Testing is an essential tool when it comes to software development. This week I listened to the podcast “Code Newbie Podcast”. In the episode “How to make sense of the testing landscape”, they talk about the importance of testing, and why people test, what kind of testing is done. They bring in Sergei Egor who is the CEO of AtomicJar. His company works with Testcontainers a java library that supports JUnit testing. The reasoning for looking into this episode, in particular, is so I could get more familiar with how testing works and its intricacies. The experience with testing my programs is very minuscule, and so having a veteran QA developer explain the little details of testing is just what I need personally.
Sergei goes on and explains that when discussing software, new features are being implemented almost all the time and so in order to make sure that those features work as intended, testing must be done. When working on a program, it’s fair to assume that the program must meet a certain amount of requirements and so testing ensures that all those requirements are met. The whole point of Sergei’s company is to optimize testing for software companies that don’t want to create miscellaneous code just to test a certain feature or task. The question that Sergei brings up is why can’t it be done through the backend of the program? And so Sergei’s company basically specializes in these software developing phenomena.
He goes on to explain different types of testing such as automated testing, performance testing, and the most popular type of testing which is unit testing. Unit testing is basically low-level and it will tests individual methods and functions. I’ve worked with unit testing in my previous computer science courses so I am most familiar with this type of testing. I would use unit testing in java for the most part. I had a project that involved me creating a banking account class, and basically, I had to create a JUnit class that would test all my methods such as my “getFirstName” method or my “getBalance”, so on and so forth. It would basically tell me if these methods succeed or not. When Sergei started talking about unit testing it felt very familiar with assignments I’ve worked on in the past.
When he started going into performance testing which is a test that evaluates the speed and reliability of an application, it made sense on paper but with my little experience, I began to struggle with how one may go about doing something like that. With the implementation load test as Sergei mentions, load tests are used to work the application to its full potential and record its response times and request counts as he goes on to explain. Something else that I found interesting while listening to Sergei is the question of why can the developers test the code so that the company doesn’t have to hire a crew of QA engineers. In Sergei’s words, he explains that many developers don’t want to think of all possible inputs for a program they would rather have an algorithm such as property testing do it for them.
Property testing relies on properties and makes sure that a program abides by those properties. It would be interesting to see this type of testing work with my mobile application which word recorded finance information- it would typically crash when inserting a record of payments occasionally and so having an automated testing algorithm such as property testing could have helped me figure out why my program failed in some instances. Automated testing would serve me in testing my programs and their functionality, running multiple tests whenever it’s needed.
Link to “Code Newbies Podcast” S18:E4: https://open.spotify.com/episode/51gZ0yQATZQ8twsJzRb7YE?si=39e4585f9db143c0
From the blog CS@Worcester – FindKelvin by Kelvin Nina and used with permission of the author. All other rights reserved by the author.
