Scrum process management and programming management

 

As a Master, in addition to writing core code, the daily work is basically to participate in process management and programming management:

 

Work required for process management:

Precondition:

PO completes requirement writing

 

Work content:

1. Participate in the requirements review in the requirements stage

2. Split tasks in planning meetings, estimate hours, prioritize, and plan iterations

3. Record the cost in the development stage, follow up the experience, and conduct code review.

4. In the testing phase, perform good defect tracking.

5. Release the product for review and summary.

 

 

Required work in programming management:

1. Use git or svn for version control, create a new version before the start of each sprint, fix bugs in the old version, and merge into the new version after the repair is complete. (Version control may be different for each company)

2. Plan the code structure at the beginning of the project. There are 3 commonly used code structures:

1) "Chuan" type code structure

     All team members are responsible for one or some functions, and everyone has been doing it from the upper-level page, the middle business logic, and the underlying data access.

       The disadvantage is that there are few functional bugs that experts are responsible for, and many novice bugs, which affect the overall quality

2) "Three" type code structure

     All team members are developed in layers (generally presentation layer View , business logic layer Model , data access layer Data , and if it is MVC mode, there is also control layer Controller ).

     The advantage is that the quality is obviously better than the " Chuan " type code structure of vertical division of labor , because if the underlying code such as Model and Data that is repeatedly accessed is handed over to the master, there will be few defects in these codes; " In the same code, if a defect is found in one place, the defects in other places are also modified at the same time, and the repair cost of the defect is very low; and for the newly developed functions in the future, when these codes are called again, you can directly enjoy close to zero cost and zero defect. Effect.

3) L-shaped code structure

    It is suitable for loose pair programming. The underlying library is basically in charge of the master at the initial stage, and the novice can ask the master to try to develop and maintain the underlying library during subsequent business updates.

Guess you like

Origin http://10.200.1.11:23101/article/api/json?id=327078563&siteId=291194637