I approved your profile but wanted to send you a note.
"The moral rationale for this lets-improve-internet-security-by-hacking-passwords, a bit like me burgling all my neighbours in order to have them all rushout and buy alarms and window locks, is pretty suspect but hey! the credits are excellent."
Rainbow tables are pre-computed tables for attacking password hashes. Most of the big security breaches lately such as Sony PSN stored plaintext passwords and there were no hashes to break.
Some password storage just uses a hash (such as Microsoft password storage) and no salt. This makes the hashes vulnerable to pre-computation attacks. *nix password storage has been using salts since the 70s. Lots of webapps also just store an unsalted md5, sha1, etc. Also, you have to have the hashes in the first place. Some of our sets are actually being requested by CISSPs working in infosec for audits because there is about 1 company that sells tables and they cost $1000USD for a single table set.
I begin all my talks stating that if simple measures were taken that the project wouldn't exist and I wouldn't be talking. Yes, we really did start off with the goal of giving insecure password storage a greater spotlight so that we raised awareness to the point of making our project obsolete. However, it is failing to raise the awareness that it should and mostly we're focusing especially on the NTLM side for auditing.
If you'd like to learn more about the topic I've been invited to Norway and spoke twice on the topic:
http://securitynirvana.blogspot.com/201 ... nline.html
http://securitynirvana.blogspot.com/201 ... chive.html
I'm speaking from a less math/theoretic point of view and more practical side this Tuesday,
http://dallas.naisg.org/meetings.asp, though sadly it will be audio recording only.
Our primary mirror for completed tables is hosted by a University in Italy and are working to find at least a secondary mirror or reliable hosting on a university network for torrent seeding.
James Nobis - quel