As a senior tester, there are always some industry secrets that are difficult to tell, and I am no exception. Maybe you recognize these secrets, maybe you sneer at them, but no matter what, I hope to bring you a little insight, a deeper understanding test, and go further.
1. Manual testing is irreplaceable unless it is not for human use
Artificial intelligence replaces testing. I believe everyone can see this argument from time to time. But I firmly believe that manual testing is irreplaceable, unless the software and system produced are not for human use.
People's aesthetics and habits are changing all the time, and there are even cultural, moral, and legal constraints on top of this. If you only rely on artificial intelligence to test, it is tantamount to handing the nuclear bomb detonator into the hands of a child.
2. The job of a tester is never simply to do testing, unless you are very simple
The core job of a tester is to do the test, but the tester never just does the test. Newcomers may not agree with this point of view, but if you carefully read the test ideas such as test left shift and precision test, you will find that testers are also product, development, design, and even operation and maintenance.
3. Technicians hate writing documents, but testers should be good at writing documents
For technicians, documentation can be described as the enemy. Often the above requirements are just to find a template on the Internet and fill in the content to deal with it. But for testers, documentation is the most important output of the work product. We must not only learn to write documents, but also be good at using statistics, charts and other tools, so that the work results can be displayed intuitively, so as to gain extensive support. Otherwise, no matter how much you do, without the support of documents, it can be regarded as rhetoric by others.
4. Software quality can be compromised, but the factual basis is a necessary prerequisite
I am very pessimistic that software quality can be compromised. However, compromise needs a factual basis. The testing organization needs to find bugs in the software as much as possible, and the decision-maker needs to give the scope and reasons for the compromise based on this. Otherwise, we only know about compromise and do not talk about the factual basis. Under the infection of this pessimism, the quality of testing work will be greatly reduced, and it is difficult to evaluate the consequences of compromise.
5. Software quality is not important, what matters is that users pay
There are two software in the same industry, one has a beautiful interface and easy-to-use functions, but it is thrown into the trash can, and the other has a decent interface and functions, but it is always repaired by people. Why? In the final analysis, whether there are users who pay the bill. In this case, software quality is really not important. What matters is whether the company's brand is strong, whether the advertising is strong, whether the discounts for burning money are strong, and whether users pay the bill.
Software quality is only an additional attribute of software, and it is suitable as a bargaining chip for both parties to compete for market share under the same conditions. If the conditions are not equal, trying to enter the market through quality is tantamount to nonsense. Think about why there is only one WeChat.
epilogue
Manual testing can never be replaced. Testers never do simple testing. Testers should be good at writing documents. Software quality can be compromised, but the factual basis is a necessary premise. In the end, software quality is not important, what matters is that users pay for it.
