Keyboard for one: 9 of the most impressive solo programming efforts

Notable examples of coding brilliance that came from the mind (and hands) of one programmer

A one man band
REUTERS/Carlo Allegri

Editor's Note: If you are unable to advance to the next slide, try disabling AdBlock. We apologize for the inconvenience.

Writing software can often be a solitary task (unless you’re pair programming, of course) done by a person sitting alone at a computer. However, most developers still work as part of team, dividing up coding responsibilities for a single piece of software and generally helping each other out during the process. But sometimes programmers work truly alone on an application, game or tool and, sometimes, a lone genius coder creates something so impressive that other developers take note.

Use the arrows above to read about 9 software programs written by one person that the developer community has deemed to be particularly impressive pieces of code. Some of these programs were written to solve real problems, while others were created just for the challenge; some of them continue to live on and be used today, while others are essentially museum pieces. Regardless, what these programs all have in common is that they’re generally recognized to be feats of solo programming brilliance.

See also:

10 signs you’ve been coding too much for too long

Exposing the source: 16 pieces of classic software whose code is now accessible

Superclass: 14 of the world’s best living programmers

15 GOTO schools for the best programming jobs

Lazy like a fox: 8 ways great programmers save time and energy

Head-scratchers: 10 confounding programming language features


Fabrice Bellard’s JSLinux
ITworld/Phil Johnson

Fabrice Bellard’s JSLinux

When you’re widely acknowledged to be one of the greatest programmers in the world, you do something like write a browser-based PC emulator in JavaScript “for fun.” That’s the reason that Fabrice Bellard, known for things likes his QEMU hardware emulator, gave for doing just that when he wrote JSLinux. It emulates a 32 bit x86 compatible CPU running Linux with 32MB of RAM and includes a C compiler (Bellard’s own TinyCC) and emacs (Bellard’s QEmacs). And when you’re as smart and as good of a programmer as Bellard, you write it all in about 7,000 lines of code.

“I can't believe anyone would attempt this in Javascript, it's insanity.” Raffi256

“This is some truly amazing work” themoop

“Very impressive.” Alexandre C.

“Looking at Fabrice Bellard's accomplishments is a very humbling experience for any programmer.” Julianjon

David Horne’s 1K ZX Chess

David Horne’s 1K ZX Chess

Sinclair’s ZX81 computer, released in the UK in 1981, sold for £69 and came with just 1K of on-board memory. That didn’t stop David Horne from creating a version of chess that fit in the 672 bytes available after system overhead. Though the screen went blank each time the computer pondered its next move and the game had some quirks (the computer only had two opening moves), it was nevertheless an impressive programming achievement.

“David Horne achieved what many would even now consider impossible. He wrote a chess game, with AI, that ran on a poorly documented, buggy machine that contained only 1k of memory.” codemonkey_uk

“It really is one of the great achievements of programming, both then and now. “ John

“... the greatest achievement in the history of computer programming.” Jón Þórsteinn Petúrsson

A screenshot of BootChess
Olivier Poudade

Olivier Poudade’s BootChess

For 32 years, David Horne’s 1K ZX Chess stood as the smallest chess program ever written at 672 bytes; just recently, however, that record was broken by Olivier Poudade, a French programmer who created BootChess, which fits in a mere 487 bytes. Poudade’s program, which he wrote in assembly over a three month period, creating more than 200 versions along the way, was meant to fit on the boot sector of a floppy disk (512 bytes). Like Horne’s 1K ZX Chess, BootChess lacks some of basic chess moves such as castling and en passant; unlike Horne’s program, however, it does support queening, though not under-promotion. When it comes to miniature chess programs, it remains to be seen whether Poudade’s work is a checkmate or simply check.

“As someone who has written a chess program which even after some code optimisation still came out at ~3000 lines of C, I can tell you this is damn impressive.” Viol8

“The fact that it has any AI at all in this file size is amazing to me.” Wickwick

“... a very cool accomplishment in micro-optimization techniques.” Dutch Gun

“The code is a work of art by an exceptionally skilled craftsman.” foxyshadis

DNA sequence display at the Science Museum in London

Jim Kent’s GigAssembler

In the spring of 2000, a race was on to complete the first assembly of the human genome sequence. The race was between a private company, Celera Genomics, which wanted to patent (so it could then sell access to) the human genome sequence and the publicly funded Human Genome Project, which wanted to make it freely available to all. The latter called on Jim Kent, a programmer and biology graduate student at UC Santa Cruz to try and complete the task first. Using a cluster of 50 Linux computers, each with 256 Mb of RAM, Kent coded so hard and furiously over a 40 day period that he famously had to ice his wrists at night. Kent was able to complete his 10,000 line GigAssembler program, which mapped 400,000 bits of DNA into a sequence, just ahead of Celera, ensuring that the human genome would be in the public domain.

“The task was huge - some think it equivalent to putting a man on the moon…. ” Gary Richmond

“He's unbelievable. This program represents an amount of work that would have taken a team of 5 or 10 programmers at least six months or a year." Dave Haussler

A note saying SMTP

Daniel J. Bernstein’s qmail

Daniel J. Bernstein, or DJB, created qmail in 1995 as a secure alternative to the popular SMTP mail transfer agent (MTA) Sendmail. Written in C in a modular format, qmail’s code was intended to be less complex than Sendmail’s (it has about half as many words) because “Security holes can't show up in features that don't exist.” Famously. Bernstein has offered rewards for anyone who can find a bug in the code ($500) and bugs have been rare. qmail remains one of the most popular MTA’s in use on Unix-like systems even though the last major release of the code was in 1998.

“It's really incredible how much time qmail has been alive without any security hole bugs….” drinchev

“One bug — one bug! — was found in qmail.  djb’s programs are some of the greatest works of beauty to be comprehended by the human mind. ” Aaron Swartz

“... the defect-free-ness of the qmail code is an impressive accomplishent, head and shoulders above any other production software I have heard of; and its minimalism is without peer in my experience.” Kragen Javier Sitaker

A screenshot of the scenario Diamond Heights in RollerCoaster Tycoon

Chris Sawyer’s RollerCoaster Tycoon

Early computer games had to be coded in assembly language, in order to get the most out of early hardware. When Chris Sawyer wrote RollerCoaster Tycoon, a game that simulates the construction and management of amusement parks, in the late 1990s, assembly was no longer required. However, that didn’t stop Sawyer from writing 99% of the game code in x86 assembler. RollerCoaster Tycoon and other Sawyer-coded games were such huge hits that they earned him $30 million in royalties from Atari.

“I would never imagine that this awesome game was written in Assembly (and by a single programmer).” kuramayoko10

“He wrote that in assembly?! Jesus Christ. I think I need to go boil my brain now.” Maxpm

“That's the first time I actually believe someone deserves that much money for a single task.” Xanathos7

“He's either a genius or an alien.” Dr-Farnsworth

Ivan Sutherland’s Sketchpad

Created as part of Ivan Sutherland’s PhD thesis at MIT in 1963, Sketchpad was designed to change the way humans and computers interacted. It allowed users to draw and manipulate figures on a screen using a light pen, as well as knobs, switches and buttons. As such, it was a precursor to computer graphics, CAD software and graphical user interfaces. It also pioneered the concept of object oriented programming, by letting users to create instances (objects) of master drawings (classes). Sutherland did all this on an experimental TX-2 computer at MIT’s Lincoln Laboratory, which had 64k of memory. Being a shared resource, he often only had access to the TX-2 in the middle of the night.

“... an enormously influential program that basically invented CAD, Computer Graphics, and constraint-based systems.” Ken Fishkin

“I once asked Ivan Sutherland ‘How could you possibly have done the first interactive graphics program, the first non-procedural programming language, the first object-oriented software system all in one year?’ He said ‘Well, I didn't know it was hard’.” Alan Kay

Physics equations on the facade of a building in Stockholm

Don Knuth’s TeX

When the author of The Art of Computer Programming was unhappy with the proofs for the second edition of the second volume of the book, he took matter into his own hands and designed a new digital typesetting system specifically for technical publications. Knuth wrote the initial version of TeX during a sabbatical in 1977-78 using the Stanford Artificial Intelligence Language (SAIL) on a PDP-10 computer. TeX is famous for having very few bugs, but for years Knuth paid cash bounties to people who found any in his code.

“Since TeX was written back when computers had less memory and were slower, the code is actually pretty tricky to be as efficient as possible. And yet it is basically bug-free.” Victor Eijkhout

“don't forget that he wrote TeX ON PAPER in a notebook completely then just 'typed it in'” Kevin Won

“... it is simply amazing how few bugs there are in Tex. ... I believe the quality of TeX is due to Knuth's genius....” poliopteragriseoapte

Screen shot of GNU Emacs

Richard Stallman’s GNU Emacs

Richard Stallman was one of the creators of the original EMACS text editor, writing it in PDP-10 assembler, at MIT in the mid-1970s. In 1983, Stallman began work on the GNU operating system and wanted a version of Emacs for his new Unix-like system. He intended to build a free version of Emacs for GNU based on a (proprietary) version of Emacs created for Unix by James Gosling. Gosling’s Emacs, however, did not include a true Lisp interpreter, which Stallman wanted, so he wrote a new version of Emacs from scratch using C by himself, including a Lisp interpreter. GNU Emacs was first released in 1985 and has since been ported to many different operating systems.

“He did the initial emacs on his own because the concept was just too complex to articulate to anyone else.” Tim Post

“VI is highly overrated.  Stallman's GNU Emacs was much better code and much more involved.” David Brower