Security :: Effectiveness Of A Salt For An "offline" Attack?

Context: I happened to read through an old presentation today on OpenBSD's cryptography page called "A Future-Adaptable Password Scheme". In spite of its age, it still seems relevant and useful. One of the topics it discusses is the problem of "offline" attacks, where an attacker is not slowed down by any system (or other external) security. It's attacker vs. the computational cost of guessing passwords in such a scenario.

Specific question: On several unix-like systems (including Linux), the salt helps make building rainbow tables computationally expensive. It's not enough to guess a password and hash it; the proper salt must be provided as well, or the password will not be discovered.

However, the salt (or the hashed salt) seems to be visible in /etc/shadow. For example:

Code:

foouser:$6$U9a6HdUY$U3qFDMen0wDmL0x5WHm2OWhOgzOZ4MCQxV/oY.i5RhfXCQrLifIVkBpWOd1CbCGimVCjmfxZAaud/sXDf1.mv0:14733:0:99999:7::: So in an offline attack, a rainbow table could be built using precisely that salt, correct? (Yes, I realize /etc/shadow is not readable by non-root users, but I am considering an offline attack.) Building the salt (or the hashed salt) into the hashed password seems to defeat the purpose of using a salt altogether.