tedd, I think that the problem of the "duplicated hashes" in the database (in the case of two users using the same password) persists with a constant prefix in the passwords. Although the random salt portion get stored in the database concatenated to the hash, the attacker don't know the string length of the salt, making the attack very difficult. Cheers 2009/2/9 tedd <tedd.sperling@xxxxxxxxx>: > At 2:02 PM +0000 2/9/09, Stuart wrote: >> >> 2009/2/9 Michael Kubler <mdkknd@xxxxxxxxx>: >>> >>> These days SHA should really be used instead of MD5, and you should be >>> SALTing the password as well. >>> Here's a great guide : >>> http://phpsec.org/articles/2005/password-hashing.html >> >> Good advice. I would also advise against stripping and trimming >> anything from passwords. By removing characters you're significantly >> reducing the number of possible passwords. > > I read the article and didn't find any objection to it, but before we all > jump on the SHA bus, why can't we do this: > > 1. Allow the user to pick whatever password they want. > > 2. After entry, add a token string to it, such as 'a14fmw9'. > > 3. Do a M5() hash and store the hash the dB. > > When the user wants to log back in: > > 1. They enter their password. > > 2. We add the token string ('a14fmw9') to it. > > 3. Then we M5() the string and compare that hash with what's stored. That > will work. > > Furthermore, if the token string is stored in the script, or in a > configuration file, and not in the database (as suggested by the author), > then if someone obtains access to the database, all the dictionary and other > such brute force attacks will be more difficult because the hashes are more > complex than one would normally expect, right? > > If not so, then where am I wrong? > > Another scheme would be simply to use the user's password and generate a > hash. Then reverse the users password and generate another hash. Then > shuffle the two hashes, or take pairs, or quads, or any number of other > techniques to obscure the hash. As long at the process can be reversed, it > will work. > > From my limited view, a minor amount of work can throw a major monkey wrench > in any method of trying to crack a hash -- am I wrong? > > Cheers, > > tedd > > -- > ------- > http://sperling.com http://ancientstones.com http://earthstones.com > > -- > PHP General Mailing List (http://www.php.net/) > To unsubscribe, visit: http://www.php.net/unsub.php > > -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
