ChatGPT in Kubernetes Development
ChatGPT facilitates Kubernetes coding but with drawbacks
--
ChatGPT is taking the world by storm. Overnight, developers, designers, and product managers, or we can say, people from all over the world, are constantly testing its features and challenging its limit.
I am one of those who dip their toes into the pool to see how ChatGPT can help with coding. It is helpful from my point of view, especially in bash writing, where it beats me badly in Regex expression, for example, it can write a sed command in seconds, while might take me 20 minutes. It also provides excellent support for Golang, giving out examples and explanations for common issues and reminding you of the limitations of ReadFile
.
However, it is not smart enough to know that ioutil
has been deprecated since Go 1.16. So if I want to get the answer to the implementation in Go 1.20, the question should be more detailed. I typed “how to implement it in Go 1.20”, and finally get what I was looking for.
For Go native packages, ChatGPT can offer you the correct answers given detailed information in the questions. But, it is not so smooth sailing in Kubernetes development.
Kubernetes Development
YAML converts frequently between string and struct in Kubernetes, which can be implemented in several ways. Let’s try them one by one with ChatGPT.
- Use the gopkg.in/yaml.v3 package and its Unmarshal function can parse
byte[]
to anobject
, while Marshal function can do the reverse. When asked, ChatGPT responses with an example.