I started programming in the late 1960s, and I remember Knuth’s books as being excellent references. There was a copy of the PDR, the Physician’s Desk Reference, in every doctor’s office, but I have never met a doctor who has read the thing from cover to cover. If you wanted to sort some data or reorder a tree, Knuth was there to describe and discuss the various approaches one could take, just as a doctor might check the PDR to weigh the merits of prescribing one drug or dose as opposed to another. No doctor would read the PDR from cover to cover. What would be the point?
The IBM 650 didn’t have much in the way of memory or computational power, so getting it to do anything reasonably well was a process full of constraints. It was like painting a small room. Moving the ladder would mean moving the paint cans or blocking the door. Modern computers have lots of computer power, lots of storage and lots of bandwidth. Getting them to do something useful is a more of a balancing act. More importantly, we have seen computer hardware get cheaper while programming stays expensive, so the ability to develop, deploy and maintain may often be the real constraints as opposed to cutting storage requirements in half.
It’s a very different world, but as a software person, I find it a lot more fun.