This is a discussion on project management through instant communication in a certain day and friend.
I will re-see the discussion, I found that I said more, less listening :). I am not like this. Mainly, I am more familiar with the company's company (development habits, programmers, deadline setting, etc.), very worried that this project can succeed, so it is desperately persuaded that friends have adopted some evasive risks.
Since the purpose in this discussion is to convince, many concepts are not necessarily accurate. This project has about 4 people to participate, using C developed applications on mobile phone platforms, estimated that it takes about half a year, should be considered a small project.
The following is the full text of the discussion:
I: How is your project?
Friends: in progress
Friends: The next week's detailed design begins
Friends: I haven't started writing code yet.
Me: You let them add the words, did they add it?
I: So I haven't started coding Luo yet.
Friends: Of course they want them to add
I: Do you agree?
Friends: I said that
Friends: I am head
I: Oh, I have to worry about orally.
Friends: Well
Friends: I will let them understand
Friends: How are you,
I: relatively relaxed, recently written GUI, use wxpyhon
Friends: Oh, master is master
I: Write a 10 line a day.
I: Rely, this is still a master
Friends: Oh. . . .
Friends: Microsoft is almost 10 lines per day
Friends: This is the level of Microsoft.
I: I have to get, Microsoft is much better than us.
Me: Our call is low efficiency
Friends: Well
Friends: I have recently seen "design model"
Me: Not bad
Friends: Learning to learn
I: Who do you plan to grasp the architecture of the program?
Friends: I have to study this now.
Friends: I see that I don't have this talent here.
Friends: Now I only rely on myself to architecture.
Me: Who is it?
Friends: no
Friends: I am
Friends: The scale of the project is not very big, I can try it yourself.
Me: No one can do it for the first architecture, and you have trouble and project management and architecture.
I: Architecture in the later period is the main goal of being complained
Friends: Yeah, I am now a few jobs. . .
Friends: I am now what I think about finding the following people to discuss, see if this architecture is reasonable.
Friends: Then set it down. . .
Me: It is best to delegate two people to make a structure, you have the last veto
Me: You guide
Friends: I don't have a special outstanding person now.
Friends: I can only be like this.
Me: You can also resist your responsibility.
I: I want to design the architecture in the form of the committee.
Friends: um ~~~
Friends: understand
Friends: I will allocate this way.
Me: Architecture This thing is hard to do, but everyone wants to do, I think it is from the perspective of the supervisor, and the architecture design can be regarded as a kind of remuneration, not to see it as a job.
Me: From the perspective of pay, everyone is the most experienced. Because it is difficult to do (because the risks are too big), everyone can reduce the risk
Friends: reason
Friends: more about this knowledge
I: Don't divide the module, and you have to decide the module and interface in the form of discussion.
Friends: Well
Friends: I am basically doing this.
I: Because the interface changes (or the architecture change) are very tired, it is also very sinful. So everyone decided, we must only blame everyone.
Friends: It's just that I will mention a module first, and I will discuss everyone.
I: I still make a module by others, you can modify it again.
Friends: The problem is that there is no other person here to do this.
Me: It is best to accept you, but also think it is your own opinion.
Friends: um ~~~~
Friends: reason
Me: The form is very important, otherwise it will inevitably have an idea.
Me: It is best not to make others think that the architecture is doing it yourself.
Friends: Yes. If the software development process, some unnecessary ideas do not have a lot of manpower me: I can compromise when necessary, give up my own good design, and use others. The key is to be unanimous.
Friends: Well
I: My understanding, from the perspective of management, it can be considered that the purpose of architecture design is to get people's heart, not the technical pursuit of perfection.
Friends: handsome guy, this is too cool.
Me: I am serious.
Friends: Yesterday I just saw such a sentence:
Friends: The architecture is like politics is the art of processing various uncertain factors.
Friends: When I saw what you said, I thought of the above sentence.
Friends: I am also very serious
I: If you let everyone discuss architecture design, there will be no uniform comments. As long as you are good at balance, more people, more compromise, more morality. Finally, everyone is very happy and has reached the same. As for the quality of the architecture itself, it will not be too bad.
Friends: It makes sense ~~~~
I: Ha, I am very close to him. I mean to turn architecture design into politics
I: The key is that you have to set it out.
Friends: You are also Master A
Friends: Well
Me: meat
Friends: 8-)
I: Participants are engaged in architectural (or interface), the more people, the better. I am afraid that you will follow the market, you can't persist in the form of the committee.
Friends: Well
Friends: Sometimes I think it is unity, I will use my power to regulate.
Me: Basic principle is that the interface is simple and less. You can let them get out of the interface, and you cut off half of the interface.
I: You'd better reflect the authority as an arbitrator.
Friends: Well
Me: It is not everyone's debate, you let everyone make half, compromise with each other.
Friends: This may make everyone get involved
I: I will be very tired as a party that arguments.
Me: There is no argument on the surface. It is not a morale.
Me: The actual morale depends entirely on everyone's opinion
Me: There are different faces to say that I have a good heart.
Friends: reason
Me: There is no argument on the surface, there is no morale.
Friends: Well
Friends: I fully agree
I: I actually do arbitration is very simple, even my mind is not used, it is to let everyone make halfway
I: compromise with each other
Friends: I think the most important thing for the manager is to let everyone have a strength.
I: Yes
Friends: Second is the project
Me: This is also the most beneficial to the project.
I: Because the project is high risk, it is necessary to use discussion compromise.
Friends: Well
Friends: The discussion of the project is very necessary
I: I don't have to have it, don't express your opinion.
Friends: Well
Me: It is best not to publish
Friends: This is not this, you don't post, you have no authority.
Friends: I feel compromising, it is published. . .
I'm right,
Me: I mean sometimes gives up your right opinion
Me: It is for compromise.
Friends: Compromise is also an alternative to an alternative to express your ability, prove that yourself in the project.
Me: I mean that it is best to avoid this kind of thing.
Friends: um. . .
I: and should be done.
Friends: understand. . .
I: You can catch quality control.
Friends: Well
I: Quality control you can watch "test driver development"
Friends: Good, must see
I: Drive the quality of development
I: You can take this part as long as you get the test.
Friends: Well, I fully understand this.
Me: This book tells a very important truth, which is the life pulse of development.
I: Important to tests should be preferred than the development. For example, write the test code first, write the development code.
Friends: Test is the lifeblood for development? Unfortunately, I know too little about the test now.
Friends: Seeing that I should see this book immediately.
I: is not a unit test, is an automatic test based on assertion (technique than the judgment to be advanced)
Friends: Oh ~~~~~~~~
Friends: "Test Drive Development" Who is written?
I: When you design, you have to turn off, use the structured design as much as possible, do not use object-oriented technology. Kent Beck Friends: Well
Me: Use inheritance as possible, polymorphism
Friends: I know this
Friends: Well
I: Access the data to pass the function, don't access it directly, you can do this.
I: The rest of the freedom
Friends: These are a bad performance in packaging
I'm right
Friends: I have limited life
Friends: I hope that they will play freely on the basis of certain
I: Well, mainly considering everyone's C level,
I: Polygon, inheriting is very easy to write difficult code
I: This is a place where C is easy to make mistakes.
Friends: C level is my most worried question
Friends: I have only one C programmer here.
I: Nothing is worried, put C as a structured C, quality is definitely guaranteed
Friends: good
Friends: I understand. . .
Me: You should put your energy in the automation of the development process when you start.
Friends: What does this mean?
Friends: Nothing to understand
Me: Is there any IDE now?
I: This is actually not conducive to multiplayer cooperation.
Friends: Yes
Friends: Well
Me: You can fully automate the process and deployment of the process and deployment of the process and deployment process through the script (the BAT file under DOS), the command line compiler, makefile. That is, you can complete the build and deployment as long as you point a mouse.
Me: I mean is built daily.
Friends: Oh ~~~~~
Friends: a little deep
I: This is more simple than writing code.
Friends: I have no research in this regard.
Me: Just write the dos ba file.
Friends: Oh
Me: For example, your app is to use two DLLs, an EXE.
Me: There are three people developed now.
Friends: Well
I: Open Visual Studio
I: Edit, press the Build button.
Friends:
Me: Let's copy these things to one place, package, right?
Friends: Well
Me: Now you ask you to completely use any programs with the GUI interface, write a simple script. Double-click this scriptor, do all things (including packaging)
Friends: understand
I: This is the industry popular management technology, which is doing this.
Friends: Good technology
Friends: understand
Friends:
I: Meet, you don't want to develop procedures, suppose you are a programming idiot, even Visual Studio doesn't know, how to manage it?
Me: Click on the mouse once.
Friends: huh
Friends:
I: This technology is useful, no need to program the foundation, actually the specific script can also be entrusted to others. You can ask if you ask.
Friends: Well
Me: The key is that you have to stand in the manager.
I: Suppose you don't understand anything, but you have to have the easiest and easy way to perform quality control.
Friends: Standing at the manager's angle ...
I: So you will try to remove people in the process.
Friends: Russia
I: I refer to people's negative factors, so they want to automate. Every day you have a job is a point and you can only have a mouse.
Me: You have a job every day.
Friends: um. . . . . . .
Friends: To do the above process, it looks very simple
Friends: I am afraid it is very difficult to realize it.
I: Yes, your management work is a melody.
Me: How can it be difficult?
I: Is it what to do before?
I: I have a unilateral to get the script, even if the IF is logic.
I: is simpler than VB.
Friends: This is achieved on the basis of the early development work.
Friends: The quality monitoring of the early development work is more difficult
I: So starting to integrate from the first day of development.
I: This is naturally conclusive.
Friends: This is a daily building a
I: There is no system to integrate this big stage.
I: I started system integration from the first day. I: Oh, huh
Friends: a bit of understanding
Me: You have to assume that you are programming idiot.
Me: You will have the only management job is a misunderstanding
Friends: Seeing is good practices
I: Can you give up management in the early stage?
Friends: can't
Friends: absolutely can't
I: So start the system integration from the first day of encoding.
Friends: um ~~~~~~
I: If you don't do it, let them find a way. Otherwise, you give up management (because you will only click on the mouse)
Friends: big touched
