All bar 2 of those; I tried to read "Code Complete" in the first edition, but frankly I thought the advice was pretty superficial and I couldn't bring myself to waste the time on finishing it (although bear in mind I'd been coding large systems for many years before it was published so I wasn't exactly the target audience and thus likely to be impressed by it), and "Unix Power Tools, Third Edition" is end-user stuff rather than a developer text - I have a copy of the UNIX sixth edition source code, and did my own UNIX implementation (and worked on POSIX compliance in others) years beforehand anyway.
Those are two of the four books listed that frankly don't manage to make the grade.
Instead of "Code Complete", substitute either the
Ada 95 Quality and Style Guide, or perhaps for a more general book to make one think about the importance of clarity in exposition of code,
Nancy Leveson's great work
Safeware on general system safety. Bertrand Meyer's
Object-Oriented Software Construction, 1st edition is also far more important to read than Code Complete (alas the second edition isn't nearly as good, in my opinion).
Instead of the listed UNIX end-user guide, read Maurice Bach's
Design of the UNIX Operating System, although for a different perspective on operating system design you can always try
VAX/VMS Internals and Data Structures (I do have the less interesting
OpenVMS equivalent, but I regret passing up the chance some 15 years ago to own the original VAX/VMS version).
Throw away "Design Patterns", which is utter garbage, a pale shadow that rips off and repackages the mere tiniest sliver of what you can learn from
Smalltalk-80: The Language and its Implementation which is several hundred times more important. Alan Kay is a giant.
And as for THHGTTG, while I loved it and indeed listened to the original radio series when it was first broadcast, it's not really different to similar comedic entertainments familiar throughout the British Commonwealth like the great BBC Goon Shows or Monty Python. Instead, the books that are candidates for this slot in such lists is either (for those around my age or older) Robert M. Pirsig's classic "Zen and the Art of Motorcycle Maintenance", or for a younger audience Douglas Hofstader's "Goedel, Escher, Bach" (I found it somewhat contrived, but it's positively affected many people). Both have important things to say at the meta level; about the state of mind in which to approach things, in Pirsig's case, or about proof and beauty in Hofstader's.
And frankly, it's not right to commit the sin of leaving off Knuth's great work
The Art of Computer Programming. I have both first and second editions, and although they are somewhat dated there is still no real substitute for having read it and absorbed what they have to offer (including attempting the exercises).
Not sure what I'd drop to make room for it, though.
Two other honorable mentions that don't quite make it to the all-time hall of fame but for which I have immense affection: the collections of Martin Gardner's "Mathematical Games" column in
Scientific American (I was the sort of irritating child who took up counting to 1024 in binary or playing with
hexaflexagons and
Kaleidocycles after reading about them in his column) or perhaps
Winning Ways which serves both both as a compendium of game theory and a great collection of fun things to do (games like
Sprouts for instance are plain fun as well as mathematically interesting).