[Installation] excessive technological hype

On stackexchange.com, a post commented on the over-hyped technologies in the last 20 years. Most of the results agree, and some disagree. I have selected 10 from the previous 15 (I removed the Java wore and TDD, because I think they should not be over-hyped, and they are both good ), listed in the order of the original post: (I added some comments later. You are welcome to discuss them)

　Top 10 Technologies and concepts over-hyped

• Unified Modeling Language (UML)

UML is a good tool for programmers to exchange ideas, but it is far from the design tools that programmers really need, such as: whether the design meets requirements, architecture design, data flow, and so on. Only a few programmers use this tool to exchange ideas, instead of using it in their specific work.

• SharePoint

Now more than N companies are using Microsoft's internal Intranet. However, installation and maintenance are costly. However, the market has been very successful. It is not technically speaking, it is rather bad for technicians. The Design of SharePoint has not carefully analyzed the business process. It is just a place where documents are stored. It seems that we can do anything, but if you want to use it to manage your project and track your project problems, you will find it extremely difficult to use.

• Extensible mark-up language (XML)

XML. We have mentioned a lot before (xml1, xml2). We use it to encapsulate program data for configuration files and network transmission formats. Our program processes XML, which is slow and Economic. Without tools, we can hardly maintain XML files. It is really not economical to use XML for Data encapsulation. Isn't yaml and JSON simpler than it? I don't know who came up with the idea of using XML for program configuration files. It's silly to look at the configuration files in Unix/Linux, which is easy to read and maintain. It's really high technology.

• Web Services for soap, XML-RPC, and WSDL

This thing was fierce a few years ago. Everyone believes that this is the future of programmers. Unfortunately, the complexity and inconsistency are quite disgusting. The soap S is still simple! It seems that nothing about XML is good. However, I personally think that CORBA is more evil than him.

• CORBA

As a complementary technology in Linux/Unix, which is more hyped than its more vicious COM technology, this technology is nothing better. It's quite complicated. It's like this in theory. This is something that has never been put into practice. Then start hyping.

• Cloud computing

This is something that relies on hype. That is to say, we can use different calls, such as computers, tablets, mobile phones, and mobile devices to do what we want to do anytime and anywhere. Google's chrome notebook ads demonstrate this technology, but how many people put their work results on the cloud. More people prefer to use computers or platforms that they can control. Google is obviously inferior to Amazon in this regard. like Amazon EC2, you can start your remote system anytime, anywhere in the world.

• SOA-Service Oriented Architecture

This is a concept that no one really knows what it is. After many years of hyping, many people try to understand it, but the final result is a yawning and reading other things. No one has mentioned it now. Some banks in China have engaged in many so-called SOA applications under the encouragement of IBM. The result is that the system is very complicated. Of course, IBM is also indispensable.

• Software Industrial Process

There are many so-called industrial processes in software development. Using these processes seems to be able to control quality. Outsourcing companies and local Chinese companies like these things, such as ISO and cmme. These processes cannot be difficult to say or have good news, especially for workers who don't think about what they just want to follow. The hype in these industrial processes is that the so-called use of these processes can predict the project cycle, quality control, development and management of previous needs, and so on. It brings the process to the predictable level of theology, as well as to the political level. Because SQA's audit process and statistics and report process are inevitable in these processes. These are not software development processes, but they are indeed quite political. Companies that use these industrial standard processes are usually creative and problematic.

• Agile Software Development-agile development

First of all, I admit that many of these experiments are quite effective and theoretically sound, and there are many good methods. However, there are still hype components (the following comments, I am waiting to be scolded). For me, in China, the hype of "Agile development" is like a TV shopping, thoughtworks consultants in China are not experienced in software development. To be precise, they have consulting experience, without specific project implementation experience (some consultants did not even write a line of code to learn how to program and develop software), they can communicate with them to feel that they are excited about agility, in addition, it is only agile, and the slogan of one process and one agile is worse. They believe that agile processes are close to religious beliefs, and their spiritual world is very North Korea. Because no matter what you talk to their consultants, they only say agile and never analyze what are the characteristics of the project? What is the style of the person who develops this project? What are the characteristics of customers? Is there any stakeholder who cares about the software (such as programmers, testers, customers, and managers? XP and scrum have become the most powerful tools for push engineers.

　　The process should be self-developed by the project team, rather than being instilled and used by the dogma. Different teams, projects, and people have different styles. The best process is to use the process that suits you. For example, in the football team, Brazil is playing personal art football, Germany is playing overall and disciplined football, and Italy is playing defensive football, however, they all have the power to win the World Cup championship. If you want to let the Brazilian team go to the German or Italian style, then it will be a tragedy.

Obviously, thoughtworks is very similar to turning all Chinese software companies into agile, Which is doomed to be a cup of cake in China, and can only win the companies and projects that are overwhelmed, there is no suitable project, and it only relies on various hypes (such as some conferences and Publicity ). They always think that Chinese users and programmers need to spend time constantly educating, but they never think about their own reasons-Education and indoctrination will never win.

　　My personal advice to them is not to think that the world is just as you think. To learn to respect programmers and projects, there are still many non-technical things. Listen to what programmers and customers say, let's analyze the characteristics of the project from the actual situation, instead of injecting Tao's theory into everyone.

• Object-Oriented Programming (OOP)

I will not talk about it much. I have said this before. All things are in the article that object-oriented is a scam. But one thing I want to tell you is that the object-oriented design pattern is being abused. Design Pattern teaches you two things: 1) how to simplify the complexity; 2) how can we reduce the coupling of objects. Instead of a formula, more programmers have learned "popular design-Mode Programming ".

Appendix: Technologies I do not think are over-hyped

• Write Once run anywhere

This is a bit confusing for me. I don't know why it is so advanced. This is the slogan of Java. I think Java is still successful in cross-platform development and there is no excessive hype. Virtual machines do. For companies that require different hardware and operating system platforms and constantly upgrade and replace them, this is indeed a good solution to platform dependencies. I feel that this technology is not overhyped, at least on the Java side. This is not so much about EJB as it is about over-hyped technology.

• Test driven design (TDD)

Writing a program from the test case may be a method that many programmers are not used. In fact, this is a good programming method, which ensures that no code changes will break other code that has not been changed. The code can ensure quality in a continuous integration. However, we need to know some side effects of TDD (the disadvantages of TDD are also mentioned in ten good programming ideas): 1) TDD may make programmers perfunctory, the test case is correct. 2) TDD may cause you to ignore the software design and architecture as well as the scalability and reusability of the program.TDD is just a method, not the core of the program. Of course, the speculation of TDD in recent years is also a little overhead. I am questioning the argument that "TDD is a design method" and other "amazing technologies" have emerged.