Arkisto: February 2012

Automating Testing to Revolutionize the IT-business?

27. Februaryta, 2012 | Kirjoittaja: Antti Niittyviita

In an article published in Tietoviikko in January they talk about automating testing and about the atmosphere of revolutionizing the IT-business. I opened the article with interest, but my brains skipped a beat right after the first sentence:

Software tester’s work and entire testing is about to change.

The principal meaning of testing is to bring up the flaws in the product, which have a negative impact on the end user’s user experience. Even though testing could grow as the products become more complex, the meaning of testing will not change. Is it meaningful then, to drive software tester’s work and entire testing on the road to change?

In addition to the Tietoviikko article you can hear from several sources the same song, where they sing that you should put in more efforts to automate testing. To some, automation seems to be a guideline on par with the New Testament for the rocky road of working life. You hear people swearing on it and allocating more and more resources into it. I wonder, are we on the express highway to change, or driving down the ramp to a slow and stuck suburb…

Even though test automation is worth it in certain situations, I feel like we cannot talk about testing when it comes to test automation. When you automate the tester work, we end up only with machine-based checking. There is a drastic difference between testing and machine-based checking. For as long as the end user of products are human, you should not replace the testing with machines. The machines find small and technical ‘errors’, when people on the other hand find feel-based errors. I say that an essential part of errors is found by the latter method. That is the part, which significantly influences the product’s potential to sell.

If testing increases while the number of testers decrease, we are on a very precarious road in software development. On that road, significant flaws remain undiscovered as more and more money is sacrificed on the altar of automation. It is worth it then to stop and ask: Is this equation really financially sound?

From Lissabon to Florida

14. Februaryta, 2012 | Kirjoittaja: Antti Niittyviita

The New World was discovered and documented by the Europeans for the first time in 1492. At that time, an explorer by the name of Cristopher Columbus found America when looking for a sea route to India. The find heralded a new wave of sea travel over the Atlantic.

The navigation methods used for crossing the seas were crude and most often based on following the position of the stars and the Sun. The destination was successfully reached only when the course was realigned often enough, with observations based on the navigation.

  1. Navigation
  2. Realigning the course

That is a fact that even the Vikings knew over a millennia ago. Despite this, most modern information systems projects always start from the assumption that the course is plotted only one time and that is the design phase. After that, you close your eyes, keep your thumbs up and keep on hoping.

Time after time we can read examples on the media about projects that have lasted for 5 years and then gone very wrong. The goal is hit years late and millions of euro poorer.

I would suggest that those owners of information systems projects, who dare not blindly believe in there being enough time and money, would become acquainted with this picture and would stop to think for a moment about the secrets of sea travel.

There are two ways to reach success and the goal. You can either navigate and realign the course often enough, or you can trust your blind luck.