Top Plan Ape 10 best Practices article

This article turns from: Eetproject teacher Album

Http://forum.eet-cn.com/FORUM_POST_10011_1200263220_0.HTM?click_from=8800111934,6106462476,2014-04-18,EECOL,NEWSLETTER

watermark/2/text/ahr0cdovl2jsb2cuy3nkbi5uzxqvtgvleglhb2jpyw8=/font/5a6l5l2t/fontsize/400/fill/i0jbqkfcma==/ Dissolve/70/gravity/southeast ">

1. Choose the 1th language carefully

the programming languages are different. Just a little different. But the difference between people in language is big. When you choose a language, you choose a community.

Suppose you want to build prototypes at a high speed (especially for founders who want to iterate over the product). Well, use Ruby or JS .


2, you are not a programmer!

Great developers can build and develop applications. Amazing developers can do this at the same time as the business. The People's Congress on the business side don't know the code. But you can certainly understand the motives behind a particular feature.

What other people say about what the developers do, not to understand why they do so, both sides miss the opportunity. This kind of thing is too common.

Proficiency in programming is a lofty career goal.

Once you've achieved this goal, don't forget to think about yourself.

Don't be a company or waste your time on worthless things.

if you want to finish the schedule. Project communication (even if this is not a prerequisite) prior to the start of the technical work, due to the ever-changing response of others.


3, keep agile, continuous delivery

early announcement, constantly announced, side rap side announced.

Constantly testing.

A good test package is a combination of a policy and a canary in a coal mine.

It can help you to find errors earlier in the production cycle. And the sooner the error is found the easier to solve.

High-speed failure. Coding (and living) I want to know as early as possible where not to work, and not let it proliferate or spread.

Full release. High-speed failure, repair defects. Continued.

Write your own active test for all your code!

The

Practice test-driven development as much as possible.


4, keeping control of the test process

    Many projects are suffering from many test cycles. This can slow down the project, leading to high-level issues in the Organization as a whole.

Program apes should focus on unit tests and semi-regression tests of their own code. They know the code base better than anyone else, and who knows what changes they can affect. Sometimes such changes are missing due to the limited scope of the QA test, resulting in significant problems in the production process.

If you want to develop a defect-free code as soon as you can, never try to put it back.

We are more aware of this.

To check the coverage of the test. Ensure 100% no dead ends.

5, if you are a freelancer, to learn to say no, even if the face of money

have a reasonable assessment of time and cost. and double it. Suppose everyone says, "This should be very easy," then do

6, Honor belongs to the past-theory is one thing. But the practice is more important

the best way to improve the quality of software development is to develop software.

Many ambitious start-up project architects spend a lot of time reading, about the latest tools, about the open process. That sort of thing.

A lot of people spend their leisure time so much. But it's very easy to delay you. Don't do that by using your brain as much as possible to strengthen the part of your mind that is responsible for developing software.

continue to explore. Many of the coders I've seen have several business projects on hand.

Doing Business projects forces you to explore new technologies and then learn to create all aspects of your application.

You may need to do front-end html/css, backend API integration, database optimization. Do mobile apps. You also have to set up your own server.


7, pair review is your secret weapon

    knot pair programming is necessary. Two program apes jointly develop the same module that can examine each other's code.

The development team also holds weekly code review meetings. Let each developer give feedback on other people's code to explain how to improve the code better. This can create a collaborative culture that puts the developer's ego aside.

8. Avoid premature optimization like a plague

Refactoring is only done when problems and solutions are present in front of you-premature refactoring is a huge waste of time. Do not invest half a year later may be thrown away no matter what is intact. Premature optimization is the source of evil.

Do not optimize prematurely!

I keep seeing the project master worrying about scaling up to 1 million of users before they reach 1000.

9. Your code is written only once. But others will read it thousands of times

You write code that the machine will parse to run. But others need to read your code, understand it, and tinker with it. You have to be clear that your code will have a future audience.

Code is also a written form of communication.

It sounds strange, but you always have to think about your future. Ask yourself: Let's say you have amnesia. Can you understand the code you've written?

Read through your documents. There are so many changes in design that sometimes staring at code updates doesn't necessarily follow.

Keep the document updated. The future of people (including yourself) is easier to understand. I don't know how many times I've looked back at my code and I always thought, "What the hell am I doing?" "Just want me to write a good stare." There are a lot less headaches in the future.

10, this is a lofty occupation. Use your skills in a good place.

helping others is a deep-seated need for people. Find ways to use your work to improve the human race and you will have a grasp of success.

Top Plan Ape 10 best Practices article