------------ ------------ restore content begins
- Discuss your test plan
- Do we need to test until our software is PERFECT?
- What is “good enough” for testing?
Members: Du Feng proud, Mr Chan, Fan Qiqi, Yang Qiang, Yang Qian Gang
1, brief test plan
Test plan include: Test Plan Description, test project description, test range, test tools and strategies, projects pass or fail criteria, pause and restart testing standards, testing deliverables, test assignments, environmental requirements, responsibilities , forecasting and staff training needs, schedule, risks and contingencies.
Summary general idea is: First of all members of the team responsible for writing code in the code after the completion of each module, will test the testers to see whether the function to achieve. At the same time the test team members to select test tools, test software selection and testing environment. After each module site functionality has been achieved, the members of the group responsible for the overall test site testing, including testing on site performance, a brief evaluation of the use of satisfaction with the site, test site's usability. Code written for each module members to improve the site based on the test results.
The overall approach using white-box testing methods. White box structural testing, transparent box testing, based on the logical drive test or test known as code. White box method comprehensive understanding of the internal logical structure of the program, to test all logic paths.
The overall model: W Model: User Needs - Needs Analysis - outline design - detailed design - coding - unit testing - Integration testing - Acceptance testing - unit testing design - Integrated Test Design - System test design - Acceptance test design - integration - Implementation - Delivery .
Overall step is divided into: (1) content of the test cell (mainly white box, a black box supplement )
(2) integration testing (white box and the black box binding)
(3) Software Configuration Review
2, we need to test the software until we complete it ?
Panel Discussion deemed necessary. The classic definition of software testing is: under the conditions specified in the program to operate, to discover bugs, measure software quality, and evaluate whether it can meet the requirements of the design process. Software testing is the software development based on specifications and internal structure of each stage of the program carefully designed a number of test cases (that is, enter some data and get their desired results), and use these test cases to run the program to find bugs the process of. By this step to find software testing to find differences and problems expected results and actual results, whether the measure to assess the software to meet the requirements and characteristics of the design.
"The purpose of software testing" The main problem is the need to find, but also to prevent problems and reduce the risk for many purposes. And after the product release, it is mainly based on user feedback to constantly updated perfect, but still perfect after the first test release.
3. What is "good enough" test?
You can measure the software finished and the expected effect of the vast majority of software testing is good enough test.
Front of the computer software testing, software testing must follow the first principle, that principle is not fully complied with. If the test is the principle incomplete incomplete testing process involves a large part of the principle of immunity, may play a role to help test the software. Immunity test content testers can complete immunity to its proportional with a certain degree due to factors like software testing Therefore, if you want to make software testing more fluid, more effective test results, first of all need to follow such principles, this class principles throughout the development process, ongoing testing, rather than full-time test.