[tor-talk] Possible upcoming attempts to disable the Tor network

[a]

Updateable list: one single party has keys to make Tor work or not work
Hardcoded list: lots of repos of different OS are responsible for their
users' Tor to work


Jim wrote:
> Thomas White wrote:
>> Generally some criteria for DirAuths are people who are trusted by the
>> community, aren't raided very often/have legal troubles/could be
>> legally compromised, people with an extensive understanding of how Tor
>> works and people who can secure their systems from attacks. Remember
>> every time a dirauth is added or change IP a new tor version must be
>> released as they are hardcoded so it isn't a simple process to add a
>> new one, not to mention intra-dirauth communication and verification.
> 
> I don't know about the intra-dirauth communication, but as far as the
> directory authorities being hard coded, would it make sense to put them
> in a small configuration file instead?  If for some reason it is deemed
> undesirable to allow users to alter that file, there could be a
> requirement the file be signed with one or more private keys and have
> the corresponding public keys hardcoded into the code.  It would be much
> easier and cheaper to distribute a new, small, signed configuration file
> than to release and distribute a whole new version of tor.
> 
> Jim
> 
>