Are there alternatives to aws-mde/universal-image with CodeCatalyst custom devfile?

Yes, it is possible to use other public Docker images in a devfile with CodeCatalyst targeting VS Code, but there are some important considerations and limitations to keep in mind.

While CodeCatalyst provides default universal images that include commonly used programming languages and tools, you are not limited to only these images. You have the option to use custom runtime environment Docker images in your devfile, as long as they are public.

If you need a specific environment, such as one with Python 3.12 and pip 3.12, you can use a custom public Docker image that meets your requirements. This image can be from Docker Hub, Amazon Elastic Container Registry (ECR), or any other public repository.

However, there are a few key points to remember:

1. The custom image must be public. CodeCatalyst does not support private images in devfiles.

2. If you're using a VPC-connected Dev Environment, your options are more limited. In this case, you can only use universal images or private Amazon ECR images.

3. When using a custom image, make sure it has Git installed, as this is a requirement for CodeCatalyst.

4. Be aware that changing the image or component name in your devfile can result in data loss in the /home directory of your Dev Environment.

To use a custom image, you would need to specify it in your devfile. For example, you could use a public Python 3.12 image from Docker Hub or create your own custom image with the specific tools and versions you need, push it to a public repository, and then reference it in your devfile.

Remember to thoroughly test your custom image to ensure it works correctly with CodeCatalyst and includes all necessary tools for your development workflow.
Sources
Configuring a devfile for a Dev Environment - Amazon CodeCatalyst
Specifying runtime environment images - Amazon CodeCatalyst
Specifying universal devfile images for a Dev Environment - Amazon CodeCatalyst

Based on the AI answer, I tried redhat's universal-base image but found it less helpful than the amazon universal image. whereas, if I try quay.io/devfile/python:latest I do not get a working image.