I wonder how they define "classics" - two of my cherished titles from MIT Press are "IBM's Early Computers" (1985) and "IBM's 360 and Early 370 Systems" (1991).
They're examples of what I consider "The Perfect Computer Book".
Depth of information, organization, etc. I wish there were more books like these covering other platforms - like one about DEC's minicomputers, another covering Cray, etc.
Plus they're nice and thick, without a third of it being indexes and footnotes.
Another vote that both of these are great books - richly informative, classics of history of computing/technology. I own hardcover 1st editions of both.
Yes, I read it in the mid-80s and loved it, and own multiple copies. Steven Levy's _Hackers_ is a similarly good book.
I found out years later that one of the moderators on a discussion site I frequent is Tom West's daughter, and I sold her a laptop at one point.. Isn't it weird how "small" the world can be?
IBM's actual mainframe-product manuals (hardware and software) were frighteningly thorough, as I recall... even those that were probably read in detail a few hundred times (ever, anywhere) at most.
Yes, I cut my teeth in technical reading on the 370 PoPs manual (Principals of Operations). IIRC it was the authoritative assembler reference. Many IBM manuals had a distinctive turn of phrase for operations which would crash. I can't remember exactly what it was, but something like "results will be unpredictable".
Nothing starts your day quite like a 3:00AM call to come in and resolve an "Abend - S0C7".
In the Production Control Center (liason between operations and the programming organizations), we loved S0C7s because they were pretty much always program errors rather than difficult-to-diagnose hardware issues.
They're examples of what I consider "The Perfect Computer Book".
https://mitpress.mit.edu/books/ibms-early-computers
https://mitpress.mit.edu/books/ibms-360-and-early-370-system...