.devcontainer for pandas-profiling

[Anselmoo]

Missing functionality

A .devcontainer could be nice for CodeSpace Users and Organizations for quickly working on pandas-profiling, especially for hotfixes and releases.

https://code.visualstudio.com/docs/remote/containers

Proposed feature

Adding a .devcontainer folder including Docker-Image and devcontainer.json

Alternatives considered

No response

Additional context

No response

[sbrugman]

Hi Anselm, thank you for contributing. I'm not familiar with this devcontainer.
The benefits are clear from the linked documentation, but not the need.
What other projects are supporting this feature? Is this config needed at repo level, or can it be provided by individual developers?
Given that it's a commercial microsoft/github feature, I'm converting the issue to a discussion so that we can poll interest for this feature.

[Anselmoo]

Hi Anselm, thank you for contributing. I'm not familiar with this devcontainer. The benefits are clear from the linked documentation, but not the need.

What other projects are supporting this feature?

AutoKERAS is a prominent example, where I saw it for the very first time. Furthermore, they are using a specific TensorFlow image, and vice versa, if there is a pandas-profiling image, this could be used, too?! I assume.

A selected list of REPOS using .devcontainers + mcr.microsoft.com/vscode/devcontainers/

1. Ruby Rails
2. Prisma
3. jhipster

Is this config needed at repo level, or can it be provided by individual developers?

The config is provided on repo level.

https://github.com/keras-team/autokeras/blob/master/.devcontainer/Dockerfile

FROM [tensorflow/tensorflow:latest](https://hub.docker.com/r/tensorflow/tensorflow)
RUN apt-get -y update
RUN apt-get -y install git vim
COPY setup.sh /setup.sh

They are using a MS free Docker Image. In general, the MS images are coming from https://hub.docker.com/publishers/microsoftowner

Especially, in combination with VSCode + Jupyter it might have the potential to provide a ready-to-go solution for contributions without taking care of all requirements.

Given that it's a commercial microsoft/github feature, I'm converting the issue to a discussion so that we can poll interest for this feature.

Sounds like a great idea ... 👍