Readers Respond to Bugging Out

You can't please everyone. Readers raise quibbles and questions about our

The Wired News inbox was overrun with kudos, questions and quibbles following Simson Garfinkel's special report, Bugging Out, examining the top 10 software bugs of all time and efforts to correct the programming errors.

What about Y2K? Was the Russian pipeline explosion really the result of a bug? Following is a sampling of responses to that report chosen for publication by Wired News editors.

Before we start, though, we'd like to draw your attention to our online comment tool, in case you've missed it. Just click the Rants + Raves button on the toolbar located below the headline of every story page. Registration is easy -- we ask only for your name and a working e-mail address.

Now, on to the good stuff:

What about ... ?

Nice article, but I was surprised you didn't include Y2K. It has to have been the most widespread and costly bug to fix of all time.

-- Gil Taylor

What about the Patriot missile bug?

-- John Byrne

Let's not forget about the Airbus A320 fly-by-wire bug that reportedly contributed to a crash in France in the late '80s.

-- Paul Deaver

What about the Mars Orbiter that crashed? The U.S. engineers writing the navigation code were working in imperial units, but the European engineers that wrote another part of the code were working in metric units. Talk about a clash of cultures!

-- Paul Wilson

Interesting article. I am sure you will get lots of feedback about other bugs, but here is a biggie: Aug. 14, 2003. A bug in a General Electric controller causes a blackout for 50 million folks in eight states. I remember where I was when the lights went out on that day, how about you? 8-)

-- Mike Bailey

Soviet gas pipeline explosion

Interesting stories, including the Soviet gas pipeline. But the gas pipeline bug is not a real bug. It was sabotage. Bugs are mistakes and unintentional. The "bug" was clearly intentional and had the desired result (at least for the CIA).

-- Jonathan Tham

Entertaining article. However, the largest non-nuclear manmade explosion occurred in 1917 in Halifax, Nova Scotia, killing or maiming over 10,000 people. The Washington Post article you link to actually says: "The result was the most monumental non-nuclear explosion and fire ever seen from space." Of course, there weren't any space observers in 1917.

-- Paul de Man

Origin of the term "bug"

Simson Garfinkel opened his article by retelling Rear Adm. Grace Hopper's famous story about the moth in the Harvard Mark II: A technician looking for a bug in the computer found a dead moth in a relay, taped it into the system log book and wrote the words, "first actual case of a bug being found."

It's a nice story, and Hopper used to show off the log book to prove it was true, but it does not explain why we call computer problems "bugs."

Why did the technician save the moth? Why did he write what he wrote? He did it because it was funny. He did it because electrical engineers had a decades-old tradition of blaming mysterious faults on "bugs in the wires" or "bugs in the system" and here, finally, was the proof that they were right! He found an actual bug in the system.

-- Jim Large

Pros and cons

From the article: "What engineers didn't know was that both the 20 and the 25 were built upon an operating system that had been kluged together by a programmer with no formal training."

So programmers with formal training don't write bugs? I find this lack of analysis amazing.

-- Bill Dugan

Where's Microsoft?

It is with great irony that I point out that none of the top 10 worst software bugs were found in a Microsoft product! Even the "Ping of Death," which Garfinkel identified as affecting Windows, was not identified in the Computer Emergency Response Team advisory he quoted! Check that advisory carefully; Microsoft was not on the list of vendors.

After further digging, I couldn't find any instance where the Ping of Death affected a Windows operating system. Not Windows for Workgroups 3.11 (which included an IP stack), not Windows NT 3.1 or 3.5, and not Windows 95 (which had been in use for a year prior to the CERT advisory).

Kindly tell Garfinkel to stop blaming Microsoft for other vendors' problems.

-- Gordon Fecyk

The good ol' days

Ah, you guys are all too young. Should have been around in the '50s when all was new and we really screwed up.

-- Gene Racicot

The author responds

Many arbitrary decisions need to be exercised in creating any top 10 list. Although I can't go into every bug suggested by readers as to why they were or were not included, I thought I should address some of the more common questions.

Although the original version of the "top 10 bugs list" included the Millennium Bug, also known as the Y2K bug, I decided to remove that bug because it wasn't a single bug. Instead, Y2K was a large collection of systematic errors in date handling spread across the entire computer industry. If Y2K is a single bug, then it ranks up there with programmers who start writing code before they understand what problem they are trying to solve.

Several readers complained that the changed code that resulted in the Soviet gas pipeline explosion wasn't a bug at all, but perhaps a feature that had been intentionally inserted -- a feature that apparently worked quite well. I disagree. Logic bugs are errors in program implementation or design that cause them to function in ways that are not wanted or not expected. If a program malfunctions, it ultimately doesn't matter if the malfunction was the result of an intentional program alteration or a programmer who typed a period instead of a semicolon.

It's true that Rear Adm. Grace Hopper didn't coin the word "bug." But then, I didn't say that she did. Our error in reporting this tidbit was originally specifying the year as 1947 instead of 1945. Indeed, we found both years in a variety of different sources. We finally decided to go with 1945 because that is the year on the official Navy history website.

Another bug that flew into these articles was my erroneous spelling of Edsger W. Dijkstra's last name. I also forgot his middle initial -- something that he was very particular about, according to those who knew him. If you wish to learn more about EWD, I recommend reading the website that has been set up in his tribute.

-- Simson Garfinkel