I’ve spent a couple of hours cleaning up my home office this weekend and the results are actually noticeable. I still have too many shelves crammed with books that I will in all likelihood never need again, but I can’t bear to part with them — I am sentimentally attached to many of the computer science textbooks from school and career.
My oldest books are the little handbooks once published by DEC (Digital Equipment Corporation). I bought these in junior high school and they are badly worn, but still of value to me. At that point (1972?) it was very hard to find any information at all about computers or programming. Fortunately, the school’s computer teacher chaperoned a friend and I to downtown Boston to hunt down some books. If all else fails I can still figure out how to program a PDP-8.
I have architectural manuals for most of the chips that I’ve programmed — the Intel 8080, the MOS Technology 6502, and the Motorola 68000. I wrote substantial amounts of 6502 and 68000 code in my younger years — stories still to be documented here for posterity.
Odds are that I will never again need to write any PL/I, but I sure have good memories of my PL/I Structured Programming book. I really liked that class and that language. We used the Waterloo PL/W compiler at Montgomery College. I used to enjoy throwing in a random deck of cards (yeah, punched cards) consisting of a mix of FORTRAN, COBOL, and IBM 360 assembly language and watching the compiler turn them into some semblance of a workable program. In fact, I actually tracked down the original articles on the construction of the compiler and learned a whole lot about language processing and error recovery from reading them.
From there I went on to David Gries and his classic book, Compiler Construction for Digital Computers. I never quite grasped everything that he was talking about, but I definitely gave it my best shot and went on to acquire a whole shelf of compiler books. I doubt that I will ever write a compiler again, and if I did there are now high-level tools for the job, but I still have fond memories of the subject. I did find that most of the books gave somewhat short shrift to code generation, which I found to be the most interesting aspect of the whole program.
I’ve got the Smalltalk-80 language reference, the first three volumes of Knuth‘s Art of Computer Programming, and a zillion more. Of course I have a couple of linear feet of O’Reilly books; what developer doesn’t?
They are all old friends, and there’s no reason to let go of friends!