Twitter Hack Details Revealed
Jack Spine writes "Twitter co-founder Biz Stone has confirmed both to ZDNet UK and Wired's Threat Level blog that a dictionary attack was used to hack Twitter. After the hacker distributed details on the Digital Gangster forum, celebrities such as Britney Spears and Barack Obama had their accounts defaced.
Wired spoke to the alleged hacker, while ZDNet UK got in contact with someone who had been on the Digital Gangster forum at the time."
Cracking the site was easy, because Twitter allowed an unlimited number of rapid-fire log-in attempts.
Twitter is doubly at fault here. First, it's not that hard to detect rapid-fire password attacks. Even Unix (way before Linux) knew to kick you out after 3 failed attempts. Second, they should enforce better passwords for their employees (not necessarily for regular users, that's another discussion).
He decided not to use other hacked accounts personally. Instead he posted a message to Digital Gangster offering access to any Twitter account by request.
That's where the 18-year old kid is at fault. He showed a lack of hacker ethics. Good hackers may discover an exploit, but they don't do harm.
When I hacked my university's computer network (Vax machines on Bitnet back in 1990), I did it with the knowledge of the sysadmin staff. And once you have made your point, you stand back.
--
FairSoftware.net -- geeks starting fair and open software businesses together
Slow down cowboy! It's been 1 minute since your last failed attempt to login.
Do you even lift?
These aren't the 'roids you're looking for.
Easy, increase the amount of time between the password being supplied and the pass/fail response being sent. If the script has to wait for 5 seconds to see if the password is bad, it increases the dictionary run time by a LOT. The only way around this is to run multiple iterations of the script, each with a section of the list to run. This makes them much easier to spot by other filters.
However, a legit user waiting 5 seconds for the login to complete probably won't generate a lot of complaints.
Comment removed based on user account deletion
This is one of my favourite security conundrums.
How do you limit someone's login attempts to an account without allowing an account to be denial of serviced?
Captcha - hurts young, old, and disabled users. It can also make it hard for normal users if poorly designed (as many are).
IP Limit - Very easy to bypass with a proxy list.
Hard Account Limits - Denial of service
Thus is the problem. How do you limit logins without hurting legitimate users?
One approach is to still allow the login but to insert artificial delays. Maybe your password cracker can guess several thousand passwords in one second; too bad, because the site will only allow you to try one every three seconds. Even a fairly weak password can be extremely difficult to guess this way, though it is no substitute for strong passwords that are never sent as cleartext.
It is a miracle that curiosity survives formal education. - Einstein
Comment removed based on user account deletion
wait wait wait... you're on slashdot... news for nerds... and you pay for porn?!
Please hand over your geek card on the way out.