Myths and Misconceptions in the History of Computing

---

🕹️ Games & Arcades

Space Invaders caused a national 100-yen coin shortage in Japan. Myth. The game (Taito, 1978) was a real phenomenon, but the claim that it drained Japan’s 100-yen coins is an urban legend. Mintage rose sharply in 1979–1980, but from an unusually low 1978 baseline, and numismatic research attributes the increase to coin hoarding and melting after the 1967 switch from silver to cupronickel, not to arcades. Operators emptied their cabinets regularly, keeping coins in circulation. → The History of the Video Game Industry

Atari buried millions of E.T. cartridges in the desert because it was the worst game ever made. Myth (distorted). The burial was real (a 2014 dig at the Alamogordo, New Mexico landfill confirmed it) but it was a routine tax write-off and warehouse clear-out of ~700,000 cartridges and other stock in 1983, not a shamed disposal of one hated game. The excavation recovered many titles (Pac-Man, Centipede, and others), not just E.T. The pile symbolizes Atari’s overproduction during the 1983 crash, not “the worst game of all time.” → Nolan Bushnell and Atari

The Magnavox Odyssey only worked on Magnavox televisions. Misconception. The first home console (1972) connected to the antenna terminals of any standard TV. The belief that it needed a Magnavox set was fostered by the impression Magnavox dealers left (it suited them to sell the console and a television together), and it held back sales of Ralph Baer’s machine. → Ralph Baer and the Magnavox Odyssey

---

🌐 Networks & the Internet

The Internet was designed to survive a nuclear war. Myth (mostly). The packet-switching research behind ARPANET was about reliable communication over unreliable links and efficient resource sharing, not specifically about surviving a nuclear strike. The “nuclear survivability” framing came partly from Paul Baran’s separate RAND work and was retrofitted onto the Internet’s origin story. → Fun Fact: The Internet Nuclear War Myth

Tim Berners-Lee invented the Internet. Misconception. He invented the World Wide Web (1989–91), HTTP, HTML, URLs, an application running on top of the Internet. The Internet itself, the TCP/IP network, was built earlier by Vint Cerf and Bob Kahn. The Web is not the Internet; it is one of many things that run on it. His own press FAQ answers the question with “Sorry, not me!” → Tim Berners-Lee and the World Wide Web

Robert Cailliau co-invented the World Wide Web. Myth. Berners-Lee states it plainly in his own FAQ: the idea, the URL/HTML/HTTP specifications, and the code of the first browser and server were his alone; Cailliau joined more than a year after the March 1989 proposal. Cailliau’s contributions were different and real: the proposal that got CERN staff allocated, 18 months of lobbying the directorate for the 1993 royalty-free release, the first WWW conference (1994), and an early Mac browser. First convert and indispensable advocate, not co-inventor. → Tim Berners-Lee and the World Wide Web

V. A. Shiva Ayyadurai invented email in 1978. Myth. He wrote a program called “EMAIL” as a teenager, but networked electronic mail already existed: Ray Tomlinson sent it across ARPANET in 1971, building on 1960s time-sharing mail. His 1982 copyright covers his program’s code, not the concept (copyright isn’t a patent). → Email: The Killer App

“The cloud” is a wireless, satellite-based system in the sky. Misconception. The name is marketing; the cloud is intensely physical, hundreds of vast data centers full of humming servers, drawing gigawatts and guzzling cooling water, stitched together by a global mesh of undersea fiber-optic cables that carry the overwhelming majority of intercontinental data. The cloud is “just someone else’s computer” in a windowless warehouse, reached over wires. → The Cloud Computing Era

Email was invented after (and because of) the World Wide Web. Misconception. The chronology is backwards. Ray Tomlinson sent the first networked email across ARPANET in 1971, establishing the @-sign address still used today; Tim Berners-Lee did not propose the Web until 1989, nearly two decades later. Email is one of the internet’s oldest applications. Webmail (Hotmail, 1996) put email inside a browser, which is likely the source of the confusion, but the medium long predates the Web it now often rides on. → Email: The Killer App

Germany was the second country outside the USA on the internet. Contested claim. The line appears in Karlsruhe’s own histories, tied to the first German email of August 3, 1984. It overstates the case: CSNET added international members in 1984, and Israel connected before Germany, which makes Germany the third CSNET country rather than the second. The genuine, uncontested facts stand on their own: Germany was among the earliest countries linked to the American research internet, and Karlsruhe was its relay. The ranking is the institutional flourish. → Germany’s IT Industry

The “deep web” and the “dark web” are the same thing. Misconception. The deep web is simply everything search engines do not index (email inboxes, banking portals, paywalled databases, intranets) the vast and mundane majority of the web. The dark web is a tiny subset reachable only through special software such as Tor, with addresses (.onion) that resolve nowhere on the ordinary internet. All dark web is deep web; almost no deep web is dark web. → Tor and Anonymity Networks

---

🌉 Places & People

Silicon Valley was an inevitable, spontaneous tech paradise. Misconception. The Valley’s rise was shaped by specific, deliberate forces (Stanford’s industrial policy, decades of military and federal funding, and a few pivotal individuals) not by geography or luck alone. The “myth of the place” obscures how much was built on purpose. → Silicon Valley: How a Place Became a Myth

Apple’s bitten-apple logo honours Alan Turing’s poisoned apple. Myth. Designer Rob Janoff had never heard of Turing in 1977; the bite is there for scale (so it doesn’t look like a cherry), and the “byte” pun was spotted only afterward. Turing’s death by poisoned apple is itself less certain than the legend, the apple was never tested for cyanide. → Alan Turing and the Enigma

Alan Turing single-handedly broke the Enigma. Oversimplification. Polish cryptographers (Marian Rejewski’s team) broke Enigma and built the first “Bomba” in the 1930s, handing their work to Britain in 1939. Turing’s redesigned Bombe and the thousands at Bletchley Park were decisive, but the break was cumulative and collective, not a lone feat. → The Polish Cryptologists Who Broke Enigma · Alan Turing and the Enigma

“Kimble” (Kim Schmitz / Kim Dotcom) was a master hacker who broke into NASA, the Pentagon, and Citibank. Myth. The exploits that built his teenage fame were largely self-promotional and never substantiated; his documented convictions were for trafficking in stolen access codes (1998) and stock manipulation (2002). His talent was publicity, not penetration. → Kim Schmitz (Kim Dotcom)

Ada Lovelace was unambiguously “the world’s first programmer.” Contested. Her 1843 Note G (the Bernoulli-number algorithm) is widely recognised as the first published program for a general machine, but historians debate how much was hers versus a formalisation of Babbage’s earlier worked examples. She did find and fix an error in his version (real, independent contribution) but the bare superlative flattens a genuine debate. → Ada Lovelace and the Analytical Engine

---

🏢 Companies & Founders

Bill Gates said “640K ought to be enough for anybody.” Myth. Gates has repeatedly denied it and no primary source exists; the line first appeared unattributed around 1990. The 640K limit came from the IBM PC’s hardware memory map, not from Microsoft. → Bill Gates and Microsoft

Apple was founded and built in a garage. Myth. Co-founder Steve Wozniak says “we did no designs there, no breadboarding, no prototyping… no manufacturing.” The Apple I was designed at his HP cubicle; the garage was mostly a staging area. → Steve Jobs and Apple

Steve Jobs (or Apple) invented the graphical user interface. Myth. Apple invented neither the GUI nor the mouse. The windows, icons, and pointer Jobs saw at Xerox PARC in 1979 were built there in the 1970s, descended from Douglas Engelbart’s 1968 “Mother of All Demos.” Apple’s real achievement was to commercialize and refine the idea into the affordable Macintosh (1984), productization, not invention. → Steve Jobs and Apple

Bill Gates (or Microsoft) wrote MS-DOS. Myth. With no OS of its own and an IBM deadline looming, Microsoft bought QDOS (“Quick and Dirty Operating System”) from Tim Paterson at Seattle Computer Products for $50,000, adapted it, and licensed it as MS-DOS, itself modeled on Gary Kildall’s CP/M. Gates’s genius was the licensing deal (selling to clone makers), not writing the code. → Bill Gates and Microsoft

Tim Berners-Lee became a billionaire from inventing the Web. Myth. He did the opposite. Berners-Lee and CERN deliberately released HTTP, HTML, and the URL into the public domain in 1993 with no patents and no royalties, so the Web would become a universal open standard rather than a proprietary walled garden. Charging for it would have fragmented the Web into incompatible licensed systems and destroyed the universality that made it valuable. He gave away what might have been the most valuable patent portfolio in history. → Tim Berners-Lee and the World Wide Web

Linus Torvalds wrote all of Linux. Misconception. Torvalds wrote and still gatekeeps the kernel, but the kernel alone is not a usable operating system. The shell, compiler (GCC), C library, and most of the userland came from Richard Stallman’s GNU Project (begun 1983), which is why many insist on the name GNU/Linux. The kernel itself is now the work of thousands of contributors and hundreds of companies; Torvalds writes a small fraction of each release. His decisive acts were starting the kernel and running its development. → Linus Torvalds and Linux

Steve Wozniak designed the Apple II single-handedly. Misconception. Wozniak was the genius behind the digital logic, the color graphics, the disk controller, the architecture were his. But turning that board into a mass-market product took others: Rod Holt designed the innovative fanless switching power supply, and Jerry Manock designed the moulded plastic case that made it look like an appliance. The machine that sold millions was Wozniak’s logic inside a product other people helped make shippable. → Steve Wozniak: The Engineer’s Apple

---

Every barcode secretly contains 666, the “Mark of the Beast.” Myth. The UPC guard bars at the edges and center superficially resemble the bar pattern for the digit 6, but they are fixed timing references that encode no value at all, and a real 6 carries a distinguishing wide white space the guard bars lack. The resemblance is coincidence, not conspiracy. → Barcodes and the UPC

⚙️ Machines & Firsts

ENIAC was the first electronic computer. Contested. The Atanasoff-Berry Computer (electronic, 1942), Colossus (1943–44), and Zuse’s Z3 (1941) all predate it on different criteria, and a 1973 court voided ENIAC’s patent as derived from Atanasoff’s work. ENIAC’s accurate claim: first general-purpose, programmable, fully electronic computer in sustained operation. → ENIAC and the First Electronic Computers

The computer mouse was invented once, by Douglas Engelbart. Misconception (incomplete). Engelbart’s prototype (1964, patented 1967/1970) has priority, but the mouse was invented at least twice. Rainer Mallebrein at Telefunken in Konstanz independently developed the Rollkugel RKS 100-86, publicly offered on October 2, 1968 (two months before the Mother of All Demos) and it already used the rolling ball that Engelbart’s lineage adopted only in 1972 at Xerox PARC. Telefunken’s patent bid was reportedly rejected for lack of inventive step, and the parallel invention was forgotten for decades. → Douglas Engelbart and the Mother of All Demos

John von Neumann single-handedly invented the “von Neumann architecture.” Misattribution. The stored-program design in his 1945 First Draft of a Report on the EDVAC drew heavily on work by the ENIAC engineers J. Presper Eckert and John Mauchly; Eckert had described storing program and data in addressable memory in early 1944. The draft circulated under von Neumann’s name alone, and the credit stuck. → Von Neumann Architecture

The “Turing machine” is a real machine Alan Turing built. Misconception. It is an abstract mathematical model (an infinite tape, a head, a table of rules) defined in his 1936 paper to reason about what is computable. Turing never built one and never intended to; it is a thought experiment. Every real computer is a physical approximation of the idea, but the Turing machine itself is pure mathematics. → Alan Turing and the Enigma

Babbage’s engines failed because Victorian machining wasn’t precise enough. Myth. The “genius defeated by primitive tools” story is largely wrong. In 1985–2002 London’s Science Museum built Difference Engine No. 2 from Babbage’s own drawings using only tolerances achievable in his era, and it worked. The projects collapsed over funding disputes, his bitter split with engineer Joseph Clement, and his compulsive redesigning, a failure of money and management, not mechanics. → Charles Babbage: The Visionary Architect

Altair BASIC was written on an Altair. Misconception. Micro-Soft owned no Altair. Gates and Allen developed the interpreter on a DEC PDP-10 mainframe running a software simulation of the Intel 8080; the code first ran on real hardware only at the MITS demonstration in early 1975. The borrowed-mainframe detail matters because the machine was Harvard’s, funded for defense research, its unauthorized commercial use triggered a university inquiry into Gates just as he was publicly calling hobbyists thieves for copying that same BASIC. → The Open Letter to Hobbyists

Moore’s Law is a law of physics. Misconception. It is neither a natural law nor inevitable physics. Gordon Moore’s 1965 article was an economic and engineering observation about transistor cost and density, a forecast, not a derivation from first principles. What kept it true for fifty years was the industry organizing around it: the semiconductor roadmap turned the prediction into a coordinated target everyone had a commercial incentive to hit, a self-fulfilling prophecy. Physics is what is now ending it, as feature sizes approach atomic scale and Dennard scaling has broken down. → Gordon Moore and Moore’s Law

Apple built its first iPhone in 1983. Misconception. A widely shared 1983 Apple concept by designer Hartmut Esslinger is captioned as an iPhone ancestor. It was a corded desk telephone fused with a Mac, operated by a stylus, a Snow White-era design study that was never built and only surfaced in 2011 from Stanford’s Apple archive. Pen input on a landline is not the capacitive multi-touch pocket device of 2007; the connection is one of ambition, not engineering. → Fun Fact: Apple Designed a Touchscreen Phone in 1983

---

🐛 Software & Bugs

Y2K was a hoax, nothing happened, so the threat was fake. Myth (prevention paradox). The quiet rollover was the result of an estimated $300–600 billion in remediation, not proof the risk was imaginary. Unfixed systems did fail in minor ways. “Nothing happened” was the outcome of the work. → The Y2K Crisis

Grace Hopper coined the word “bug” when she found a moth in a relay. Myth. A moth really was taped into the Harvard Mark II logbook on 9 September 1947 (“first actual case of bug being found”), but the term was already engineering slang: Edison used “bug” for technical faults in the 1870s. Hopper popularized the story; she did not originate the word. → Grace Hopper: The Queen of Code

Pascal was an inherently slow programming language. Misconception. The slowness was an implementation choice, not the language. Wirth’s original CDC 6000 compiler (1970) produced native machine code; the “slow” reputation came from the portable p-code implementations, above all UCSD Pascal, which compiled to bytecode run by an interpreted virtual “p-machine.” Turbo Pascal (1983) compiled straight to native code and was fast. The p-code approach was ahead of its time, not misguided: bytecode-on-a-virtual-machine returned with Java’s JVM (1995) and .NET (2002). → Niklaus Wirth and Pascal

COBOL is a dead language. Myth. Old is not dead. An estimated 800 billion lines of COBOL remain in active production (more than any other language) processing trillions of dollars in transactions daily and running the core systems of most major banks, insurers, and governments. COBOL has been declared obsolete roughly once per decade since 1970 and keeps running; the standard was last revised in 2023. The real risk is the workforce retiring faster than it is replaced, the opposite of a dead language. → COBOL: The Language That Runs Banking

---

🤖 AI & Minds

A chatbot (“Eugene Goostman”) passed the Turing Test in 2014. Overblown. It fooled 33% of judges in five-minute chats by pretending to be a 13-year-old non-native English speaker, a gimmick to excuse weak answers. Turing set no “30% in five minutes” pass mark, and the organiser was widely criticised for hyping it. It was scripted misdirection, not understanding. → The Rise of Artificial Intelligence

Artificial intelligence is a new, 21st-century phenomenon. Misconception. The term was coined by John McCarthy for the 1956 Dartmouth Conference, and the foundations go back further: Turing’s 1950 “Computing Machinery and Intelligence” and the 1943 McCulloch–Pitts artificial neuron. Neural networks and machine learning are decades old; what changed recently was enough data and compute to make old ideas finally work. → The Rise of Artificial Intelligence

HAL 9000’s name is “IBM” shifted back one letter, a hidden dig at IBM. Myth. The one-letter offset is real but coincidental: Arthur C. Clarke repeatedly denied any intent, said he would have renamed HAL had anyone spotted it before release, and explained the name as Heuristically programmed ALgorithmic computer. IBM was in fact a consultant on the film. → Computing in Popular Culture

Deep Blue beat Kasparov with human-like chess “intuition.” Myth. Deep Blue understood nothing. It won by brute force, evaluating ~200 million positions per second with custom chips and alpha-beta search. The Game 2 move Kasparov read as deep strategy may have come from a software-bug fallback. Its strength was raw computation on one narrow game, the opposite of human understanding. → Deep Blue and Computer Chess

ELIZA (1966) was the first program to generate emotional, natural-language text. Misconception. Christopher Strachey’s Love Letters program, written on the Ferranti Mark I in 1952, predates it by thirteen years and works the same way: sentence templates filled from word lists, producing text that reads as feeling with no understanding behind it. ELIZA was more sophisticated and interactive, but not first. → Christopher Strachey

---

🔮 Quotes & Predictions

IBM’s Thomas Watson said “there is a world market for maybe five computers” (1943). Myth. Unattested in any speech, document, or contemporary article; debunked as early as 1973 and first printed only in 1981. Likely a garbling of a real 1953 remark by Watson Jr. that was actually optimistic (expected 5 orders for the IBM 701, got 18). → The IBM Mainframe Era

Bellman named “dynamic programming” to hide his mathematics from a research-hating Secretary of Defense. Myth (self-mythology, contested). The story is Bellman’s own, from his 1984 autobiography: Secretary Charles Wilson supposedly loathed the word “research,” so Bellman picked a name “not even a Congressman could object to.” But Bellman used the term in print by 1952, and Wilson only became Secretary of Defense in 1953; as Russell and Norvig note, the tale “cannot be strictly true.” The name more likely reflects the era’s ordinary meaning of “programming” as planning. → Richard Bellman and Dynamic Programming

Ken Olsen (DEC) said “there is no reason anyone would want a computer in their home” (1977). Out of context. He said it, but meant a single computer controlling the whole home (lights, heat, meals), the science-fiction “home of the future,” not the personal computer. The PC sense is a later misreading. → DEC and the Minicomputer Era

🕵️ Hackers, Crime & Espionage

“Hacker” always meant a criminal who breaks into systems. Misconception. At 1950s–60s MIT a “hack” was an elegant, clever, often playful technical feat, and a “hacker” was someone who delighted in finding them, a term of admiration. The criminal sense was attached later, largely by 1980s news media covering computer intrusions. Practitioners coined “cracker” for the malicious sense to preserve the original; the conflation drives bad law and bad culture. → The Hacker Culture

The Morris Worm was built to expose security flaws (and was the first computer worm ever). Myth (two errors). Robert Morris’s own account, accepted by the court, was that the 1988 worm was meant to measure the internet’s size; the mass disruption came from a reinfection bug, not a deliberate payload. And it was not the first self-replicating program (Creeper and academic worms predate it); it was the first to spread across the internet and cause widespread damage, which is why it’s misremembered as “the first.” → Robert Tappan Morris and the First Internet Worm

John Draper discovered that a Cap’n Crunch toy whistle produces the 2600 Hz tone used to hack the phone network. Myth. The 2600 Hz signaling frequency was known to phreaks before Draper, who by his own account learned of the tone and the whistle from others in the phreaking underground, blind phreak Joe Engressia (“Joybubbles”), for one, could whistle the tone by mouth. Draper adopted the “Captain Crunch” name and popularized the blue box; he was the famous face, not the sole discoverer. → John Draper: Captain Crunch and the Blue Box

The KGB hackers stole vital nuclear secrets and gravely damaged US security. Exaggeration. The 1985–86 espionage by Markus Hess and the Koch ring was historic as the first documented state-sponsored network intrusion, but by most assessments, including Clifford Stoll’s, the material exfiltrated was largely unclassified or low-sensitivity documents reachable from networked machines, never genuinely secret weapons data. The precedent mattered; the haul did not. → The KGB Hack: Koch, Hess, and the Cuckoo’s Egg

OneCoin was a real (if failed) cryptocurrency. Myth. OneCoin had no functioning blockchain at all. Prosecutors showed the “coin” was a number in a private database the company edited at will, with a price simply announced by founder Ruja Ignatova rather than set by any market, and no way to actually spend or freely sell it. It was a $4-billion Ponzi scheme wearing crypto’s clothing, the technology investors were told they were buying never existed. → Ruja Ignatova: The Missing Cryptoqueen

Craig Wright proved (or might be) Satoshi Nakamoto, Bitcoin’s creator. Disproven. Wright’s 2016 “proof” was a signature copied from a public 2009 Bitcoin transaction, not a fresh cryptographic signature only Satoshi could produce, which he never provided despite years of incentive. In 2024 the UK High Court ruled he is not Satoshi, that his supporting documents were forged “on a grand scale,” and that he had lied to the court repeatedly. The claim is settled, not open. → Craig Wright: The Man Who Said He Was Satoshi

Hackers break into systems mainly through brilliant, complex code. Misconception. The Hollywood image (a prodigy brute-forcing a firewall) is rare in practice. The overwhelming majority of successful intrusions exploit people, not cryptography: phishing, pretexting, and other social engineering that trick someone into surrendering a password or clicking a malicious link. Verizon’s annual Data Breach Investigations Report consistently finds the human element in a large share of breaches; Kevin Mitnick called social engineering his sharpest tool. The weakest link is usually the user, not the firewall. → The History of Hacking

In the 1984 Btx hack, the Chaos Computer Club stole 134,000 DM from the Deutsche Bundespost. Myth (garbled). Two details are routinely wrong. The money was run up from the Hamburger Sparkasse (Haspa), whose Btx access Wau Holland and Steffen Wernéry had captured, not lifted from the Bundespost, which merely operated the insecure Bildschirmtext system. And it was never kept: the CCC publicly returned every mark (134,634.88 DM), notified the Hamburg data-protection commissioner, and called a press conference to expose the flaw. The point was demonstration, not theft. → Wau Holland

🔒 Security & Privacy

Passwords must be changed regularly (every 90 days) to stay secure. Outdated advice. Forced rotation backfires: users pick predictable variants (Spring2024! → Summer2024!) and reuse them. In its 2017 Digital Identity Guidelines (SP 800-63B), NIST reversed decades of guidance, don’t force routine expiry; change only on evidence of compromise, and prefer long passphrases checked against breach lists. The UK’s NCSC agrees. → The History of Passwords and Authentication

WEP Wi-Fi encryption was broken because its keys were too short (40-bit). Misconception. Longer keys did not help. The “128-bit” (104-bit) WEP fell to the same attacks as 40-bit WEP, because the weakness was in the RC4 key-scheduling and the 24-bit initialization vector reused in the clear, not the key length. The Fluhrer–Mantin–Shamir attack (2001) recovered keys passively; by 2007 the PTW method cracked a 104-bit key from about 40,000 packets in under a minute. Adding key bits to a structurally broken protocol bought nothing, which is why 802.11i deprecated WEP entirely in 2004. → The History of Wi-Fi

Open-source software is inherently secure because “many eyes” review it. Misconception. “Given enough eyeballs, all bugs are shallow” (Linus’s Law) assumes someone is actually looking. Critical, widely-used open-source code can go years effectively unread, as the Heartbleed flaw in OpenSSL (2014) and Log4Shell in Log4j (2021) showed. Open code is necessary but not sufficient for security; visibility only helps when paired with real review and funding. → The History of Hacking

“Incognito” / private browsing makes you anonymous. Misconception. Private mode only stops the local device from saving history, cookies, and cache; it does not hide activity from the websites visited, a network admin or employer, or the ISP, and it does not mask the IP address. Google’s own incognito splash screen says so, and the Brown v. Google suit alleging Google still tracked incognito users settled in 2024 with Google agreeing to delete billions of such records. Real anonymity needs Tor or a VPN, not a private window. → The Privacy War

Cookies were invented to spy on users. Misconception. The web cookie was created by Lou Montulli at Netscape in 1994 to solve a technical problem: HTTP is stateless, so a server forgets you between clicks; there was no way to keep a shopping cart or a logged-in session. The cookie gave the web a short-term memory. Surveillance was a later repurposing: third-party tracking cookies turned the same mechanism into cross-site behavioral profiling. The tool was mundane; the abuse came afterward. → The Privacy War

Cambridge Analytica swung the Brexit referendum. Myth. Cambridge Analytica pitched its services to the Leave.EU campaign and publicly boasted of a role in the 2016 EU referendum, and the boast hardened into popular legend. But after a three-year investigation the UK Information Commissioner’s Office concluded in 2020 that it found no evidence the firm was actually involved in the official Brexit campaign or misused data to influence the vote. Its self-promotion outran its proven impact. → The Privacy War

Public-key cryptography was invented by Diffie and Hellman at Stanford in 1976. Contested first. They hold the first public invention (New Directions in Cryptography, 1976), but the concept was created earlier and secretly at Britain’s GCHQ: James Ellis conceived “non-secret encryption” in 1969, Clifford Cocks worked out an RSA-like scheme in 1973, and Malcolm Williamson found a Diffie–Hellman-like exchange in 1974. The work stayed classified until 1997, so the public credit went to Stanford. → Computer Science Firsts

Macs don’t get viruses. Misconception. Apple’s own “Get a Mac” ads (2006) claimed the Mac “doesn’t get PC viruses,” and the line hardened into folklore. It rested on market share, not immunity: malware authors chased the far larger Windows install base. The 2011–12 Flashback trojan punctured it, infecting over 600,000 Macs (~1%) through a Java flaw with no password needed, and Apple, which maintained its own Java build, shipped the fix nearly two months after Oracle’s, after the infections. Apple quietly dropped the “doesn’t get PC viruses” language around 2012 and ships XProtect precisely because Macs aren’t immune. → The Antivirus Industry

The 2026 Claude Code leak exposed Claude’s AI. Misconception. On March 31, 2026 Anthropic accidentally shipped the full source of its Claude Code agent to the npm registry inside a leftover JavaScript source map. What escaped was the client software: the tool-call loop, permission layer, and integrations that run in the terminal, about 512,000 lines of TypeScript. It contained no model weights, no training data, and no safety pipeline, which is why leaked-code forks such as OpenClaude still have to point at somebody’s model API to function. A source-code leak exposes how the software is built, not the intelligence it talks to. → Famous Leaks

---