Procedural ideas.

I think what matters in programming is not how gorgeous the code is, but the ability to translate the user's needs into machine language.

Many of your ideas are the common ideas of programmers at first, and you begin to think that modular design can finish the work more easily and quickly, and the program runs more efficiently.

If you have worked in a big software company, you won't be confused. Because software companies of a certain scale have accumulated rich module and library resources for a considerable period of time, programmers only need to splice together like a buffet according to different projects, and they have a basic framework.

The most important thing is to explain the user's needs, and then design the system framework, which will save a lot of detours.

Speaking of programming, I think it is very important to require programmers in the company to have absolutely standardized programming habits, otherwise there will be many problems in teamwork and the basic library will not stand the test of time.

What's more, the modular thing you mentioned is not omnipotent, and it is closely related to the project field you often engage in. You can use them easily, but others may not use them. For example, if you are an information system, then a good query analysis module can be used in many places. User information management should be done according to the complexity, such as changing passwords regularly, such as refining permissions to columns, such as simple user passwords. Choose according to different projects.

Some people advocate automatic programming and use commercial system templates for configuration. I think it is ok for enterprise applications, but it is not desirable for software development. Because you don't know the level of the unit that provides the template, and you don't know if there are many bugs in it, but the demonstration is beautiful. Once you use this thing, your project has problems, and you can't find the specific reasons through debugging, so you can only eat the bitter fruit yourself. This is a suggestion, which is a bit off topic, but I am afraid that your thinking will reach a certain level and you will be full of praise for this method.

Finally, most projects need database as background support, so we must pay attention to the problem of database design, otherwise it is easy to make the program complicated because of unreasonable library design, or the efficiency will be seriously reduced after a period of use, leading to rework of the program. That's all. I hope you can learn something.