It's a bit of a shame that the world at large doesn't get to see his code in NT. It is by far the most gorgeous C code I've seen. In fact, in the beginning, there have been times when I used to look up his code just to feel inspired (think of it as 'code inspiration').
Getting to meet him and work in the same team as him for the last few years has definitely been the highlight of my Microsoft career.
Also, my wife (HN username:arithmetic) will tell you that getting a autographed copy of Showstopper was one of the best gifts I've gotten her :)
I say under appreciated, because he seems like a rather quiet figure who does not do a lot of showboating or seek attention and because of those things he tends to be forgotten.
Also, is he not the lead developer on the Windows Azure Platform?
Do you recall the title of the book? I am unable to find any reference to it.
That turns out to be one of the better engineered compilers, at least according to David Conroy (author of MicroEmacs, author of the first handful of compilers for Mark Williams, hardware engineer for DecTalk and who should be on this list).
That is quite humbling to say the least.
Certainly if you wrote games for the Atari ST, Amiga or any of the 8-bit platforms you would've written it in assembler. Compilers just weren't efficient enough at producing the fastest code possible at the time.
On the other hand, Sawyer did have some rationale to continue going all-ASM, since he had been building up the engine code over the course of the decade, and when he first made it for Transport Tycoon(1994), hand-optimizing the rendering code was probably the only way to achieve a smooth high-res isometric renderer.
I think the PC world encouraged the use of higher level languages as waiting for VBL's became more important than squeezing all the juice you could because of the potential variations.
Thanks for your comment, it's inspired a new sense of respect for 486 progrmamers (25/33/50? SX, DX? - we can't worry about that, we have TIE fighters to render!)
I know others that work best in a team and definitely benefit from talking out the entire design beforehand, and then coding pieces of it together.
So yes, there's a very good chance that those programmers who think they work better alone actually do work better alone. It's probably experience talking, and not just hubris.
* (Yes, I know there's no 1 'right' answer. The answers they come up with are always as good as they get, though.)
By comparison, DJB is a force of nature when it comes to programming, but reading his code can be a little difficult (try understanding qmail's source base as an example). I have to admit though, _why's is the most readable of all truly great code, despite the fact that it's at least mostly in ruby.
That, and teaching a course where the class uncovers 44 Unix security holes is quite impressive:
"Some of his most notable contributions were the vi editor, NFS, and csh. Joy's prowess as a computer programmer is legendary, with an oft-told anecdote that he wrote the vi editor in a weekend. Joy denies this assertion. Joy's accomplishments have been sometimes exaggerated; Eric Schmidt, CEO of Novell at the time, inaccurately reported during an interview in PBS's documentary Nerds 2.0.1 that Joy had personally rewritten the BSD kernel in a weekend."
"Joy was also a primary figure in the development of the SPARC microprocessors, the Java programming language, Jini / JavaSpaces and JXTA."
"BBN had a big contract to implement TCP/IP, but their stuff didn't work, and Joy's grad student stuff worked. So they had this big meeting and this grad student in a T-shirt shows up, and they said, "How did you do this?" And Bill said, "It's very simple — you read the protocol and write the code."
Love it. Reminds me of the Feynman Algorithm mentioned here on HN a few days ago. Regarding how Feynman came up with so much brilliant, groundbreaking work, that the steps he supposedly followed were:
1. Write down the problem.
2. Think real hard.
3. Write down the solution.
L. Peter Deutsch for writing the PDP-1 version of Lisp at the age 12 (he was the son of an MIT prof and was hanging out with MIT Hackers back in the 60s) he also wrote Ghostscript
I know some truly amazing Lisp hackers—some of the most productive programmers I've ever met—who considered nsgmls an amazingly and slightly disturbing feat.
Ishaan Chattopadhyaya IMO, not many outsiders know of him but he single-handedly rewrote MapQuest Search.