Knuth "The Art of Computer Programming"
Abelson & Sussman "Structure and Interpretation of Computer Programs"
Hamming "Numerical Methods for Scientists and Engineers" (This is a classic.)
Brooks "The Mythical Man-Month"
Glass "The Universal Elixir And Other Computing Projects Which Failed"
Barron "Assemblers and Loaders"
Aho & Ullman "Principles of Compiler Design" (also known as the Green Dragon Book) (I tend to prefer this one to the newer Red Dragon Book by Aho, Sethi & Ullman.)
Boehm "Software Engineering Economics"
a complete collection of Dijkstra's EWD notes series (search around http://www.cs.utexas.edu)
I'm not really impressed with any of the operating systems textbooks I've seen.
Recommended:
Thornton "Design of a Computer: the Control Data 6600" (very hard to find, maybe impossible)
Kidder "The Soul of a New Machine"
Levy "Hackers"
Almost 20 years ago, Don Good's group at UT Austin delivered the Message Flow Modulator to the Navy. The program was basically a technology demonstration: it showed that a nontrivial piece of code could be designed and formally verified to be correct.
Written in the Gypsy 2.0 language, and running on an LSI-11, the Message Flow Modulator sits on a serial line and filters message traffic. Messages that match a set of rules (that are defined by the user) are passed through, messages that don't match are dropped. The program was about 1000 lines of executable code, and about 1500 lines of type declarations and specification statements. Gypsy was a PASCAL derivative, carefully pruned and extended so that every statement had precisely-defined semantics, that could be mechanically verified, using an automatic theorem prover. Don's group developed the code, and the formal correctness proof.
The acceptance test suite was developed by a different contractor, out on the West Coast. The Message Flow Modulator saw the acceptance test suite for the first time at PAX River NAS, during the actual acceptance test.
It passed. No deviations, no waivers, no ifs, ands, or buts. It passed. On the FIRST try.
This had NEVER happened before in all the history of military computing. Don's group got a LOT of attention as a result.
I worked with Don's group in the late 1970s. I learned a great deal from the experience.
Slashdot Mirror
Knuth "The Art of Computer Programming"
Abelson & Sussman "Structure and Interpretation of Computer Programs"
Hamming "Numerical Methods for Scientists and Engineers" (This is a classic.)
Brooks "The Mythical Man-Month"
Glass "The Universal Elixir And Other Computing Projects Which Failed"
Barron "Assemblers and Loaders"
Aho & Ullman "Principles of Compiler Design" (also known as the Green Dragon Book) (I tend to prefer this one to the newer Red Dragon Book by Aho, Sethi & Ullman.)
Boehm "Software Engineering Economics"
a complete collection of Dijkstra's EWD notes series (search around http://www.cs.utexas.edu)
I'm not really impressed with any of the operating systems textbooks I've seen.
Recommended:
Thornton "Design of a Computer: the Control Data 6600" (very hard to find, maybe impossible)
Kidder "The Soul of a New Machine"
Levy "Hackers"
Almost 20 years ago, Don Good's group at UT Austin delivered the Message Flow Modulator to the Navy. The program was basically a technology demonstration: it showed that a nontrivial piece of code could be designed and formally verified to be correct. Written in the Gypsy 2.0 language, and running on an LSI-11, the Message Flow Modulator sits on a serial line and filters message traffic. Messages that match a set of rules (that are defined by the user) are passed through, messages that don't match are dropped. The program was about 1000 lines of executable code, and about 1500 lines of type declarations and specification statements. Gypsy was a PASCAL derivative, carefully pruned and extended so that every statement had precisely-defined semantics, that could be mechanically verified, using an automatic theorem prover. Don's group developed the code, and the formal correctness proof. The acceptance test suite was developed by a different contractor, out on the West Coast. The Message Flow Modulator saw the acceptance test suite for the first time at PAX River NAS, during the actual acceptance test. It passed. No deviations, no waivers, no ifs, ands, or buts. It passed. On the FIRST try. This had NEVER happened before in all the history of military computing. Don's group got a LOT of attention as a result. I worked with Don's group in the late 1970s. I learned a great deal from the experience.