In a very general sense, the great systems of the past: Multics, ITS, the Lisp Machine, the Alto, UNIX(TM) and 2/4.x BSD, and early Windows NT. (Linux not so much.) Xen nowadays, perhaps. TeX, and let's add TCP/IP, Ethernet, and CHAOSnet. Public key cryptography and RSA. Colossal Cave Adventure and Zork. The Whirlwind computer.<p>Hmmm, and of course the great languages, LISP, Scheme and Clojure nowadays, C, FORTRAN (each in their niches, of course). (C++ not at all, and I've sworn never to write another line of Perl after heavy use for more than half a decade. Hopefully the same with C++, but then there's LLVM....)<p>People who I only know through their books:<p>The great students of project success and failure: Fred Brooks, Gerald Weinberg, Richard Gabriel, Joel Spolsky, Ivar Jacobson, the Anti-Patterns crew.<p>The great explainers: Guy Steele, W. Richard Stevens, Harold Ableson (actually knew him, and we share first names :-) and Gerald Sussman, Bjarne Stroustrup, Andrew Appel (compilers) and Richard Jones (GC), Paul Graham, Alan Kay, David Moon, some more Structured Programming types who's names may be less familiar: P. J. Plauger, Brian W. Kernighan for that (not the C book), and Edward Yourdon.<p>But to actually do something, it's to build something that's needed and has a shot at being enduring. From Perl scripts to manage my email, whatever a company needs to survive and thrive, to something open source.<p>I don't think any <i>particular</i> person inspired me in an overwhelming way, so much as the generally quiet competence of the programmers and system administrators I've worked with since 1977, the teachers and professors who knew their subjects, how to teach them well, and respected sincere students like myself, salesmen who could sell, small businessmen who could keep their companies afloat, etc. Solid competence with or without flair is highly underrated.