9 COMPUTER BUGS WITH MAJOR CONSEQUENCES

9 Computer Bugs With Major Consequences
By Janet Burns, Mental Floss, 9 September 2016.

On September 9, 1947, programmer Grace Hopper and her Harvard University peers famously described processing issues with the Mark II electromechanical computer as a “bug.” In their case, the culprit was a dead moth stuck in a relay switch, but the word has been used more generally to describe technical and mechanical glitches since the 1870s, when Thomas Edison used the term to describe problems with his inventions.

Advances in computing, the web, and the internet have spawned some nasty bugs over the past several decades, which the National Institute of Standards and Technology estimated in 2002 was costing the U.S. economy upwards of US$59.5 billion per year. Many companies use bug bounty programs or enlist software-savvy patrons to hunt down bugs before they cause too much trouble - but despite all our efforts, more than a few have gotten the best of us.

1. Mariner 1 is brought down by one misplaced hyphen (we think)

Image credit: NASA/JPL

The National Aeronautics and Space Administration (NASA) has seen its share of disasters resulting from computer bugs - including during its first planetary mission. NASA launched the Mariner 1 on July 22, 1962, with the intent of sending the probe on a flyby of Venus. But not long after launch, the Atlas-Agena rocket booster blasting Mariner 1 into space started to lose contact with the ground signal, probably due to a faulty antenna. NASA had planned for this scenario; the rocket was supposed to reject the signals and keep on the correct trajectory. But something went wrong with the backup software, throwing off the rocket’s trajectory. There was nothing left to do but abort the mission - and so, just 293 seconds after launch, Mariner 1 was purposely blown up.

What exactly caused the rocket's software to go haywire is murky, but later reports referred to a tiny error in the guidance system’s coded instructions - variously a dropped hyphen, an “overbar transcription error,” or a misplaced decimal point, depending on the source. According to WIRED, “[officials] may also have been less than forthcoming with hard facts owing to the high-profile nature of the mission…[and] in the face of the much-ballyhooed space race with the Soviet Union, which was well underway by then.” But whatever the glitch’s true cause, sci-fi heavyweight Arthur C. Clarke likely helped Mariner 1 be forever remembered as just another casualty of missed typos that was “wrecked by the most expensive hyphen in history.”

2. Ariane 5 Flight 501’s primary and backup systems crash in 0.05 seconds

When the European Space Agency’s Ariane 5 rocket launched Flight 501 on June 4, 1996, it did so using working code from the Ariane 4. However, just 36.7 seconds after launch, the Ariane 5's more powerful engines reportedly set off an arithmetic bug in the flight computer, triggering an overflow condition that crashed both its secondary and primary inertial reference systems (the backup tanked first by 0.05 seconds). This caused the rocket’s primary processor to overpower its engines, and - 40 seconds after launch - the craft disintegrated. Thankfully, it was an unmanned test flight.

3. AT&T’s machine network repeatedly crashes itself for nine hours

Image credit: Sirgorpster/Wikimedia Commons

Communication errors don't just plague space-bound computers, as a major 1990 snafu from telecommunications giant AT&T illustrates. On January 15 of that year, a bug in AT&T’s new version of its software for controlling #4ESS long-distance switches crashed the company’s computers by creating a chain reaction of crash-and-reboot signals that repeated every six seconds. Soon, a network of 114 long-distance switches locked in a hopeless loop of self-sabotage. After approximately 60,000 people were left without long-distance service for around nine hours, AT&T was finally able to fix the glitch by swapping in an older version of the software.

4. Google thinks the entire web is malware - itself included

Image via TechCrunch

For up to 55 minutes on the morning of January 31, 2009, Google’s search engine warned users that everything on the web - including its own homepage - was malware. As Google’s then-VP of Search Products & User Experience Marissa Mayer explained on the company’s blog, an updated list of known malicious websites included a single, stray, all-too-common forward slash (or “/” ) as an entry - telling Google’s browser, in effect, that all websites were no-go areas.

5. Windows locks out paying customers for acts of piracy

Image credit: Mike Cohen/Flickr

On August 24, 2007, Microsoft unleashed its automated antipiracy processes on legitimate new Windows users thanks to a computer glitch. For as long as 19 hours, people who tried to install legally acquired copies of the operating system were informed by Windows Genuine Advantage, Microsoft’s in-house antipiracy software, that their actions and copies were illegal; new would-be Vista users also had some features shut off. According to Microsoft, what happened was that they sent pre-production code to servers that hadn’t been upgraded to account for changes to product key encryption/decryption, which meant the servers declined all activation requests.

6. The Mars Climate Orbiter arrives at its destination...and disintegrates

Image credit: NASA/JPL

Humankind’s numerous attempts to send crafts to Mars have often met with bad ends because of computer bugs or other technical glitches, and NASA’s efforts are no exception. In 1999, the agency’s US$655-million robotic Climate Orbiter probe had finally completed its journey to Mars, where it was supposed to orbit the planet and, eventually, serve as a communications relay for a future Mars lander. But when ground computers that controlled Climate Orbiter’s thrusters calculated its trajectory down to the planet using pound-seconds rather than the newton-seconds NASA expected, the orbiter entered the atmosphere and burned up.

7. World of Warcraft suffers through the “Corrupted Blood Incident”

Image credit: beketchai/Flickr

World of Warcraft is known for its highly detailed challenges, but a 2005 programming oversight led to an in-game outbreak that was a lot more realistic than either designers or players saw coming. At the time, Blizzard Entertainment had just added a new instanced dungeon area to the game that featured a monster capable of infecting player characters with a very damaging, spell-and-potion-proof plague that wasn’t supposed to leave that area.

It was soon discovered, however, that players could transmit the disease to other player characters and even some non-player ones after teleporting back to the game’s capital city - and an outbreak in the thousands ensued. But it was, at least, a learning experience: The “Corrupted Blood Incident” ended up providing a window for disease and terrorism researchers into crowd mentalities and self-preservation behaviors, with Tufts researchers saying the simulated outbreak "raised the possibility for valuable scientific content to be gained from this unintentional game error."

8. The Mars Polar Lander thinks it’s landed, plummets from 130 feet up

Image credit: NASA/JPL

NASA’s exploration of Mars hit another bug-induced stumbling block on December 3, 1999, when time had almost come for its Polar Lander to get settled on the planet. After atmospheric entry, it was supposed to jettison its solar panels and cruise-stage equipment, release its dual Deep Space 2 microprobes, and make a controlled landing on its supportive retrorockets the rest of the way down. Instead, the spacecraft’s computer seemingly interpreted forceful mid-air vibrations on the lander’s legs - likely caused by turbulence - as evidence it had set foot on Mars. So, at around 130 feet above the Martian surface, the US$120 million (before launch vehicle), car-sized bundle of technology turned its boosters off and landed with a splat.

9. A trading glitch loses Knight Capital US$440 million in 45 minutes

Image credit: Jin Li/Bloomberg

On August 1, 2012, a software glitch caused computers belonging to Jersey City-based Knight Capital Group to buy and sell shares of stocks unchecked for 45 minutes after the stock market opened. The company was forced to sell off its erroneously purchased shares the next day, losing a staggering US$440 million. According to a statement released by the company, the event "severely impacted" its capital base (a pool of funds the investment firm used in its daily business of buying and selling) and left it “actively pursuing its strategic and financing alternatives.”

Top image credit: testbytes/Pixabay.

[Source: Mental Floss. Edited. Some images added.]