This week I am diving into an article by David Rodenas, PhD – a software engineer with a wide array of knowledge on the topic of Software Testing. I found the article, “Improve Your Testing #13: When F.I.R.S.T. Principles Aren’t Enough” which is one of many posts from Rodenas that helps provide insights to Software Testing that only a pro could provide. Through this article, Rodenas walks through each letter of the FIRST acronym with new meaning – not replacing what we know but instead enhancing what we know. As the author teaches us these new ways of understanding, examples are provided to look for ways in our own work in which these can be applied.

The acronym can be read as: Fast, Isolated, Repeatable, Self-Verifying, and Timely, but taking this one step further we can acknowledge the version that builds on top of this as: Focused, Integrated, Reliable, Significant, and Thoughtful. It is obvious that these definitions are not opposites of each other, they should also exist cohesively in our quest for trustworthy software.

One pro-tip that sticks out to me here is keeping your code focused by avoiding using the word “and” in the test name. When I first read it – it seemed kind of silly, but in a broad-sense it really does work. The writer relates this to the Single Responsibility Principle – by writing tests with a clear focus, our tests are fast and purposeful. Another takeaway is the importance of writing reliable and significant tests. Tests should not only validate but also provide meaningful information for what went wrong to cause them to fail. A test that passes all the time but fails to catch real-world issues is not significant. Similarly, flaky tests—ones that pass or fail inconsistently—break trust in the testing suite and should be avoided. Rodenas also emphasizes that integrating tests properly is just as important as isolating them. While unit tests should be isolated for precision, integration tests should ensure that components work together seamlessly. A good balance between both approaches strengthens the overall reliability of a software system.

Ultimately, this article challenges us to go beyond simply following FIRST principles and to think critically about our testing strategy. Are our tests truly adding value? Are they guiding us toward our goal of thoroughly tested software, or are they just passing checks in a pipeline? By embracing this enhanced approach to testing, we can ensure that our tests serve their true purpose: to build confidence in the software we deliver.

From the blog cameronbaron.wordpress.com by cameronbaron and used with permission of the author. All other rights reserved by the author.

I chose this blog because I wanted to learn about methods for ensuring thorough test coverage while minimizing effort. During my research, I stumbled upon an intriguing article on TestSigma about decision table testing. I was particularly intrigued because it offers a structured approach to managing complex decision-making scenarios in software applications. Because I am still working on my programming and testing skills, I found this technique particularly useful for systematically identifying various input conditions and their expected outputs.

Decision table testing is a black-box testing technique that allows testers to organize input conditions and associated output values in a structured table format. This method is especially useful in software applications where multiple variables affect the output. By organizing all of the factors into a decision table, testers can ensure that all possible combinations of inputs and expected outcomes are thoroughly covered, with no gaps in test coverage.

For example, consider an online banking system in which a user can only transfer money if his or her account balance is sufficient and the authentication process is successful. In a decision table, this scenario would have four different combinations:

• Sufficient balance and successful authentication.

• Enough balance and failed authentication

• Insufficient balance with successful authentication.

• Insufficient balance and failed authentication.

Each combination would be associated with the corresponding expected outcome, ensuring that all scenarios were tested. This structured approach ensures that no critical test case is overlooked, allowing testers to avoid missing edge cases or rare conditions that could lead to system failures.

I chose the TestSigma article because it explains decision table testing in an understandable, simplified manner. Although most other methods of testing may be involved, this one is broken down in depth and includes real-life examples, making it easier for even beginners to understand. The other articles I found were either too technical or too shallow and disorganized, whereas this one managed to strike a balance between the two.

Decision tables are a simple way to solve complex decision-making scenarios in software. It is structuring the input conditions and expected outcomes in order to provide more comprehensive test coverage while avoiding redundancies. The above article provides the most lucid and applicable example of the technique, making it an extremely useful tool for students like me. I’m still honing my programming and testing skills, learning structured testing methods like the decision table, so I can create very efficient, thoroughly organized, error-free tests, preparing me for a future in software development and testing.

Blog: https://testsigma.com/blog/decision-table-testing/

From the blog CS@Worcester – Matchaman10 by tam nguyen and used with permission of the author. All other rights reserved by the author.

Intro

Software testing is an essential part of the development lifecycle, ensuring that applications function correctly and efficiently. The blog post “Static Testing vs. Dynamic Testing” by TatvaSoft provides an insightful comparison and description of these two fundamental testing approaches.

Summary Of The Source

The blog post explains the key differences between static and dynamic testing, their benefits, and when to use each approach:

1. What is Static Testing? This type of testing is performed without executing the code. It involves reviewing documents and conducting code inspections to detect errors early in the development process.
2. What is Dynamic Testing? Unlike static testing, dynamic testing requires running the code to identify issues related to performance, security, and functionality. It involves techniques such as unit testing, integration testing, and system testing.
3. Advantages of Static Testing: Helps detect issues early, reduces debugging costs, and ensures adherence to coding standards.
4. Advantages of Dynamic Testing: Identifies runtime errors, ensures the software behaves as expected in real-world scenarios, and validates functional correctness.
5. When to Use Each Method? Static testing is best used in the early stages of development to catch errors before execution, while dynamic testing is crucial before deployment to validate real-world performance.

Why I Chose This Blog

I chose this blog because it breaks down the question of static vs dynamic testing really well. It has clear sections that serve their purpose of answering the important details, such as what they are, their benefits and disadvantages, and even comparing them. It’s a great blog to check out for anyone new to this by keeping it clear and understandable. 

Reflection

The blog reinforced how skipping dynamic testing can lead to undetected runtime issues, which can be costly and damaging after release. One key takeaway for me is the necessity of balancing both testing methods. Relying solely on static testing may overlook execution-related issues, while dynamic testing alone can result in avoidable errors slipping into the later stages of development. A collection of both is needed or at least most optimal, but practicing good static testing early makes it so that the dynamic testing which comes later is less prone to errors.

Future Application

I think in the future, when going about testing software, I will definitely keep in mind these two methodologies and probably incorporate them both as I think static testing is very valuable to prevent any errors or bugs before running but dynamic is really useful to actually see the end functionality as you’re testing. It’s an important topic to know as correct testing methodologies and practices keep code clean and working properly.

Citation

TatvaSoft. (2022, November). Static Testing vs. Dynamic Testing. Retrieved from https://www.tatvasoft.com/outsourcing/2022/11/static-testing-vs-dynamic-testing.html.

From the blog CS@Worcester – The Science of Computation by Adam Jacher and used with permission of the author. All other rights reserved by the author.

This week in class, we learned about test doubles. Test doubles are make-believe objects in code that take the place of actual objects for testing purposes. Test doubles are a great resource if you are working on a team project and your collaborator’s code is incomplete or not there at all. The test doubles we went over in class are dummies, fakes, stubs, and mocks. And being honest, I didn’t have the best understanding of it at first. But creating stubs in Activity 12 gave me a greater comprehension, as well as interest in it. Upon doing a little more research on stubs, I found a blog all the way from 2007 called Unit Testing with Stubs and Mocks. The blog compared stubs and mocks to each other, as well as which test double is more useful.

Stubs

Stubs are used by implementing part of your peer’s code as a concrete class. How much of the collaborator code needed depends on the minimum code needed to properly test something in the class. A stub’s job is to do nothing more than return the expected value needed for the test to pass. Stubs can also be implemented inside the test class. This saves the tester a lot of time because the stub class now becomes a separate, anonymous declaration. As a result, the same stub does not have to be reused across multiple unit tests. And since the stub becomes a separate declaration, the number of stubs used in a project becomes significantly lower.

A drawback from this however, is that the separately declared stub does not account for any changes that your peer might make to the code. This can mean a lot of work for the tester because new methods and variables may need to be declared. As a solution to this case, it is recommended that a base class is created in the test directory of your project. This way, if a collaborator makes an update to the code, you limit the amount of work you do by only updating the base class.

Mocks

Mock objects are used to obtain a high level of control over testing the given implementation. Mock object generators such as EasyMock and JMock are great tools that create mock objects based on the class being tested. If a test fails in EasyMock, it will create an obvious message. When testing with stubs, it is uncertain whether the test will fail or not. This however leads to constant changes made to the test cases. But through constant updating, the unit tester gains a greater knowledge of the implementation.

When to Use

In the case of which is better between stubs and mocks, there is no wrong answer. There are scenarios where using stubs will fit the test better, but there will also be scenarios where the usage of mocks is the better option. For example, if you need to nest a method call on the collaborator, then stub objects are the way to go. Mocks are more robust than stubs, but stubs are easier to read and maintain. However, mock users say that it is worth it to use mocks over stubs in the long term.

Reference

https://spring.io/blog/2007/01/15/unit-testing-with-stubs-and-mocks

From the blog CS@Worcester – Blog del William by William Cordor and used with permission of the author. All other rights reserved by the author.