Get your own Free Home Page
The
first bug
Deadly
bug
Morris’s
bug
Police
bug
Robin Hood bug
Pentium
bug
Useful
bugs
Do you have a story about a bug? Send it to me, I'll publish it here
The first computers (ENIAC etc) used gigantic boards with millions of components. They crashed often. One of the reasons were bugs (yes, real insects). Sometimes a bug would short-circuit the wires, so operators had to search for (dead) bugs. So the word bug become a synonym for computer problems.
At least four people died in 1985, exposed to fatal doses of radiation from Therac-25 accelerator (Atomic Energy of Canada Ltd.), used for radiation treatment of cancer. Software bugs caused the machines to incorrectly calculate the amount of radiation being delivered to patients.
One of the most famous bugs become known in 1988. Robert T. Morris, Jr., the Cornell University graduate student, found a security hole in UNIX and wrote a "worm" program, sort of a virus, which was supposed to spread over Internet. But the worm had a bug and multiplied 14 times faster than intended. The result - Internet was swamped and overwhelmed in a few hours. It took weeks to recover, and estimated damage was a few hundreds millions of dollars. Morris sad "It was a mistake. I'm sorry."
A computer in Paris read files on traffic violations and then mistakenly sent out letters charging 41,000 traffic offenders with crimes including murder, drug trafficking, extortion, and prostitution. It happened in 1980.
A British bank mistakenly transferred $2 billion to customers in only 1 hour, when a bug permitted payment orders to be issued twice. Since there was no way to distinguish real from duplicate transactions, the bank had to depend on the honesty of its customers to recover the money.
The Pentium bug (1994), probably the most widely reported bug in history, was an error in the lookup table used to perform floating-point division in Intel's Pentium microprocessor. Pentium would, under peculiar consequences, return a wrong answer when dividing real numbers. Probably more significant than the bug itself was the fact that Intel's reputation was hurt -- Intel knew about the problem, decided to keep it a secret, and then downplayed the defect when it was discovered independently. It is estimated that Intel may have lost upward of $400 million due to the Pentium bug.
And they did not have enough – Pentium 2 and Pentium Pro have another arithmetic bug. Not quite as dangerous, but the bug.
A long time ago, I had a HP-41CV programmable calculator. Nice little computer, with a 2 kHz microprocessor and up to 4 kilobytes of RAM. The programming language was sort of an assembly, but you could not access internal processor and other registers. But, fortunately, it had bugs... and those bugs enabled users to peek into HP-41’s internal registers and write quite a few programs that normally could not have been written. So the bugs could be useful.
Members of PPC club, independent organization of HP-41 users, did most of the research. Late 1970s were the golden era of PPC club – they documented and classified the bugs, made use of (so called) Synthetic Programming and even produced the fantastic PPC ROM with 8 KB of highly optimized synthetic code. And the documentation... imagine a 500+ pages book in an extremely fine print, documenting PPC ROM routines and HP-41C bugs. Useful bugs.
In case someone is interested, PPC club is history (I wrote about that once, but I am afraid you will not understand much of it – it is in Serbian). In 1984 the board of directors made Richard Nelson, founder and editor of PPC Journal, resign. He left and made another club, CHHU, but both PPC and CHHU were disbanded in 1988. It seems that it is dangerous handling bugs... maybe I should have taken a hint before starting this page.
Get your own Free Home Page