Software testing: present and future

To comment on some technology topics with such a title (… future) could be certainly challenging. If this theme happens to be software testing and within the confines of a blog limitations, definitively it could be pretentious as well. Nevertheless, we need to have that conversation started somehow. For that purpose, I will be using my experience of few months of outreaching and talking to customers, software programmers and testers while setting up Data Camino. Those conversations involved not only software testing of course, but a good range of topics within the IT landscape from staffing to the latest technologies including Big Data ecosystem. I am committed to writing a bit more, in coming blogs, about this software testing subject relying on my 20+ year experience at the now Dell – EMC’s R&D.

Most of the topics on software testing surfacing in talking to people were mostly around the preparedness and skills of the software test engineers, automation testing, tester certifications, testing methodologies, and techniques. It should be clear that the software industry has growth substantially. So that you have an idea of how, in 1998 the software industry in the USA was contributing much less to the economy than the automobile industry. By 2000 their contributions were about the same. In 2007 (beginning of financial crisis) the software development was contributing to the economy about 40% more than the auto industry and equipment. That is a huge of a difference. The software industry is expected to continue its growth. Here is where software testing comes into play. In theory, you do not deliver your completed software project to customers (mass consumers or enterprise) without being properly tested. I said in theory, and I will come back to this at another time. We can make the math and see the logic: if software development increases, so will do the test activities.

It is not a secret how costly the testing phase in software development could be. Obviously, this depends on the size, nature of the project, and resources of the development organization. Over the years these organizations have faced some problems: a low percentage are entirely satisfied (regarding time-budget conforming to all specified functionality), projects completed with problems (including quality), or some only canceled. In trying to overcome these challenges, some software process improvement (SPI) and quality movements have emerged and play a significant role as today. Some are relatively old (TQM, Six Sigma, CMMI) and some are trying to impose a new paradigm in incorporating all stakeholders in one whole process (Agile, DevOps, BizDevOps). There is a debate on the future of the so-called context-driven approach for software testing which often looks like the opposite of these new methods.

We should revise these latest SPI trends and find how they see the role of the traditional manual tester embedded on them. This function of the tester is going to be one of the themes for our next blog. After this sort of “open statement” on the issue and promises made, I can tell you that we will have software testing for a while. Providing a clear answer to the following five traditional questions is another matter: who does the testing? What gets tested? Why the tests? How to test? , how to evaluate (pass/fail)? Demands by SPI’s core (cost, speed, quality) and new technologies will impact test engineers in different dimensions. That applies in particular to the skills set level where this professional must embrace automation as a permanent companion. At the end of the day, all come down to efficiency, accuracy, focus and trust in facing the customer’s needs. That’s what we do. We devote ourselves to excellence.