robsonde has it right. If something is encrypted there must be a way to decrypt it. The algorithm used by unix for passwords is a one-way hash that cannot be reversed. JtR works by trying all possible passwords until it finds one that happens to result in the same hashed value. That is not what is meant by decryption. I will say the JtR is very clever at guessing weak passwords though.

We are required to used strong passwords here at work. To verify compliance I have been asked to run JtR. So far:

Code:

guesses: 1  time: 20:17:21:23  c/s: 324328  trying: lmpsps* - lmpsos!

After 20 days, 17 hours, 21 minutes, and 23 seconds, JtR has guessed one password. Note that the author of JtR chose the verb "guess", not "crack", not "decrypt" or anything like that. It may take decades to guess all of the remaining 186 passwords. And I downloaded dozens of carefully chosen dictionaries to help JtR out, which is the only reason that JtR even has one guess so far.