Do you want to investigate the cause of BUG?
Author: kpxl
Do you want to investigate the cause of BUG? There are many controversies for this problem. Some people think that the reason for the BUG is the development of the development. Software tests can be found enough; others think that software testing can find the reason why they can find BUG as much as possible. Which point is correct? I personally think this is the benevolent, and the wise is witnessed. It will have different views in a different level of products. The views discussed here are only representative of personal opinions.
To figure out this question, first make a few definitions, first of all, what is QA? Simple from literally is Quality Assure (Quality Assurance), CMM requires the following: Safeguard system system; promote process improvement; guidance project implementation; increase transparency; review project; assistance problem solving Provide decision-making reference; conduct defect prevention; realize quality objectives. Second, what is software test, software testing is based on software development specifications and procedures for the internal structure of the program and the internal structure of the program (ie input data and its expected results), and uses these tests to perform procedures to discover Process of program error. The software tester is the executor of this process.
As can be seen from the above definition, QA focuses on not only quality, but the entire software process, the guarantee is the first process and system, that is to say only the process and system can make good products. Software testing is to provide QA personnel to QA personnel through their own activities, so that they can find out the abnormalities or institutions in the process. The task of visible software test is more than just testing, but also report the abnormal situation of the project to the QA, so it can only be reported that BUG is not enough.
In fact, the purpose of QA and software tests is the same, which is as possible to make the published product more in line with the needs of users, and do not have bugs as much as possible. The difference is just a concern of the entire software process, one is just the final quality. Therefore, in order to engage in software testing, don't pursue the cause of BUG, first make it clear that what is the benefit of the entire software process, or what is the benefit of the end quality?
For development, it is generally possible to reproduce this bug, so for those who have a chance of 100%, software testers will describe the steps of BUG as long as they describe the steps of BUG, Execute the roles of users and permissions, useful operating systems and browsers, and then write clear the actual results and expectations, basically almost, develop according to these descriptions, and know that it should be changed What is it. At that time the software tester (may not be the person who originally reported Bug), according to the description of the bug, it is clear that this bug is really changed. However, if a BUG has a chance of 100%, or a chance of occurrence under certain specific conditions is 100%, but it does not exist in general. Test staff may just discover this problem, but it will be 100%, and there is no designation of this issue when reporting BUG, and the development of this BUG is not able to reproduce, then call the test personnel, so repeated A few times, although the final problem was resolved, but for the entire project, it was a lot of time. If the problem is found. You can try more, or try to try an environment, you may find a chance of less than 100%, such as illegal data, special characters, special user privileges, special dates, or other parameters you don't know in the system. The impact, or the problem of operating system, or the problem of browser settings, and it is possible to find the version of the browser, etc., the process of finding these reasons, is a self-improvement process, but also to accumulate itself test experience. The process is also a process of proved the important process of testing the role, which is the process of certifying the value of the test. Of course, the level of testers in domestic software companies is not very high. I want to understand the development of the code and the test is more difficult, so I don't advocate the code to watch the code to test, only need to test Think more, especially when there is a problem, think about why this problem will happen, which will affect anything else in the system, and where there will be such problems, so wait until the development and modification, Submission tests can also be targeted, especially how effective regression tests are effective in the case of regression test time, and ensure that no major hidden dangers, I want to be related to development level, but the relationship is the largest Or testers are familiar with the system, and whether it has the idea of software development.
The reasons for pursuing BUG, not a good one, you need long-term exploration and summary, you will be very annoying, you may be very depressed, but you will find the fun, think about it, want to report a bug for developing a bug At this time, as the BUG's report has a detailed conditional data that happens this bug, and the development of data such as the test platform is easy to reproduce this problem, and it will have a great recognition of the test staff. I think my feelings are not a few words, I can say it!
