On July 6, 1998, the indigon of Hong Kong's new airport was officially opened. This cost of $ 20 billion is finally angry. However, when the new airport opened operation, an embarrassing problem appeared: the electronic signal showed the wrong information, the telephone system was forced to close, the baggage did not see, the passenger goods were stranded.
The 20 billion US dollars of complex architectural structure is indeed one of the world's largest building projects. In the project, half-world excavator teams are busy involving this vast project. Dozens of contractors and construction projects are waiting to coordinate to complete thousands of different and often associated complex activities that are different from each other at the same time. Although the project is facing the desperate situation from time to time, the project is finally coming out of the desperate situation.
However, the dilemma described in this article is still produced, and this problem has lost $ 600 million. But this problem doesn't matter to the main building project. They all come from computer hardware and software.
In fact, the issue of software development encountered by Hong Kong New Airport is one of the failure cases of software industry. The success of the software industry is just an exception.
Start meant means failure
The Standish Group surveyed for approximately 8,000 software projects in the United States. The results show that 84% surprising in all software projects have not been completed on time, complete or install all previously required features, and more than 30 in all projects; % Was canceled before completion, and others significantly exceeded the deadline and (average) exceeded 189% of the budget.
Another study found that more than 50% of the highly complex project were canceled. In other words, such a project will be higher than the chance of success from the beginning.
Software is always fascinating. In the early 1980s of the last century, the US Bank experienced a nightmare that failed MASTERNET system. In order to grasp the opportunities brought by the information technology, the Mei Lijian Bank became the giant of the trust industry. In the fall of 1982, the United States Bank decided to establish a software system for business and customer management. After 18 months of detailed research analysis, a budget was $ 20 million, and the project plan's deadline was December 31, 1984. However, a New Year's Eve came again, and the system did not appear. Until the year of 1986, after a year and a half of the initial period, banks were able to demonstrate a system with too many vulnerabilities.
In March 1987, the system was officially working in late 27 months. At this time, the IT nightmare of the United States of America began. Banks can release accounting reports in 3 months, so the customer's trust began, and corporate customers took away $ 4 billion in funds. Finally, the management of the United States of America gave up this project.
This kind of disaster may happen quietly, but the disaster of Denver Airport, "Aria" rocket explosion, has been the explosion of the Boeing "Delta" rocket, all the software gives people a breeze ring.
Who is the culprit?
Why is the software industry so easy to fail? Who is the culprit?
...........
Attempt to shorten the project schedule by increasing people. Some developers' mistakes think that it is considered that the project man can shorten the project schedule. This is usually wrong, there are two reasons: First, development work is often unable to split between many workers, mutual dependence enables certain task requirements to be in order. As Brooks said: "The child is always taken for 9 months, regardless of how many women are arranged." Second, the time of manpower increases, the time spent on communication will grow synchronously.
....................
"A talented software developer creativity can be ten times that of ordinary people, even if they have the same pass experience." McConnell said quite emotionally. McKinsey Company has agreed to this in the investigation. "A talented carpenter will do a lot of decisions when building a good project. He made the right thing through intuition." Barizza, a senior development manager said, "It is also the same for a developer. Many decisions about structure and coding are quite intuitive, depending on long-term experience and talents. It is difficult to bring together them, but the results will have a huge difference. "Software development project should be equipped with the best people. "It is best to find a programmer with creativity, not to expect a programmer to become creative." McConnell said. The same principle is also applicable to the team.
Create a powerful team structure. Not every programmer is a star player. "When tissue grows, you will have some mediocre developers - this is almost inevitable," said Platinum CEO Philipski, "The top masters can be sucked."
But successful companies have learned how to combine stars with those who are not equally excellent. Like a hospital first-aid team, there are one or two top programmers, others assist them.
...............
Take the top talent. When a few years ago, a top programmer in Platinum said he would have to leave the company, because his wife got a good job in California, the company's COO Hutan Yusky completely refused. Because I don't want to lose this kind of employee, Huaman Yusky arrangements will build terminals in this developer in California. - And his huts on the mountains of Nevada. Since then, this happy employee (obviously not an early man) at noon, game or hike, work for Platinum, from 3 pm to 3am. "If we don't have a way to establish a way, overcome geographic boundaries, I may have been out today." Hutan Yusi said.
....................
Synthetic components?
For decades, the development manager has been calling new "magical solutions" to bring a fundamental or real revolutionary progress in software development. Today, the synthetic components are significantly changing software development games.
Many software experts agree with the change of software components. Cox once observed that if the software development energy from "Building all" turned to "repeatable components that simply integrated customization, this means huge productivity and quality improvement.
But using component development is also facing all kinds of difficulties. For example, a large number of pre-investment needs to change the entire ideological form of developers.
