Seven software testing bug types

We are yet to discover software that is bug-free. We also frequently detect new defects. Some errors are more common than others, appearing with a degree of regularity. Today we look at seven software testing bug types which will be very familiar to many testers. Hopefully, our list will help you to perform fewer random searches but do let us know in the comments section below.

Errors with communication & documentation

eTestware are huge advocates for establishing a robust line of communication on any project. No matter the size or scope, nor the number of people involved, clear communication is key. When a tester cannot find a new feature that has been implemented, it is either due to a lack of communication or poor documentation – in the worst cases, it can be both. Clear documentation is another cornerstone for a successful testing experience, so make sure to treat it with the importance it deserves.

Top Tip

Let the software testing experts at eTestware lighten your testing load. We have huge experience in cruise, travel, healthcare and insurance. As part of theICEway, we have access to resources around the globe, and our main eTestware office is located in a true ‘tech hub’ – Estonia. All of which means that we can help on software testing projects both large and small. We collaborate with your own teams to ensure that all objectives are shared and met.

Error handling issues

Error handling testing is performed by both developers and software testers. It checks if errors that may occur in the future can be handled. In this scenario, mistakes made should be met with an informative error message. Should this not happen, there is every chance the same mistake will be made repeatedly.

Calculation Errors

These are caused by issues within the coding, mis-matches in the data type and incorrect algorithms. They can be very costly if not detected, with numerous high-profile examples. One such example would be the European Space Agency’s Ariane 5 rocket, which exploded 40 seconds after launch. The reason? An unpredictable conversion from a 64-bit floating point number to a 16-bit signed integer value. The cost? An estimated $500 million. Always check the numbers.

Errors with missing commands

There have been many instances where a pop-up window cannot be closed without performing an unwanted action(s). It may simply be that there is no button to exit, but a logical option should always be provided in such cases.

Functional errors

It is not unreasonable to assume that software should be fully functional. For example, if a ‘click here’ button is present and clicked upon, the user can justifiably expect an action to take place. If nothing happens, you have a functional error – and these are fairly commonplace, especially in complex software. Of course, some of the errors can be much larger than others!

Boundary errors

Before any type of testing is completed, testers should be aware of the limitations involved. The maximum text size, memory limit and the number of users at any one point are of huge importance. If any of these remain unknown, there are sure to be problems – such as with the infamous “Millennium bug” back in 1999. At that point, most software recognised years by their last two digits (i.e. 78 equated to 1978). The year 2000 then would have been recognised as 1900. The problem was anticipated, but a lack of exact knowledge led to circa $300 billion in costs around the world. Know your boundaries!

Syntax errors

Human error will always be a factor whenever there is an element of human interaction. Grammar and spelling mistakes, then, are often found in software programs. We have also noticed that whenever translation is involved there is often a high number of mistakes made. It is therefore vitally important to switch between languages in this scenario.

We hope this has proved of use but if you would like to go deeper, contact us to find out more…