User Experience of AWS CLI #8293
mmazurowski
started this conversation in
General
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hey 👋
I wanted to start by saying that I am very much impressed with how stable AWS CLI is and how well it can be integrated into workflows.
Given that, I also wanted to understand better some of design decisions behind the particular commands. Some of AWS APIs are very easy to grasp in cli format, for example S3, IAM, cloudwatch.
Some, for example DynamoDB feel very much low-level. To give particular example in order to create one shell script with DDB Query you need to do some manual labor of serializing condition expressions to string or pass json file with attributes as per docs.
While this is perfectly fine for things that you do one time and forget, being able to memorize commands, use autocompletion or have helpers to solve that kind of issues would allow to use CLI even more frequently on a day to day work. I could imagine that serialization and attribute subsitution could be done automatically similarly how it's done by many DDB packages for node or other languages.
Example:
Above is not a critique, I am sure that there are / were good reasons to leave some of work up to the person who will use CLI. I would really love to understand what factors decided with such design. While my use-case could be easily solved by writting CLI application to wrap SDK and expose API that's easier to use, I feel having official tools to handle that kind of workloads would be beneficial for the community.
Let's imagine following pseudo code:
My assumption is that it lies behind code-generation tools for smithy models as most of SDKs are generated based on that
Beta Was this translation helpful? Give feedback.
All reactions