Feature request: ipykernel as a project dependency option

[hunterhogan]

For this description, pretend "Quick Create" does not exist.

1. Select Create Environment.
2. (Select environment manager.) Assume venv.
3. Select a Python Environment.
4. Enter directory name.
5. Select Install project dependencies.
6. ms-python.vscode-python-envs magically creates a list of good project dependencies the user may want to install.

The change

When ms-python.vscode-python-envs is searching for dependencies, add the following:

1. Look in the list of VS Code extensions for ms-toolsai.jupyter.
2. If found, get the "Select a Python Environment" value.
3. Figure out if VS Code has access to an ipykernel for that Python version. (I suspect you could do this by calling a function in ms-toolsai.jupyter.)
4. If VS Code has access to the kernel install, offer the kernel as a project dependency in step 6, above.

ms-toolsai.jupyter

I'm miffed that the folks at ms-toolsai.jupyter don't understand or don't care. See microsoft/vscode-jupyter#16839.

• A long time ago, installing a Microsoft product typically required the user to babysit the installation and make 806 choices at 467 different points in the installation process. At the moment, ms-toolsai.jupyter is like that, but worse. I may think I have Jupyter installed, so I try to use it (such as by running an interactive window), and VS Code asks me more questions about installation.
• "The way users address these issues is by having does such a requirements.txt or the like": generalized, "Instead of a small group of people changing this, a large group of people should just learn to adapt." That statement is always wrong. Some similar statements can be right, but 1) this statement is always wrong, and 2) I'm already giving away a valuable education for free.
• I heard a rumor that Microsoft has software--for internal use only--that will find and correct spelling and grammar errors: the utopian concept is to improve communication.

Not a binary choice

This feature, or variations of this feature, could (should?) be implemented by both ms-toolsai.jupyter and
ms-python.vscode-python-envs.

Quick Create

Handle the kernel dependency the same as the other dependencies.

To paraphrase Blaise Pascal, I wrote a post of appropriate quality because flawless writing requires an unreasonable amount of time.

[eleanorjboyd]

Thank you for the information and and I agree we want to have jupyter notebooks and the python environments extension work very well together. Want to loop in @DonJayamanne and @rebornix - is this something you have considered from the Juypyter side?

[github-actions]

Thanks for the feature request! We are going to give the community 60 days from when this issue was created to provide 7 👍 upvotes on the opening comment to gauge general interest in this idea. If there's enough upvotes then we will consider this feature request in our future planning. If there's unfortunately not enough upvotes then we will close this issue.

[DonJayamanne]

@eleanorjboyd @hunterhogan
If you try the following approach, the required packages (including IPyKernel) are installed .

• Open notebook
• Run a cell
• Now you're prompted to select a kernel
• Select Python Environment -> Create Python Environment

[hunterhogan]

@eleanorjboyd @hunterhogan If you try the following approach, the required packages (including IPyKernel) are installed .

• Install the Jupyter extension.

* Open notebook

* Run a cell

* Now you're prompted to select a kernel

* Select `Python Environment -> Create Python Environment`

iftfy

[hunterhogan]

@eleanorjboyd @DonJayamanne

If you implement the following feature, microsoft/vscode-jupyter#16839, the required packages (including IPyKernel) would be installed.

• Install Jupyter by Microsoft

[eleanorjboyd]

Seems like that feature is implemented so this is resolved?

[hunterhogan]

I don't know whether or not it has been resolved: I added ipykernel to my pyproject.toml files so I wouldn't have to deal with this. Nevertheless, I just did a quick, non-definitive test that suggests to me that kernels are not installed. I added a new 3.14 venv without installing my dependencies and this screen seems to suggest the 3.14 kernel is not present: