Development and testing in software development

We know that developers and testers are right to some extent, because developers always think that my products are perfect and there are no bugs, but they always try their best to pick you up, as a result, there is a "conflict ". In the performance evaluation of development, many companies have bugs generated by thousands of lines of code. Of course, the fewer the bugs, the better, however, there is also an average number of bugs submitted every day for the performance evaluation of the test, so it seems that this contradiction is really unavoidable, because everyone needs to eat "mixed meals.

 

However, we all know that a product cannot be bug-free, more or less, large or small, and there will always be bugs. Otherwise, Microsoft will not release patches as often as this happens, even companies like Microsoft may have bugs, so we can stay calm about the existence of bugs.

 

Although this can be done right away, we still need to pay great attention to bugs. Since bugs are generated, we have always failed to think about or think wrong in some aspects, so I suggest you analyze the distribution of bugs, find out which areas are especially prone to bugs and pay special attention to them during the development process. For our company, we have previously said that we useTechexcel devsuiteSystem Management, so we will add several mandatory check items in the development, such as compatibility with different databases, support for different languages, and whether we have considered different browsers, after the development completes the code, if you do not check these items, you cannot directly submit the development task to the test for testing. In this way, you can avoid some bugs to some extent.

 

However, even if it is avoided, some bugs will still be found. Why should we test it if there are no bugs? In many companies, the number of testers is larger than that of developers. companies like Microsoft may have a 2-1 relationship. Why do we need so many tests?

On the first hand, of course, their products are too big and complicated.

Second, the quality of a product cannot be achieved by development alone, because although development can make the product, it may also be usable, but they may not be able to consider other aspects, such as user experience, for example, in stress testing, for applications in different languages, or even applications in different operating systems, development alone may not be able to cover all these aspects, even if they do, can you ensure that there are no problems in which environments? After all, development and programming are always compiled in one environment, even if he tested it, he could not test it in all environments.

Third, because a product/function needs to be tested in many external environments (operating systems, databases, browsers, and networks), another function requires a lot of tests (normal input, abnormal Input, critical, pressure value, etc.), so even if it is a function, there are a lot of places to test, not to mention the great achievements of the product. In addition, we know that a person is strong, and the test points he can think of are always limited, so I need another person to re-test a function point that has been tested (in this regard, because our company usesDevsuite
In the solutionDevtestTool to manage the test coverage, so slightly reduce the configuration of some testers ). In addition, for a developer, because the feature is implemented by him, others have discovered the problem and asked him to fix it. In fact, he can fix it quickly and the code is easy to learn, therefore, if a test is configured with a development program, the number of bugs that may be discovered cannot make the development fully busy. From the perspective of leaders, this is a waste of costs.

 

Therefore, considering these reasons, large companies usually have a lot of testers. Of course, the current situation has changed a lot. With the introduction of automated testing, manual work is required to decrease, many companies have begun to reduce manual tests. However, developers know that automated testing is far from enough if a product is complex, I believe that manual testing will still exist for a long time, at least within the scope I can see. It seems that there is no way to replace it with automated testing.

 

However, in China, most software companies I have come into contact with do not have too many configurations for testers. Of course, I do not think they neglect the software quality. They may think that the functions have been made, the developer just needs to test it. The tester just needs to run it all at the end. I believe that this is a cognitive concept of how to ensure the quality of software. I think this will ensure the quality of products. You think that will ensure the quality, however, from the perspective of our company, we still pay more attention to quality, which may also be related to our company background. Foreign companies are more in line with foreign companies. Therefore, our company's current development and testing configuration is greater than, but it is worse than Microsoft.