A Reddit thread this week asked how people learned FreeBSD in the mid-90s and the 2000s, before YouTube, before Stack Overflow, before the endless blog-post tutorial economy. Forty-six upvotes, ninety comments, and a lot of recognizable stories. Here is my answer, unvarnished:
That was it. There was no alternative. You bought the Handbook as a printed book -- or downloaded a PostScript copy and printed it at work when nobody was looking. You read the man pages because the man pages were the manual. You broke the system, fixed it, and broke it again. If you were lucky, you had one person on IRC who would answer your question without rolling their eyes.
What That Kind of Learning Produced
Two things that are still valuable in 2026.
The first is tolerance for ambiguity. If the only way to figure something out was to read source code and run experiments, you learned quickly that the right answer was often not in any one place. You had to triangulate -- the man page said one thing, the header file said another, and the actual behavior was a third thing. The operators who came out of that era developed a habit of treating documentation as a starting hypothesis, not as ground truth. That habit survives every generational shift in tooling, because the underlying reality -- documentation drifts, code ships -- has never changed.
The second is mechanical sympathy. You knew what the operating system was doing because you had watched it do it. Kernel messages scrolling past on the console. sysctls you had changed on a hunch. Services you had debugged from rc.conf down to the binary. When a problem showed up in production years later, you had an intuition for where to look because you had already been there for your own reasons, not because a tutorial told you to go there. That intuition cannot be taught in a video. It can only be grown by the slow process of living inside a system until the system stops being alien.
The Modern Version
The point is not that learning was better before YouTube. It was not. Abundant, accessible, high-quality teaching material is an unambiguous good, and anyone who tells you otherwise is romanticizing scarcity. The point is that the underlying discipline -- read the source, break things deliberately, own the outcome -- is still available to anyone who wants it. The tools have changed. The mindset that produced the original generation of FreeBSD operators is still the fastest way to produce the next one.
If you are learning FreeBSD today and feeling overwhelmed by the options, here is the shortest path I know. Pick one feature -- jails, ZFS, or pf. Install a VM. Do not stop until you have broken it and fixed it at least three times. Keep notes. Do not look at a tutorial except as a sanity check. At the end of that exercise you will know that one feature better than eighty percent of the people writing blog posts about it. Then pick the next feature and repeat.
That is how this works. It is how it has always worked. The Handbook is free now and better than the printed version we paid for in 1998. The man pages are still the manual. The only thing missing is the willingness to sit with the problem long enough for the system to teach you.
The original Reddit thread is worth reading for the other generational perspectives -- some going back to the mid-90s, others more recent. The common thread across all of them is the same one I keep coming back to: the operators who know FreeBSD best are the ones who broke it most.
Need a FreeBSD operator who has broken it and fixed it more than three times? About Tim or schedule a consultation.