TorBEL data formats and active testing: feedback requested

Harry Bock hbock at
Tue Jun 1 22:12:33 UTC 2010

Hash: SHA1

Hi all,

I've been working on TorBEL ( for GSoC
this year, and I'm looking for feedback on my current specifications for
data export and for an active testing mechanism.  I've bounced some
ideas around some of the developers in #tor-dev but I'd like to extend
the discussion to the mailing list.

Currently, I've written a brief draft of the "generic" data format,
which describes what fields each exit router/IP entry will contain, and
a specification for the CSV export format.  I've also started a
specification on how I plan to perform basic router testing and why.  I
tried to keep them similar to the RFC-style proposals and specs Tor has.

You can find the specifications in Git at:

Some things I need feedback on:
  - Are the fields listed for each router/IP adequate information for
consumers of exit router data?  If not, what should I add, remove, or
  - What other data formats are useful, easily distributable, and easily
parsed, and why? Keep in mind that consumers of this data will want
frequent updates, and the data format should be as compact as possible
to conserve bandwidth.
  - What are the most interesting ports we should be looking to test?
Should we test interesting ports even if they are explicitly rejected by
the router's ExitPolicy?  We can't test all ports on all exit routers,
so we have to limit ourselves to ports consumers would be most
interested in.  The data collected in [1] and [2] are a good start, but
I feel we need more feedback on the subject.
  - Any other comments, questions, or concerns about my approach thus
far as more than welcome!


Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla -


More information about the tor-dev mailing list