that both OSes allocate the lowest FD everytime a new socket is
created which makes thing much more easier to manage a bitmap. Apart
from a performance stand point of view, it appears that there is no
compelling reasons to move the connection registry from an HT (current
state) to a bitmap. Feel free to speak up here if you see any reasons!
Finally, Luke Gallagher will continue to contribute tests to the project
so big thanks! 
After the Tor dev. meeting this month, hopefully things will be close to
ready for a full on release, I'll most probably write a blog post about
this effort on the why it has been rewritten, the current
state/limitation and upcoming features.
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 603 bytes
Desc: Digital signature
More information about the tor-dev