A year ago, at the outset of 2018, we (specifically, Vince Power, an experienced IT professional who also happens to be a very nice guy) took a look on the Sauce Labs blog at some of the trends that might define the world of software testing in the new year.
As all good software testers know, testing involves not just vetting something before you release it into the world, but also revisiting work that you produced previously in order to assess your accuracy and learn from mistakes you might have made.
With that attitude in mind, this blog post revisits the 2018 software testing predictions piece to consider what we got right, what we got wrong, and which predictions fell somewhere in between.
DevOps and Test Automation
The article’s first big prediction was that “Agile methodologies and DevOps concepts” would assume an even greater role in the world of software testing over the course of 2018.
Quantitatively speaking, it’s hard to confirm that this happened, because it’s not really possible to collect hard data about how Agile and DevOps are used in software testing. Plus, it’s not as if these concepts weren’t already important to software testing before 2018 began; they certainly were.
Qualitatively, however, there are signs that DevOps in particular became even more influential for software testing over the past year. One example is the release in March of the Sauce Labs Continuous Testing Cloud, which is designed to make it easier to run continuous software tests as part of CI/CD pipelines, as well as to build continuous feedback loops that incorporate test data.
Anecdotally speaking, I can also say that I heard a lot of talk in 2018 from software testing teams about increasing test automation. Most were already doing some automated testing, but wanted to automate much more. Given that automation is one of the core goals of DevOps, this is another example of the influence of DevOps over software testing.
Open source software testing tools
We also predicted that open source software testing tools would predominate in 2018, and that “there will be an increased desire to work completely within the open source ecosystem.”
I don’t know that this prediction came fully to pass. To be sure, open source tools like Selenium remain at the top of most testing teams’ lists (as they did before 2018). But closed-source tools have yet to entirely disappear.
What we can say, however, is that as open source software in general becomes even more popular, open source testing tools are rising with it. I agree that the general trend is toward building testing pipelines that run completely on open source tools, but I don’t think 2018 got us all the way there.
Moving to cloud-based testing
Cloud-based software testing was already common going into 2018, but we predicted that it would grow even more popular in the new year. That was largely because the ever-expanding number and diversity of devices, browsers and operating systems makes it increasingly infeasible to build and manage on-premises testing infrastructure that covers all of the possible environments you need to test for.
Here again, it’s hard to present hard data regarding how many more tests are being run in the cloud as opposed to on-premises these days, compared to a year ago. But I have a strong feeling that if you could collect all that data, you’d find that the prediction was true. In fact, I think it’s safe to say that on-premises test infrastructure has now become the exception. I’m not sure that I would have said that even just a year ago.
The final big prediction from a year ago was that we’d see more testing solutions that cater to Internet of Things (IoT) devices and applications in 2018. I don’t think that prediction turned out to be accurate.
Sure, the IoT continues to grow, and finding good ways to test the software that runs on it is important. But the IoT also remains a very large and loosely defined entity; the types of hardware and software that an IoT application involves can vary so widely that I don’t even know if it makes sense to talk about “IoT testing” as a singular thing right now. Maybe that will eventually change as IoT hardware becomes more standardized (if it ever does), but we’re not there yet.
IoT remains a hot topic, and it’s fun to speculate about what IoT-optimized software testing might entail, but it’s largely all still speculative at this point.
Predicting the future is always hard — even for people who specialize in making sure that software behaves as expected when you release it into the wild. All in all, most of the prevailing trends that we predicted for software testing going into 2018 turned out to more or less come true. And perhaps in the future someone will even make an accurate prediction about the IoT — We can hope!
Chris Riley (@HoardingInfo) is a technologist who has spent 15 years helping organizations transition from traditional development practices to a modern set of culture, processes and tooling. In addition to being an industry analyst, he is a regular author, speaker, and evangelist in the areas of DevOps, BigData, and IT. Chris believes the biggest challenges faced in the tech market are not tools, but rather people and planning.