Exploratory test
The advantages and disadvantages of exploratory testing are more obvious:
- The advantage of exploratory testing is flexibility and freedom, while learning, design and execution, so as to quickly detect defects in a very short development cycle.
The requirements of the testers are very high. We need the kind of testers who "come when you call, fight when you come, and win when you fight". In practice, it often happens that testers do not have enough understanding of the product and insufficient exploration, leading to missed tests. - As exploratory testing relies on a single tester, it needs to make up for it in terms of manageability and monitorability.
- The weak documentation in exploratory testing is reflected in most cases where traditional test cases are not written. Instead, it may be called "test point", "test idea", etc., and use short text to describe the goal of the exploration.
The use of mind maps in agile testing
- Using mind maps to organize test points is a method often used in exploratory testing
testing method
Guide test method: City maps usually mark some popular attractions. Popular areas are very important in the test. In the process of each release, the core functions must be covered (refer to the user manual)
Selling point test method: Encourage testers to watch demos that are sold to customers and understand which functions are the biggest selling points for customers from a sales perspective. These points may not be core functions, but they are worth treating as core functions.
Landmark test method: When planning visits when traveling, flags are usually placed on the map. This is a landmark, but there is no stipulation on the order of visiting landmarks. Different testers will choose different orders. Experienced tests will take some weird paths based on their understanding of product architecture and technology, but they are more likely to find defects.
Late night test method: After the city lights are dimmed, it will gradually quiet down at midnight, but the software will not. Testers should always pay attention to the auxiliary work done by the software during the cold time period, such as data backup and archive, maintenance and monitoring tasks, and timing tasks. and many more
All-night testing method: There are always busy places in the bustling city, such as nightclubs, they never interrupt, can the application persist to the end? When it is faced with continuous calls, input, re-reading and writing operations, if the running time is long enough, problems are likely to occur, the memory will need to be recycled, data needs to be cleared, never close it, and keep running (automation)
Long-path testing method: regard the most buried page of the application as the test target, and observe each page back and forth
Cancellation test method: Stop immediately after starting, especially some time-consuming functions such as backup and restoration, search, and cancel immediately after starting. A little divergence can also become, start a time-consuming operation, without stopping immediately start another time-consuming operation, in order to detect the self-clearing ability of the program
Lazy man test method: choose as little input as possible, no input can be input, no modification can be made without modification, and observe whether the program responds correctly
Rebellion testing method: Rebellious thinking requires inputting the most unlikely data, using reverse thinking to input data that violates the rules, or inputting/operating in incorrect order
Obsessive-compulsive disorder test method: Forcing software to receive the same data over and over again and perform the same operations repeatedly. The most important feature is repetition. This kind of thinking often breaks the developer's design code thinking. They expect you to follow the steps, but they don't think about how you should do the first step repeatedly.
Limit test method: operate one function in rapid succession, and operate two functions quickly at the same time
Scene deletion method: omit an intermediate step in the operation process
Scene operation substitution method: some users like to switch the focus with the mouse, and some users like to switch the focus with the keyboard
Destruction method: In the user's environment, the environment on which the software runs has been destroyed
Manage exploratory testing
A test management method called test session is used. A test session is defined as a test cycle that can last from 60 minutes to 120 minutes without being interrupted. The process is as follows:
- Research session: learn and understand how the department app works
- Analysis session: evaluation of functions or features
- In-depth coverage: unusual situations, scenarios and interactions