Working | Original
-
Before gaining access to the code repository, ask colleagues for the
pom.xmlfile and try to download Nexus dependencies. -
Without actual code but with the Nexus libraries, we can actually decompile the JAR or experiment with the libraries. We can do quite a few things. Think outside the box.
-
It is preferable to configure User Tokens in the Sonatype Nexus Repository in
settings.yaml. -
Don’t wait for access request ticket approvals. Before you have access rights, think about what you can do.
-
Much of the job can be done before you start working. One can familiarize themselves with almost everything in advance. Everything has open-source substitute code or materials.
-
A job often involves specific settings, slightly different code logic, and access rights or passwords to get things done.
-
Foresee the next phase of the project. Consider how the current project will progress and what users will need or expect.
-
Think of all possible ways to do the job, make progress, and prepare as much as possible before you get access rights or the materials others will provide.
-
Don’t settle; get things done and automate as much as possible.
-
Use the Python
requestslibrary to replace Postman. -
Windows and PowerShell are also viable options. The commonly used commands are limited to a few dozen. Learn them well.
-
Consider how an engineer might work on this project for 5 or 10 years, and think about how to achieve similar efficiency quickly.
-
Use notes and logs directories. Carefully copy logs from the console or frontend pages for detailed analysis.
-
Think about how to convey this knowledge to newly joined team members.
-
Think about what we can achieve if everyone does their best and we have 50 such engineers.
-
Libraries that can be easily downloaded from internal Nexus repositories can be manually downloaded and placed in the
.m2repository. -
Read the
pom.xmlfile to determine the Maven compile target and the required JDK version for the project. -
Instead of searching Jira or Confluence, directly construct a URL to go to the profile and locate pages.
-
If we encounter a bug, for example if
npm run buildhas problems in a large build pipeline, we can run it separately to ensure it works before running the entire pipeline. This makes the debug cycle faster and smaller. -
The problems or similar bugs will recur repeatedly throughout your career, so make sure you keep notes about them. This practice remains useful even in the AI era. Sometimes, large language models provide answers that aren’t entirely accurate and require you to verify them.
-
In some large companies, the progress of AI chatbots or internal AI tools may be slow. We can use our personal computers to learn related public technologies with AI tools after work. This makes it less cumbersome to work on corporate computers since we are already familiar with a lot.