This is the source of the article, I hope to go to the original site, there are more exciting articles!
Here you are for your reference only! http://champion.ewuxi.com/old/exp/foeInProject.htm
Software project enemy
Software projects have been done so many years, there is also a failure, but there is no very satisfaction with a project. Maybe I am a perfectionist, always thinking that there is a matter of the project to be full of satisfaction, always feel that it is an end, the customer can applaud, this is what we want, the boss said , Very good, we earn, colleagues said, good, we will be easy. Maybe this may be a dream, but I often think about it in the middle of the night, how can I do better. So I came up with a lot of wrong, I thought, I would like to do something wrong, I will try my best, maybe there is no right, but there will be better ways than these mistakes. I called these enemies, because I want to pay attention to them, destroy them.
First, Tianzhen Plan. Perhaps the plan is that it may be inappropriate, but it is really impressing this situation. A few days ago, I saw the report of Hongqi Canal. From now on, this is a great project to conquer the Taihang Mountain. However, this actually used the project that did not complete it. It was scheduled for 80 days when the initial time, and most people were watching reports, I am afraid that the first reflection is the same as this plan (afterwards, who will do it) . In software projects, or actually differ from programs, there is indeed a lot of such projects, which can be said that from the beginning, these projects are heavy. why? I think the reason is that the inequality of the tripartite, the software requires developers, customers and bosses (or those who are equivalent to this level, in short, in China, developers have almost very few power) to complete. However, the development of the plan is often a customer with the boss negotiation to determine that developers are no power to ask (even if they can ask. Symbolic). So, the plan is unregistered by the plan (customer) and the semi-outer line (boss), mostly is unrealistic. And the software industry does not have any standard plan or standard labor to compare, it is inevitable if it can be said. I often see, make a bridge, saying how much we have succeeded, how good everyone is doing, there are always data, how long is the international standard to complete such a bridge, and we have spent more time to complete. Also, for example, Shenzhen speed is a few days. This is clear, but so far, it is difficult to see that a programmer has completed the number of standards a day. So the customer and the boss considered the starting point of the plan to almost ignore this element, and it is simply considered from cost efficiency, customer's psychology. So we sometimes see that in the process of some orders, some software companies can report 500,000, and the other may report 50,000. If such a project has been sold at 50,000, it is a disaster to developers. The reason is very simple. If the developer is centered on the boss, under such a low cost, the boss must compress human cost. As a result, the customer is not satisfied. If the customer is centered, it will inevitably bring costs overrun, boss Dissemination is also inevitable, and in both cases, the pressure of developers is large, so developers will not be satisfied. The ideal three-win goals are smeared from the beginning, such an enemy is unable to cope with the developer.
Second, the trap of technology. Sometimes we can see the optimistic project staff say this, "We found a new tool, or we will use new technologies in the project (Java or .Net is currently using the most use), use this tool, we Will save 1/3 time, although we will use some of these times to learn how to use these technologies, but they will be effective. " Say these words, if you don't overset your own ability, it is the ability to overestimate this new technology. Finally, when the project is going to a certain stage, we have not retired when we find the ideal and reality, because the time has been flown. I used a tool in a project. When I started, it was such an attractive, it seems to have the maintenance coding of the database for the previous day for half an hour, because it will generate it for you. Most of the code. Indeed, I took half an hour to generate classes, and the developer is very simple (at least I think so), then I spent a day, church Other people using . Then, the problem slowly starts, first, the developer keeps complaining about the trigger mechanism between the incident between its internal classes, I tell them that they don't have to care, as long as the interface is concerned. But the trouble is indeed, the developer cannot face the database code directly as in the past, but to pass some classes, this makes them suspected whether the procedures written can really work. Next, starting to find these generated classes can only be modified for data for the primary key, and sometimes we need to modify a batch of data through the primary key, so we have to add some code to implement this function, at the same time, It can't handle binary fields, so we have modified some code. On a Monday morning, according to the customer's opinion, we modified some database structure, and then I used this tool to update the existing classes to keep the same as the database. At this time, the disaster happened, I found that it produced some bugs when updated, first, it has bug in the field type of the primary key, although it can correctly modify the attribute type in the class, but the error modified UPDATE The parameter type, more terrible is that it has completely covered the original file, and the modifications we have in these classes are not retained. So I have to find a backup file and compare these files one by one, and then update. This part of work, spending me two days, while the work of other people has been delayed. Since then, there is a matter of making such modifications. Wait until the project is completed, go back to compare, the time it costs is much more than what it can save, and most developers are done to the correctness of these generated code until the project is completed. Therefore, if it is an important project, developers must use new tools, it must also be added to the risk bar. In fact, my first Java's web application is also spent for a week because of Chinese issues, and this time, if I am familiar with ASP or PHP, it can save. It is dangerous to maintain an optimistic attitude toward unknown things. In fact, there is nothing tools, which can really improve work efficiency. If there is something, maybe we have been unemployed, :).
The third, research and development methods. It is important that the biggest feature of the project is that the project requires that the project requires a certain purpose within a limited time. As a developer, it is often hoping for technology in subconscious, research technology may become their greatest interest, not other. There is an example, saying that some people have an investigation in a programmer's party, asking how many people have been removed from the alarm clock before the age of 9, and more than 90% of people have gone, when I asked how many people Install, and when you work properly, only less than 10% will raise your hands. Some visible, the process of careful research is not a result. It is terrible to be in most developing groups, the more master, the easier it is to produce such problems. The definition of masters is often something else can't do. He can do it, or other people do one week, he can do it in two days. When such a person has a small difference in the project because of the sudden increase in the details of a technical production , how is the loss? The research and reserves of technology are important, but in the development of the project, such things is best to avoid. Microsoft advocates "Good Enough", if not the case, maybe Windows 95 is still there yet. In China, there is also a corresponding sentence "More perfect system, if there is no user to use, it is waste", but unfortunately, sometimes people will forget it. Its four, burying his head. Some project teams, the end of the research, all personnel are in the company's compulsory program until a stage, or the customer is reminded, and communicate with the customer. Perhaps your system is upgraded from the internal version 1 to the internal version 3 during this period of time, the speed is improved, the picture is beautiful, and the operation is convenient. But customers don't know anything. Therefore, it is important to communicate with customers, and you must communicate every week. Sometimes customers don't care about what you are doing, what is important is that the project has been in progress. During this communication process, we must cultivate customers to understand software development. If customers understand software development, then, when they want to fight for time, or when they change demand, they can more consistent ideas. For example, in a project, the customer understands that as long as there is appropriate data, the report is relatively easy. Then the customer will be more comprehensive when considering the composition of the data, and it can understand the development of the developer's time. Relatively, in this process, developers will have a better understanding of the business, and they can provide constructive opinions, bringing many savings to customers and development work.
Its five, improper pressure. The developers are introverted, not good at argument, but this does not show that they have no idea. If it is just a simple increase in work pressure, it is required to hurry, it is required to speed up the speed. The final result is to reverse, a group of lost vibrant developers are like a defeated soldier. Assign appropriate tasks and completed in a clear time, so that developers will feel that everything is in the mastery, and the effective rest makes them working longer.
In the process of the project, there are all kinds of factors lead to failure of projects, need someone to control, is there a risk control personnel in your project group? If not, find one, this does not cost a lot of time, but it is really effective, he will help you find the enemy together and eliminate it.
Zilong, August 2002
Blue Sky Copyright
