Find below the list of best practices that need to be observed while doing software tests to derive quality products.
- We should begin our testing activities at the start of software development itself.
Understanding the scope / purpose of the project will help to judge the extent / level of tests required.
- Testers should meet the requirements in detail without missing out on any of the customer's levels before testing.
- The test figures should be updated immediately when the customer issues new requirements or changes requirements.
- test case document should cover all requirements even if some requirements are not tests. These unintended items shall be marked as unregistered. Keeping the traceability of a state-of-the-art certificate will be useful for achieving this.
- The test document should help to clearly identify the hierarchy / arrangement of the tests. There should be clear methods for arranging tests if multiple tests are available with similar steps. It is not recommended to copy and paste similar test events many times, but we can only specify additional / different steps.
- The description of each exam should be clearly written after understanding the context / module of the description. Steps should be written only after their manual implementation. Expected results should not be ambiguous. If necessary, state the assumptions / assumptions.
- Planning and creation testing document is necessary, even for short-term projects. The test document must not contain all information, but it should contain at least very basic elements such as scope, schedule, risk, environment, testing
- Development / Test / Environmental Protection Program should be clear. And it's very important to move the code and maintain the issue of code in each environment without being ambiguous / confused. Testers should know which version of code / data is accessible in each environment
- Test execution should be carefully based on the tests. It is important to use the appropriate test data. It is better to create different sets of test data during testing. The test data must include the valid format, invalid format and boundary.
The test result (pass / not) should be clearly updated for each test case. It is good to mention Actual Behavior if the test case fails. The test results should send to other parties (developers, companies / customers) daily, even though all test periods are not implemented. In this case, we should add a comment to indicate that the test is still running. Compilation document / mail in examinations must clearly specify date of implementation, environment, test name and test result.
- If most test items are failing continuously, there is no need to continue.
- It will be good if we finish the test condition ( by failing still ] in an appropriate color. But nevertheless, it's not good to identify the subject with a suitable color without specifying a position. Because it's hard to see the color status when you take one color printout of the test report.
- It's good to do some ad hoc tests in addition to the exam.
- Clear / proper communication / coordination within the test group and also with other teams (developer, client / company) is very necessary.
- The bug report should be made very clear with all the necessary information, especially with the steps / bug fixes. The bug report should help designers reproduce bugs and fix it.
- It's not good if we do all the tests manually, as manual testing will take more time / effort and it's hard to control, and also that not consistent or repeatable. It is therefore better that automated testers use test tools like QTP (Quick Test Professional). Even we can use simple shell scripts and vbscript to automate the tests.
Source by Rajamanickam Antonimuthu