Achtergrond

Why software code quality tools tend to fail

Paul Jansen is the founder and CEO of Tiobe Software, a company in Eindhoven that specializes in measuring and monitoring software code quality.

Reading time: 6 minutes

The introduction of a code quality system requires extra effort from the engineers, for the benefit of having to do less testing later on. A lot of software development companies don’t acknowledge that effort, Tiobe’s Paul Jansen contends, and blame the tooling when they don’t get the quality improvement they had expected.

A long time ago I was called during the Christmas holidays by one of our French customers. ‘Paul, it’s Benoit here. I’m incredibly upset, and I need you to look at these coding standard figures right away. They’re completely wrong.’

We had introduced coding standards a couple of years earlier at this French company, along with tooling and a nice dashboard. We’d even calculated a single figure to indicate their code’s compliance with their own C++ coding standard, and it all worked fine for the next couple of years. Then management made the unfortunate decision to define its bonuses based on this compliance factor. And now, as the year came to a close, they’d discovered that they would probably miss out on them thanks to a vast number of violations.

This article is exclusively available to premium members of Bits&Chips. Already a premium member? Please log in. Not yet a premium member? Become one for just €20 for the first year and enjoy all the benefits.

Login

Related content