Locating defects and ways to reproduce them

In this article, you will learn how to reproduce a defect and what to do if the bug does not occur again.

Let's start with the definition of a defect.

A defect or bug is a discrepancy between the program's actual and expected results. Bugs are found at the software testing stage when a tester checks the program performance against the expected results described in the requirements specification.

When you find a bug, don't immediately report it to the bug tracker with the description «nothing works!». Instead, reproduce the bug. Does it appear again? Then minimize the number of steps to reproduce it and make sure there's nothing redundant.

If you use specific input data, ensure that it does not contain unnecessary information. Once you understand what kind of data and actions cause the problem, summarize the problem - write a bug report summary (description).

«Steps to reproduce» is the main field in a bug report. Write down the identified steps. As already mentioned, the number of steps should be sufficient and optimal to reproduce the problem. Don't write redundant steps, and don't skip important ones. 

To report bugs correctly, you need technical skills and an understanding of the product architecture to be tested. When it comes to web testing, you should include in the bug report the error code returned by the server, look at the details with the help of Firebug, and provide detailed information.

Imagine you have found a bug yet can't reproduce it. What to do in such cases? Should you report this defect or not? 

First, you need to be sure that you are following the same steps as last time. There is a possibility that you might have missed something. 

The main reasons why a defect can't be reproduced:

1. The defect is reproduced only during the first launch of the application. To reproduce it again, re-install the application.
2. Used old system data that has not been modified during the testing process.
3. The speed of reproducing the defect. The speed of the steps should be the same. 

There are situations when a defect is reproduced with a 100% probability, but the developer cannot repeat it in their own environment. In such cases, you need to check the product's environment:

• system version;
• component version;
• system and data state.

Regardless, bugs that are not reproducible still need to be entered into the bug tracker. It is important to remember to specify the appropriate value of Reproducibility.

To prevent problems with reproducing defects and provide the development team with the maximum amount of information about the bug, you need to use several options for monitoring actions:

• Screencast is a video recording of a screen from a computer or other digital device. It is one of the most effective ways to share what is happening on the monitor screen. It allows the tester to easily show errors in the operation of any software product. Screenshots are taken with special programs, for example: Snagit, Movavi Screen Capture, Jing, Reflector, ADB Shell Screenrecord, AZ Screen Recorder.
• Live logging is a real-time recording of system logs. Log files are files that contain system information about the operation of a server or computer in chronological order, in which specific user or program actions are recorded. The following programs are available for this purpose: Visual Studio for Windows, iMazing for iOS, XCode for MacOS, Android Debug Monitor for Android.
• Take notes. Keep a record of everything you know about the bug you're dealing with. Then you will have all the information you need at hand when working with developers on the issue.
• Action recorder. There are programs that help you repeat all recorded mouse movements and actions performed on the PC keyboard. An example of such a program is Advanced Key and Mouse Recorder. Unfortunately, this method is relevant only for desktop platforms and there are no similar programs for mobile platforms.

Software bugs that are not reproducible or randomly reproduced complicate the testers' work. 

As a rule, not many bugs can be impossible to reproduce for a skilled, experienced tester, but for a novice tester, finding an explanation and steps to recreate a bug that is randomly repeated can be quite a challenge.

However, there are some bugs found during software testing that are difficult to reproduce and explain even for testing specialists.

In this case, it is recommended to review the process and try to look at it from a new perspective. 

For cases where it seems that the error is impossible to reproduce, it is necessary:

• reject theories that do not work; 
• create new theories or modify existing ones; 
• look through all the data and try to find what was missed or ignored in previous theories.

You can also use consequential methods for software testing. For example, you need to add or remove one or more factors at a time and see how this affects the frequency of a randomly reproduced bug.

In conclusion, following these simple steps to reproduce defects, you can not only detect a bug but also narrow down the area of search for a defect by the developer for successful further fixes.