Read the "deadline" experience
This is a very interesting and deep science work, the book's genrealy's physicist George George George, the University of Colorado, the University of Colorado, Mr. Gamf, Mr. Gamf, Mr. Middle-aged bank staff - a series of small stories, I am very lucky to read the book and have been attracted by interesting physical stories, so when I see the "deadline" of the same style, the book is deeply attracted. I have. Author Tom Demarco's original nature of Moamov is also very known.
The core of this book is the summary of the owner's soup in his own diary, and records some of your experiences for this diary.
1, four major elements of quality management
Select the right person
Assign the correct job for them
Keep their enthusiasm
Help team condense and keep the team's cohesiveness
(Other everything is just a "copy")
Experience:
2, safety and change
Unless it is safe, people can't meet changes.
In all successful works (and in vast majority of other valuable work), changes are one of the basic elements.
The lack of security will make people against changes.
Escape risk is fatal, because this will make you not get the same interest with the risk.
People may have no sense of security from the objective world, but if the managers may abuse the power to punish themselves, they will not feel safe.
Experience:
3, negative effect
Threat is not the best way to improve performance
If the assigned time is not enough, no matter how scary, work will not be completed on time.
Worse, if the target is not implemented, you must honor your threat.
Experience:
4, the body part of the manager
Management involves heart, gastrointestinal, soul and nose
Leader
Believe in your stomach (believe your premonition)
Framed team's soul
Training a nose that sniffs the lie
Experience:
5, use the command war to act as a metaphor
At the beginning of the war, the real work of the manager has been completed.
Experience:
6, interview and recruitment
Recruitment involves all management-related body parts: heart, soul, nose and stomach (but mainly gastrointestinal)
Don't try to recruit alone - two parenteral stomach is better than twice the stomach.
For new employees, let them assume the same difficulty in the past, and postpone challenging targets to the next step.
Distinguisice: The person you most want to hire may also know other good people.
Listen more, less.
If you first consolidate the material, then all things will be better.
Experience:
7. Improvement of productivity
There is no "short-term productivity improvement".
Productivity is raised from long-term investment.
Any commitment to immediately see the effect is likely to be the million gold oil sold by the rivers and lakes.
Experience:
8, risk control
Manage projects by controlling risks
Create and maintain risk statistics for each project
Track the risks of rootic, not just the end of the nasty.
Evaluate the probability of each risk and possible overhead.
For each risk, predict the early sign of the avatar.
Appoint a risk control company, this person should not maintain the attitude of "I can do" inside the organization.
Establish a simple (probably anonymous) channel to make bad messages to the high level.
Experience:
9, prevent failure
Strong man
Controlling failed to improve your comprehensive results than optimization.
There is a hard work, and cancel the failure of failure as soon as possible.
Unless necessary, don't condense a team: go out to find a team that has been formed.
Maintaining a good team (as long as they are willing) to help your successor avoiding the team to condense slowly or cannot condense.
The team that condenses together - is ready, and is willing to accept new work - as one of the gains of the project.
Experience:
10. The damage to the project at the beginning of the project is wasted, the damage caused by the project is the same. There are countless ways to waste a day ... but there is no way to take back the day.
Experience:
11. Modeling and simulation of development processes
Model your process of completing the work
Use these models in the exchange of colleagues to communicate, refine thinking about project operation.
Use the model to simulate the results of the project.
The model is adjusted according to the actual results.
Experience:
12, measure
Metrical size of each product
Don't persist in units - when you wait for objective metrics, use your own subjective unit.
From all possible raw data (calculated software features) to construct metrics.
Collect raw data from the completed project to derive the birth productivity.
Continuously improve your metrics until its calculation results have the best correspondence between project works in the original database.
With a trend line with a database painting, the expected workload is displayed as a function of the humanometric unit value.
Now, for each item to be evaluated, the artificial measurement unit value is calculated and the expected working value is found on the trend line based on this value.
Use the interference level around the productivity trend as the tolerance indication of the mapping.
Experience:
13, process and process improvement
Good process and sustained process improvement is a great goal
They are also very natural goals: excellent technical workers will pay attention to them, whether you tell them.
The formal process improvement procedure needs money, spend time; specific process improvements will also delay the progress of the project. Although it will eventually reflect productive gains, they cannot offset the time of flowering improvement.
However, the project has a sufficient income from a single, correct choice method, and win back to the time and money to change this change.
During the project, do not want to implement improvements within the range of more than one method. A variety of technologies (such as an improvement of a CMM level) is likely to allow projects that are longer than those of these programs.
The risk of standard processes is that people may lose an opportunity to take shortcuts.
Especially for the project, the standard process looks very rigorous because they have created enough work (useful and useless), so that everyone is busy.
Experience:
14, change the way to do work
If you don't reduce the time of debugging, there is no way to make the project greatly completed.
High-speed completed projects are much less than the time on debugging.
High-speed completed projects is much more proportional to the design time.
Experience:
15. If you don't care about others, don't take care of others, don't want them to do some unusual things for you. If you want them to change, you must understand and appreciate their past.
Experience:
16, the effect of pressure
People under pressure cannot thinkually think more quickly.
Increase overtime time will only reduce productivity.
Short-term stress or overtime may be useful strategies because they can focus their employees and make them the importance of work. But long-term pressure is definitely wrong.
The reason why the manager will apply so much pressure, perhaps because they don't until what to do, or because of the difficulties of other methods.
The worst guess: the real reason for using pressure and overtime is to make everyone look better for the failure of the project.
Experience:
17, anger manager
The angry and humiliation in management are infected. If senior managers like to swear, low-level managers will also be sampled (just like a child who is often embarrassing is easy to become a parent of love).
If the manager uses the abuse method to stimulate the employee, this shows the incompetence of the manager, not the incompetence of employees.
Experience:
18, ambiguous specification document
The vague marks in the specification documentation have unresolved conflicts between different system participants.
If a specifications do not contain a full input output list, it is no hope: it has not begun to indicate anything.
No one will tell you that a specification document is not bad. People tend to blame themselves, not blame documents. Experience:
19, conflict
As long as there are multiple participants during the development process, there must be conflict.
Conflicts are particularly prone to the business of creation and installation systems.
Most system development teams lack the ability to resolve conflicts.
Conflicts should be paid attention to. Conflicts are not a lack of professional ethics.
It should be declared in advance: "Win" of everyone is paying attention. Make sure everyone at each level will win.
Difficulties in negotiations; easy to mediate.
If the interests of the two people are completely or partially rejected, pre-arranged, ready to ask the two parties to resolve conflicts through mediation.
Remember: We are standing on the same side; opposition to us is what we have to solve.
Experience:
20, the role of the catalyst
There is such a catalyst. Such people will help teams form and condense, maintain team health and productivity, resulting in contributing to projects. Even if the catalyst does not do anything (in fact, they will often do a lot less), the role of this catalyst is also important and valuable.
Mediation is a "catalyst". Mediation can be learned, and only small investments can learn.
Mediation should start from a small ceremony. "Can I help you mediate?" This is the first step in the event of a conflict.
Experience:
21, human error
Place you in a dead, not something you don't know ... but it is "knowing" never place your death.
Experience:
22, personnel arrangement
In the early days, the personnel have forced the project to cross key design phases (this is to make all people can do things).
If the work is given to many people before the design is completed, then the interface between the working group is complicated between the people. This will increase the internal coupling of the team, the meeting time, duplication of labor and invalidation will increase.
The ideal staff is this: In most of the project, the small core team is designed by the small core team, and a large number of people have added a large number of people in the final phase of the development (the last 1/6 of time schedule).
Terrible guess: Time to arrange urgency projects, more reasonable projects, completed, will be longer.
Experience:
23, project sociology
Let people who don't have to be relieved to leave, thus maintaining the streamline. There is a disclosed agenda and strictly implemented, which is the easiest way.
The project needs a ceremony.
Use small rituals to pay attention to the objectives and ideals of the project: small-scale meetings, zero defects, and more.
Take action to prevent people from angry.
Remember: anger = fear. Just a manager who is angry with the lower level must be because fear will do.
Opinions: If everyone understands the truth of "anger = fear", you can clearly see that people are scared. Because he can no longer conceal your fear, he will not be angry. (This cannot solve the problem of these angry people, but it will definitely make others very much.)
Experience:
24, basic common sense
The project requires both the goal and needs plan.
And both of these should be different.
Experience:
to sum up:
