Liu Tianbei
"Last Term", Tom Di, Bear Festival, 2003, Tsinghua University Press, 306PP
... k I only feel that I lost the road, or entered a singular country, this country is far from any place that humans have arrived, it is so singular, even the air does not have a little ingredient with the same air. Here, a person may suffocate because of the strange, but this strange has an unprotiable charm, so that you can only go forward, and you will be so deep. - Kafka "Castle"
Hijacked project
Why is a novel that lectured project management begins with the owner being kidnapped? Mr. Tukkins, this experience of the senior project manager of love, the company's recent victims of layoffs, is kidnapped to a completely strange country, and the committee begins with a nearly impossible task, starting a strange, difficult to measure adventure ...... This sounds completely like a Kafka-style absurd novel, but it is actually the beginning of IT famous "deadline". You will say that this is just a hanging, attracting the reader's routine. However, "suspense", originally only happened to those areas where there are risks, need to be biliary and luck: spy or anti-spy, high-altitude rescue, buy lottery, etc., where there is a risk, there is a failure or even Musk. Why will the software development project will be involved in a suspense, one in an adventure? "Last period", English is originally "deadline", directly translated is "Dead", it is said that it is that the last boundary in prison, the prisoner once more, it is more, the author is to hang it with this symbolic developer The sword? Is the authors in implicit, software projects are likely to struggle on such life and death limits, which is likely to be in danger of "being hijacked"? From the perspective of the software industry, "Project" often means the standard operation, even the synonymous language of "success." Please imagine a building project. Do not consider the arrears and cost recovery, simplicity of "failure" from design, construction perspective, is quite small. If you want to make an external trip to speculate, the software project is rarely dealing with tangible "material materials", and the probability of success should be higher than the construction industry. However, any experienced software developers will understand the proportion of "risk" that I said for software projects. Let us collect "hard data" recognized by the industry: as authority of software development management, software project research experts, this book author Dima can talk about all of his famous "person", "all projects" Above, about 15% of the project completely failed. They are either canceled, or abort, or extension, or submitted products have never been put into use. The probability of success for large projects is lower. Continuous 25 In a working year or longer project, there are 25% of the projects that have not been completed. "In fact, the title of" Chapter "Chapter 1 is" At this moment, there is a project in some project in the world "is failed." . There is no texture guide to the "Software Project Survival Manual" - suggesting that the project manager needs the Royal Air Force Pilot's adversity to survive, and the other is simply called "Death Tour". It means that if a project The manager hits into this unknown territory like the exciting explorers, so waiting for his fate without knowing. Then, maybe it can be said that software projects are in essence, first and always in danger. In the face of such a high risk, a lot of project planners who deeply think will even make multiple project groups simultaneously develop the same module while taking the optimal results (as far as I know, this practice " It is quite common in the Japanese software industry). However, there are still a lot of software projects that are not lucky, or there is not enough consciousness, or you can't bear a lot of manpower, only the former servant is hijacked, like Kfka novels, K or Sites As the captain, "continues to go forward in the way" "continues to go forward, the more beautiful". Outlook for a range of voyages
I immediately realized that I had excessive pessimistic tone. Although each adventure is definitely "death trip", it is obvious that every expedition is not returned. Take the famous loser Sutt as an example, with the Norwegian Admison Adventures of the Almost Simory, the Norwegian Admunsen Extension team has been successful. Back to the software industry, according to the above number, 25% of the failure rate can not be tolerated, but after all, most software projects are really going to transport. So, what is the secret and failure of success? If we have successfully successfully, the truth of the failed software project has been summed up, and can not be held in the future, smooth sailing? In the pure technical field, there is indeed a lot of work that is committed to working. Many experts have found that everyone is always repeating the same mistake, thereby summing up some typical error ideas in software design, and call them "reverse mode." For some specific development areas, such as Java, we have seen the monk appeared in the "Java Pitfalls", "Bitter Java", "Bitter EJB". From the book name to read the danger of traps, lessons. In the software project management, if there is such a book to record successful routes and the book of the shipwreck, we will not be able to grasp the shipping direction, avoid the notorious reef? I originally started reading "deadline" with this idea. This book can also actually play this role. With our friend Totopkins in the fictional "Morrovian" adventure, I also learned a lot of DOS & DONOTS - every chapter from a wit and wonderful story, there is a "soup" The presence of the form of the Pupins Diary ", if the time is nervous, browse these diaments in order, you can harvest a lot of real knowledge on the work division, staff equipped, project time plan, test, release, etc. But is this enough? If you can shape a project manager if you have a number of principles of familiarity, what is the most scarce resource in the world? In fact, after reading the whole book, I feel that my biggest harvest is not any specific management secret, and it is just aware of the awareness: there is no single practice or principle to ensure a successful, same, and any single defect. The project orientation will not be failed. Just like Topkins' success is not completely relying on my own experience, or, with the guidance of which all-around wise, it should be attributed to the comprehensive role of various factors and the proposal of many geniuses, if we are just a new method, For example, "test first" or "continuous integration", or even if we fully replicate all the prompts, the probability of success in the next project is not necessarily improved. Similarly, the story in the book also shows that even if you have a super decision maker on "Evil Belos,", your project is not necessarily single, and it is over. This seems to be a simple expansion of the "no silver" theorem proposed by Brooks. However, I personally prefer this as the "multi-decision theory of the software industry". Multiple decisions means that a variety of factors under specific scenarios are in a complex, dynamic and mutual interlaced relationship, emphasizing which one of the priorities can only be simplified and even distorted. As I saw in my current identification ability, I think the biggest hindrance on software development routes is the interaction of the three factors of "business, technology and management": software development first and ultimately business activities, commercial interests require development cycles, better and better The less labor costs, the better, the more demand for the later tolerance; the technology is the importance of software companies, especially how to deal with the inherent conservative and software sustained technology revolution Conflicts, becoming a problem that every item will encounter; and how the software project managers have more coordinated technology and non-technical factors, how to seem to be oligual products, in fact, the development process of unknown factors, is also monitored. The boat of the software project is the turbulence and swirls formed by sailing in these three factors.
When we stay at any single aspect, a narrow devil will quietly sneak from the other two, which leads to consequences. Overcoming these obstacles, we need patience, respect for all problems, and even a little luck. If you want to use a "project management software", you can master some skills or a "realm of" realm ", one for all the problems once, currently in my opinion: we are facing difficult, in Brooks In the sense, "essentity" is not "accidental". Even if there is a particular project to solve a difficulty, we cannot guarantee that we have immunized it. But on the other side of the coin, as the famous sentence of the German is, "Where is dangerous, there is a rescue." This intrinsic complexity of the software project may be the same as its "strange charm". If the art developed by software can be "linearly" through the principle of abstract, then we can even ask, will there be a software project only by the computer, and human developers are completely replaced? In strict scientific sense solve this problem, it may require a more clear "developmentable" definition (just like the clarification of "calculated" in the last century). A detailed investigation is completed to all of the people who are interested in the Tuling Award. However, in intuitively, according to the discussion of the above, we can believe that software development is in the process of being overcome by the machine, and the human developer is the most good. I think this is the argument that people who are really happy to be in this business: I would like to thank these difficulties. The majority of software developers will not wake up on a certain morning to find that they have been replaced by a dry computer. How do people talk to software engineering?
