What is the test person involved in the project team?
Author: Leveret
It should be said that this is not an article. This should belong to a topic discussed. No matter if the view is correct, I hope everyone can express their own insights in the forum. Friends who feel deeply in this regard.
The reason for this topic is because the intervention of the test personnel now is to the development of the development stage - after the completion of the coding work, I think it is unreasonable, so this issue is proposed. From the big way, the current test model of the process is three: V model, W model, H model. In these three models, no matter which model, the intervention of tests is intervened in the preliminary intervention of the software development process. Testing work is intervened at this time, no matter which aspect considerations, it is more beneficial than we until the development process later.
From a cost perspective: The previous intervention may increase the cost of the tester, but this cost is obviously smaller than the post-input in the late input, the cost of the previously discovered bug's modification cost than the later stage Little, if the cost is put into the intervention time, then you should get the following conclusions: the earlier the intervention time, the smaller the cost investment, the greater it. In the demand research phase, the design phase, the coding phase, the test phase, the product shipment stage we will find bugs, but the cost of processing BUG is very different from these stages, the higher the price, the higher the price. Those who have done a project should have experience.
From the risk of project risks: If the coding phase is used as the wind and water ridge, the work before the coding phase is the previous work, and the coding phase includes the work after the encoding work is working later. One problem is discovered in the early stage of the development process, and it can greatly reduce the risk issues of the project. If a problem is found in the later period, it will inevitably face the project risk issue. Many times, most of the failed failure of a project is caused by the issues found later. Of course, this is not all the factors that fail to fail.
From other aspects, it is good for the previous intervention than the later intervention. However, the situation currently encountered is: After the demand survey is completed, the test personnel will carry out demand. At this time, I feel very confused is that the testers have not participated in demand research, let the testers conduct demand research, it is not needed. Repeat a job again? Many problem testers still have to ask the people who investigate. I think the efficiency is not very high. Moreover, there will be a phenomenon of misinterpreet.
Therefore, the intervention of testers is worth considering the person in charge of a project team. It cannot be considered from investment costs, and if you want to consider the cost, consider the cost of the entire project cycle. Project risks are a problem that project leaders must face, and the most important part on this issue is that testers are changing, and of course, some companies have QA after tested people to be responsible for quality. In fact, the words about this topic should be detailed, not the three words can be finished. Today just has a feeling. What's more, sometimes many companies' regulations are different. Sometimes some experiences don't have to come back in the past, it is a good summary, which is still required for surgery. It has not changed itself to it. The inquiry rules can only be stagnant.
I hope that our testers will not just pay the program directly after programmers write programs, and there are some documents that should not be rubbish, but they are not very complete, and then we follow the process description to make a keyboard Enter the action of the mouse click. I hope we can intervene in the project earlier. Learn more, do more, and learn more. Don't be a cheap can have a could have a could have a role. That is a sorrow! ! !
In fact, sometimes testing can also be staged, such as demand research, you can understand the needs, then wait for the developer to designer, and then submit the code, then start the test, and you can do a lot of test plan according to the requirements and design. , Test design strategies, including test programs, test case design, output, and prepare relevant test data. Then, the main point is to have a very considerable, comparative test risk, etc. The following is a few peers, for your reference:
Lu is not: the program developers are in creating code activities, so many primary programmers are very clear before writing code, what to do, what to do. However, when you start writing code, over time, the purpose is no longer so clarity, but also from the original own idea, just take care of your own logic, but not very clear about the overall business logic. Therefore, in order to ensure the correctness of the programmer, the tester should write the test case before writing the code, because the test personnel care about whether the results of the code are in line with the user's business process and whether a predetermined goal is achieved. Relatively speaking, testers understand the user's demand for software than developers. Because the test is a business process, and software development is divided into modules. Therefore, our approach is that when investigating user needs, we first write test cases to specify the operating range of program developers, so that they can not deviate from user needs in the correct road. When the user needs to change, the tester changes the case, the developer modifies the code to pass the use case.
Zhou Zhou: I mentioned the intervention stage, I did so many items, I was not so strictly defined, but it is really hard to define, because whether it is demand, design, never change, everything is changing, Even after the online application, the user thinks that it is not satisfied, it has also changed.
So based on demand analysis, summary design ... The test case written is difficult to get constant. So as long as the demand, the design has changed, and the test is risky. Even if everything is written out of installation and design, but those who are not the ultimate request, and finally, this is busy.
I thought that the test was involved, understanding and understanding the ultimate demand for the user is necessary to have no loss and suitable for China's national conditions.
Of course, if the needs analysis and design can understand and understand the demand, the work of the test can be reduced, and there will be a lot of risks. As long as the test "is the function is correct", it is not necessary to test the test of the "correct functionality".
The gradual industry background knowledge has also become an active requirement for testers.
Leveret: TO: Zhou Zhou, "The gradual industry background knowledge has also become a competition requirement for testers." This sentence is being applied more and more companies, not just test knowledge, related industries Background knowledge is also very important, demand is difficult to be fixed, and the use case is always changed, which is normal. I have been thinking about it for a long time, it is really difficult to summarize the real-interventional watershed. Still living, use it according to your own actual, don't leave the trajectory of science. The road is long and the road is long, and even the up and down. . . . . . . . . . . . . .
