Cryptography isn't a strong point of mine, so thought I'd get an opinion here. For an AES truecrypt volume, would this be considered a secure password? (note that this is not the actual password to be used, but rather an equivelant pw that has the same structure and use of special chars, etc)

Basically, would this be enough to foil any dedicated cracking efforts? I don't mean like if the NSA decided they wanted to break my password, but if some script kiddie with a powerful computer and lots of time tried, would it be secure?

It has mixed-case alphabetic, numeric, and special characters. Yep. That's very secure.

I wouldn't use anything resembling a word.

This one's debatable. Try tinkering with John sometime to see how he cracks passwords. It's very enlightening. My thoughts on januarY--99***:

• A substantial part of it is a common dictionary word.
• The case change for 'january' is not as clever as it seems. (I believe there are several default John rules that tweak case in exactly that way.)
• The special characters and digits are all repeating.

It's not a bad password, IMO. Technically you are observing good practices by choosing from different character classes. I would still be more inclined to use (the similarly memorable): jan*99-RAPIDO-

----

P.S. If anyone reading this thread decides to actually use any of these passwords, please have your noggin examined. (Password dictionaries.)

Again, that part is debatable. I'd argue that your key selection is woefully inadequate if you need to be able to defend against a dedicated offline attack.

Look into something like pwgen(1), and shoot for something with good character class diversity and 20+ bytes in length. That should create enough difficulty that they'll just smack you with a rubber-hose cryptanalysis instead.

See:
http://www.passwordmeter.com/
or search google.

Personally, for really secure passwords I use hashes, although I'm sure there are even better ways to generate good passwords.