No head, no tail - Project Development Notes: UML, Idef Failure Application in Our Project

zhaozj2021-02-16  65

Title: Did not end without tail - Project development notes: UML, IDEF failed in our project

Keywords: distributed development C # project division updata UML IDEF

November 21: In the afternoon, some feelings are not spit

First declare, I was the first time to do this for the overall design of this transfer project. More than a few UMLs have been used, and I used a little IDEF. However, it is not fine.

I was trying to draw the needs of users with user case, and then I am very interested in discussing with comrades (I remember that I first defined a few big modules first, and then submitted to the foundation with a document. Each made a multi-case figure. Correspondingly draws a few sequential diagrams). Can you want to respond to it? Ha, I told you that the reaction of comrades is: "Oh .........., you should go to the customer, what do you say?"; "This is too nasty!"; "I think I am still I want to see a big tree, I can know how much things we have to do! " I made my dizzy steering.

Then I pictured directly on the whiteboard to draw the overall IDEF map, and I have a big joy and say: "This picture is good, this picture will explain the problem!". depressed!

I took another three days to turn the UML into painting IDEF0, gave comrades, and then everyone looked, talking, then put it in SourceSafe no one.

So the project is still doing down on demand, do it down ..........

From the above, from the beginning of the stage I think the project actually has a lot of problems, for example, because there may be no one may understand this in the organization of the project. (Mainly I don't understand business, also known a semi-solving for UML and IDEF). Then try to use this stuff in improper occasion, (For example, I try to take these pictures to give the customer, I feel that the customer's feelings are, this kid is talking about, this is not all I told him, he Come tell me? I have a disease!) Developers don't catch a cold for this, don't draw, I can do it, your picture I have to jump to jump, I have to do it on your picture. The painting is not full, and it is useless to play the virtual.) Wasted a lot of time.

There is no time to feel the tools for people, we use it well, and the project is still going down. Some opportunities use another development method. Next time I will describe this development method.

Here I would like to talk about my feelings of developing this matter.

l Selection of design tools used to use:

The popular statement is to use UML to be used when developing. The traditional statement is to use IDEF, but if you don't have a matter, don't use these stuff. Can understand the demand, draw a few pictures that everyone know, the form is really not important. (Especially for the company, our company's demand business is very high, the customer wants him to know, but you look at the case map to this, they still say, what is this? Dog * ...), if your documentation wants to archive, it is necessary to communicate with other people, other companies, so it may be necessary to use a kind of Dongdong to make a clear thinking. If this is not the case, the boss asks your progress after the ass every day, the programmer tells you that he will see your picture to completely understand! ! Really let me want to marry.

l Development purposes

What is development? This problem must have his own answer. I want to hear how I used to say to my boss. "Development of a product is to make customers satisfied, the minimum needs to meet the needs of customers, etc., it is best to meet the expectations of our customers! If possible, you can put development experience in the company, and you don't put development experience in each experience. "On" programmers ". When I was still in question, my friend told me: "Little Comrade, I tell you, if you want to make a project, you will lose money !!!! ! "(Giant cold). I can't accept this sentence at the beginning, but I'm quietly thinking about it. In fact, if this sentence should add a premise, this is the truth. "If you want to develop a new type of item, if you have developed In the future, the company will not develop its own business in this way, then you seriously do a project, you will lose money. " This is the reality. If you don't have a way to use the relationship, you can only use your company's technical strength to pick this type of project, your previous investment is certain will pay for money. I don't know how long I have used this half tonar UML, I learned how long IdeF used, but if this type of company, this type of project, my feelings are not as good as these No need to. Don't be given by the beautiful illusion of these stoves. The purpose of project development is to make a customer's own Dongdong at the specified time, and other first side. (Rely, when a project technical manager is too irresponsible?!) L

There is also to talk about my understanding of the development process. If you read the lasting note, you should know. I am two types of development, one is two kinds of user needs, one is an object-oriented database design. Then I do what I have to do with the two stuffs. If these two are doing a pair of contradictions in implementation. Why do this contradiction? What can we solve these contradictions if you use the CASE tool? Can't, from my practice, it can't.

I have seen a few eyes. But now I can't remember. However, I remember that it is about the development of development into demand, design, development, testing, implementing such a procedure, then doing iteration development, making such a process repeated, in order to meet customer needs. When I have gradually put this process as a truth, one day I ask yourself, why do you have to pass these steps? If this process is perfect, then why should it be iterated? Is it because the customer is always changing? Still because you can't guarantee the need to understand the needs, or because this process is problematic. In the middle, I thought about some time, I won't go to Luo. Anyway is likely to be right. However, I want to share my conclusion with you, if you have any opinions, you can discuss it.

1. The assumption that this process is that people doing this do not understand the business at the beginning (at least the default value of this process is developing business.)

2. At the beginning of this process design, it should be based on our very habitual relationship database to develop software. The relationship between object-oriented relationships and relational databases is mixed. If the business is simple, you don't have to have a relationship database? (This is a hypothesis)

3. Do this type of project is always very unconstrained, from demand to see, users want to realize business, then design people began to refine, find objects, dig empty intent from objects to write databases, this Parallel. After writing the database, the design is finished and start writing the program. The process of writing programs is one person, and there is a number of individuals who write the middle support. Everyone who is engaged in this will be more confusing. Each development phase has different tasks, the efficiency is not high (especially the small company 555555) 4. These steps are actually very obvious, that is, based on this In the development of the process, we are most grasped from the process of designing to implement, the most dangerous needs, the needs of users. This is also why we have iterated (not technical issues, business problems.) The problem that the development process can solve is just to store your data, organize. Not to do your business. This process is unhappy.

5. I always don't feel self-cultivation as a role that is not a programmer. When I realize that if we just submit to a purely-oriented system (including the interface, there is only one object on the interface called XXX system, After you go, I find that there is an object called documents, then I will find that there is an object called purchase order, then tell you a bunch of methods, one is called new. Let you modify a bunch of properties ..........), Users will say that you are idiot. I don't think back to see why we want to use object-oriented technology. why? It can be given to an answer. From the 1970s, we are forced to go to the object-oriented road. Meet the programs to write a wonderful programs. When you meet Sales to sell products, you need to say what good new technologies apply in our products, and to ignore the meaning of true software.

Grouped also sent a pass, caught up, can't solve the problem, things still have to do it. What I will write to us in the next section?

转载请注明原文地址:https://www.9cbs.com/read-24206.html

New Post(0)