Case Status
Log In


A "Lobster" is a *BIG* bug
  • RSS Feed

Last modified on 2/23/2018 2:09 PM by User.

A "Lobster" is a *BIG* bug

"In terms of status, the lobster has come a long way. Homarus americanus, or the Maine lobster, ascended from humble faire to fodder fit for royal bankquets in a relatively short one hundred years, a true success story. Prior to the nineteenth century, only widows, orphans, and servants ate lobster. And in some parts of New England, serving lobster to prison inmates more than once a week was forbidden by law, as doing so was considered cruel and unusual punishment.

Lobsters are Arthopoda, the phylum whose membership includes insects and spiders. Although lobsters are highly unsightly, the sweet, salty, sensual delight of a claw dipped into drawn butter more than compensates for the lobster's cockroachlike appearance and the work involved in extracting meat from shell. Yet in spite of prestige and high standing, the fishermen of Isle Au Haut still refer to them as "bugs".

-Linda Greenlaw, "Lobster Chronicles, The: Life on a Very Small Island"

We have no term in the computer industry for a ginormous / product-ending bug.

Take a look at your favorite labels for classifying bugs, and you will find they all have an implicit audience.

Label Audience
stop-ship product management, release team
dataloss core developers, devops 
blocker engineering management
production everyone customer facing

Why don't testers have a word of our own? A diplomatic word that says: "Hey, this (quality) sucks!"

I asked around, but couldn't find any word for this. So, I propose we call them "lobsters"

Q: Is the Zune (MP3 player) the story of a "lobster"?

On the last day of 2008, roughly 1 million Zune customers woke up to a brick. Since there were only about 3 million units sold, this simple subtraction bug in the boot code probably ended the illusions it could be Microsoft's iPod killer. Today, most people think of it as Peter Quill's walkman replacement.

The software failure marked the second time Microsoft device failed from product quality. Danger Sidekick had already lost the data of a million users. When Microsoft acquired Nokia to make Windows Mobile based Lumia phones, it was hard to be neutral on the prospects the latest Microsoft device. 

"QA" matters most when it finds the right bugs at right time

QA is one of those departments that has multiple, unrelated mandates. One of the most difficult to address is the difference between testing (finding bugs) and certification (final testing to ship). But within both those activities, there is a common, basic activity: a search for truth, and the written expression (and sharing) of that truth.

The essence of QA is different that engineering (writing code and meeting deadlines) and support (resolving issues with the customer experience). Although limited by time and constrained by the skills of the team, QA has endless possibilities in the pursuit of endless possible bugs. 

Good testers find a lot of bugs. Great testers focus on getting the right bugs fixed, before you ship.

Q: why don't testers find more lobsters?

Like "black swans", lobsters probably do not get enough attention. But why?

  • Is your organization a place that traps for  lobsters?
  • Does it understand the risks posed by unfixed lobsters?
  • Does it reward finding lobsters?

If someone asked the testers, would they have the same answer you do?