Messy ideas about software processes and Project Management

In my understanding, a project is a plan and work made to achieve one or more goals under certain conditions.
Theoretically, goals and conditions are objective factors. As long as objective factors are grasped and the feasibility of analysis is achieved, the objective will be achieved. However, this is not the case.

This is because the customer's understanding of the target, the understanding of the target with the project personnel, and the understanding of the target among the project members are more or less different. But for the existence of condition restrictions, the conditions are often more concealed than the target (people always like good things), leading to more different understandings.

The customer is generally familiar with its own business. The customer's understanding of the objectives and conditions depends on the customer's quality and ability. High-quality users are rare ). In this case, the role of the consultant (consultant) emerged, and they have a good foundation for specialized businesses (such as finance, manpower, logistics, etc ), with a keen sense of the characteristics of business application software, they can induce customers and fully elaborate on their actual business scenarios (stories ), in addition, it can distinguish and extract multiple business steps that appear to be the same in a real scenario. Make high-quality preparations for system design. (This is an ideal demand stage ~~) Only when the above situations are established can perfect requirements be met.

What if the business is clear? Design is the prelude to the production and operation of the product (of course, the product is not only a runable artifacts, but God only cares about it ). Of course, most of the time there is no design, only piece of thinking, and these pieces are often called "solutions" by some people, and then tell God, the "business needs" are changed to system design. (Super ERP such as SAP cannot be fully implemented, I think the degree of system componentization is generally not superb to jump from business needs to system design ), of course, some conscientious users will add the meaning of "Database Design" (is such a database design reliable ?). This situation is everywhere, and there is a lack of key project figures: system analysts. Yes, that is, our lovely "system score" personnel. It is not easy to make a system score, and it is not easy to have a good team score. What if there is no system score? One of the three stinks is Zhuge Liang, who gives full play to the advantages of the masses. Unfortunately, before the discussion, "everyone" usually did not have a complete thinking, and after the discussion, "everyone" often did not have a review. What a pity is that, every "everyone" cannot consider and evaluate the solution to every problem at a sufficient height. As a result, the problems behind the problems or new problems brought about by the solution will be continuously encountered in the future, the sooner you encounter it, the more difficult it is to solve it. (The larger the game, the more bosses are behind it ). (Let's look back at the initial assumption that the business is clear. If this assumption is not found, then ......)

Haha, it's time for development. Let's assume that the system design is perfect. (Currently, the common practice is to design in detail by developers, or embedCode. If there is no decent system design, you may not care about the detailed design at all .) The development language is the minimum granularity of development. Different people may write different codes for the same problem. Why? "I have finished writing it, but it will not be my maintenance in the future. Maybe I will be in this project team in the next two days." This is more convenient. Let's do it first. It's okay to get out of work, ". Without a sound development platform and system development specifications, there is no consistency between developers. You can't tell a developer: To write it like this, it will be convenient for future maintenance, and it will be advantageous to the whole project or other developers. In other words, without legislation, you say to the paper mill that it is completely useless to take good care of the environment.

Again, let's assume that we are masturbate, that is, we are masturbate, that is, we have high quality, and that the code we write is, comply with the xxx structure, clear thinking, easy to modify, easy to maintain and drop the code. By the way, there is no error code. Why? Yes, because of its super lust, we can write perfect class without the need for unit tests. (It is assumed that the sex is a super sex, sweat ......)

Of course, it is also a kind of obscenity. It is also wrong to test the combination. Therefore, if you have no mental retardation, you will do a functional test or system test to prevent god from being violent. What should I do when there is a mistake? Of course, it should be corrected by the super molt. It is of course the excellent quality of the Super molt. Of course, the reason why super masturbation is super is that after the code is corrected, there will be no errors, and there will be no mistakes in other places. Why? What does regression testing do? Is it hard for you to believe in the power of super sex? If it's true, there's no error!

With so many assumptions, I have been sweating, and I have made some effort to write it down. I will continue to talk about it in another day ~~~