A blog on Software Testing, Quality Engineering, Tools, Conferences

Month: November 2016

Developing & Testing – or the other way around?

This week I attended a local meet-up of software testers. Michael Thiele from Saxonia Systems AG gave a talk about his view on testing and developing (original German title: “Testend Entwickeln – Entwickelnd Testen”).

The main message of the talk was that test-driven development on unit test level didn’t work for him and his team. Instead, they are now following the acceptance test-driven approach of development. Starting with the requirements in the specification of the software that is to be developed one derives the actual test cases that will later be used as an acceptance criteria. Michael emphasized that this is already done in code, not in some test case management tool nobody is looking at. Of course, the first description of the acceptance tests is pretty high-level, but it’s of good use to prepare automated checks. One defines the various test cases independent from the technologies that are used to run the tests. Then the implementation of the software or user story is done and unit tests are written directly after that followed by integration tests, if multiple components are involved.

Michael and his team made positive experience with that top-down approach of TDD, especially because developers get used to ask about the acceptance criteria instead of inventing their own. So, it fosters communication between product owner, testers and developers and its a good way to find bugs and gaps in the requirements early.

What is missing in that approach is a good overview of what actually is and isn’t tested. Michael told they tried to extract human-readable test documentation automatically from the test code. I still think that the behavior-driven approach is the better way to achieve the same. With a language like Gherkin you get the test documentation for free and can still benefit from reusing existing test methods, parameterization and so on. It even has good integration into current IDEs.

Summing up, it was a very good talk about a very interesting topic that is relevant for so many software development teams.

Upcoming: GTAC 2016

Google Test Automation Conference 2016 takes place in Sunnyvale, California, USA next Tuesday and Wednesday (November 15-16).

When attending the 2015 edition in Cambridge (Massachusetts, USA) people told me that there will be no GTAC in autumn 2016 but in early 2017. Evidently, Google decided to stick to the yearly schedule afterwards. Unfortunately, I didn’t notice that last change in plans and therefore I missed the registration deadline.

Too bad because the schedule looks pretty good. These are the talks I find most promising by just looking at the title:

  • Using test run automation statistics to predict which tests to run
  • Automating Telepresence Robot Driving
  • Need for Speed – Accelerate Automation Tests From 3 Hours to 3 Minutes
  • Docker Based Geo Dispersed Test Farm
  • How I learned to crash test a server

Let’s see what that’s all about. Also, I’m excited because two of my colleagues got their talk proposal accepted. Dan and Alex will talk about automating audio quality tests: “Can you hear me?” – Surviving Audio Quality Testing. Don’t miss that one. I already know it’s cool stuff because they let me watch the rehearsal.

The good thing is that there will be a live stream of the whole event and recordings of each session afterwards. Nine hours time difference will make it hard for me to follow the whole event live. I’ll provide my personal view on best of GTAC 2016 here in a few weeks.

© 2024 BlueTopTesting

Theme by Anders NorenUp ↑