In 1947, a moth flew into an early model computer at Harvard University. The unwanted intruder wreaked havoc, causing consistent errors. After the pesky little insect was revealed as the cause of the errors, the computer’s programmer, Grace Hopper, allegedly coined the term “computer bug.” The meaning of “computer bug” has changed over the decades. Today, instead of winged creatures, bugs are coding errors. Some are relatively benign, and sometimes even helpful. Others are calamitous. Modern history is full of coding horror stories. Here are some of the bugs that have changed history, and not for the best.

1. The Mariner I space probe

In July of 1962, a few years before we landed on the moon, the Mariner I space probe went horribly off course almost immediately after takeoff. The spacecraft was supposed to fly by Venus. Had engineers not pulled the kill switch and had the $18.5 million rocket self-destruct just six seconds after takeoff, it could have landed in the middle of North Atlantic shipping lanes or even in a populated area. The near-disaster was ultimately blamed on a transcription error. The code was originally written on paper, but it was incorrectly transcribed into the computer. Five weeks later, the Mariner 2 space probe completed the mission.

2. The very first worm

A computer worm, which replicates itself so it can spread to other computers, is generally thought of as malicious (and usually, it is). The very first worm, though, appeared in 1988, about a year before the invention of the World Wide Web. It was called the Morris Worm. Robert Tappan Morris was a graduate student at the time at Cornell University. He designed the worm, which was meant to find security flaws. The worm, which was sent from an MIT computer, spread through the internet and targeted computers running on the Unix operating system. While it didn’t cause a lot of harm, it did slow the internet down to a crawl and it delayed emails for days. Morris was later convicted and sentenced to three years of probation, 400 hours of community service, and a $10,050 fine. To this day, we have no idea how much the Morris Worm cost organizations and industries. Estimates range from $250,000 to $96 million.

3. AT&T network crashed

In 1990, cell phones were still in their relative infancy. Most people only had landlines, and many used AT&T as their long-distance carrier, so you can imagine the chaos when more than half of AT&T’s network crashed for nine hours, leaving 60 thousand people in New York without long-distance service. The AT&T bug was like a game of dominoes gone wrong. As switches crashed, they rebooted, which caused nearby switches to crash, and on, and on, and on. Ultimately, 114 switches went through the crash/reboot cycle almost ad nauseam. AT&T initially placed the blame with hackers, but it was a tiny coding error in a standard software update. The fix, not surprisingly, was to revert back to the old software

4. Therac-25

While many coding errors have real-life consequences, it’s rare that people actually die from them. Between June 1985 and January 1987, a poorly programmed computer-controlled radiation therapy machine named Therac-25 killed at least six people by massively overdosing them with radiation. Investigators blame difficulty performing automated software tests and a single inexperienced coder, who based the coding for Therac-25 on the coding for Therac-6 and Therac-20.

5. Y2K

Many of us remember December 31, 1999 as the night the world was supposed to end, or at least the world as we knew it. Computers, we were told, were not prepared for the new millennia. The idea was that every computer in the world could go into a state of confusion, a century of wealth could be lost, power grids could go down. It would be pandemonium everywhere. None of that happened. For most of us, it ended up feeling rather anticlimactic, but behind the scenes, a widespread programming error cost more than $300 billion. The reason? In the early 1990s, computer programmers weren’t looking very far into the future when they decided that they could save time by formatting dates with two characters for the year. 1990 was “90,” 1991, “91,” etc. Of course, that left open the question of what would happen in 2000. Would computers think it was 1900? Programmers worked for years to clean up the simple error that could have easily been prevented with a tiny bit of forward thinking.

6. St Mary's Mercy Hospital

In 2003, a coding error killed 8,500 people, or at least it did on paper. At St. Mary’s Mercy Medical Center in Grand Rapids, Michigan, a glitch in their patient management software erroneously listed 8,500 patients as dead. Fortunately, no one actually died, but it caused enormous headaches nonetheless. The system automatically issued death certificates to insurance companies and the Social Security Administration. Even though the patients’ deaths were just temporary, many were denied benefits until the hospital eventually straightened it all out.

In conclusion: Fortunately, most coders have learned from the mistakes of the past

It’s rare to find coding mistakes of such magnitude in the last decade or so, but even still, no coder is perfect.
Source: https://www.hackreactor.com

Lordpeckxy:
6. St Mary's Mercy Hospital

In 2003, a coding error killed 8,500 people, or at least it did on paper. At St. Mary’s Mercy Medical Center in Grand Rapids, Michigan, a glitch in their patient management software erroneously listed 8,500 patients as dead. Fortunately, no one actually died, but it caused enormous headaches nonetheless. The system automatically issued death certificates to insurance companies and the Social Security Administration. Even though the patients’ deaths were just temporary, many were denied benefits until the hospital eventually straightened it all out.

Dangerously hilarious.

InfinityFabric:
Dangerously hilarious.

Quite so, but if it happens at a time like this, it would be a near disaster....

Engage the Thread: What are your worst programming bugs? How did you find it? Have a Bug in your program?
Share It Lets find it for you.