Organize an obfsproxy/PTTB test run
karsten and arma were considering an obfsproxy testrun, where a client-PTTB (#4927 (closed)) would be deployed, configured with hardcoded obfs2'ed bridges.
This should get us some bug reports on obfsproxy, some usability comments from end-users, and help us figure out #5038 (closed) and other performance-related questions.
The initial plan was to setup an obfsproxy in moria, but I believe it's a better idea to setup obfsproxy in less sensitive boxes [0].
This is a ticket to coordinate this effort.
#5047 (closed) is also important, to gather statistics and see approximately how many users used pluggable transports during the testing period.
We should also decide how many bridges to hardcode. Do we actually need more than one? I know that gamambel, marlowe and ln5 are interested in setting up obfsproxy. Still, I think that the original idea of a single bridge is better, so that we can get better answers for #5038 (closed). If too many users use the bundle, we can then add more bridges. So, who gets to setup the PTTB obfsproxy?
[0]: There are probably easier ways of compromising computers than exploiting obfsproxy security bugs, but I will be feeling better if obfsproxy-beta is not touching a directory authority.