If kernel saved into notebook not found locally, language server does not start properly #16322
Open
1 of 2 tasks
Labels
bug
Issue identified by VS Code Team member as probable bug
Applies To
What happened?
I've been having this issue for quite some time on different devices.
When I open a notebook that hasn't been modified/saved ON THE SAME MACHINE the last time it was opened, choosing a kernel does not properly start the language server. This means that I have to cycle through the kernel picker one more time to properly start the language server.
Once I save the notebook and reopen it on the same machine this doesn't happen anymore.
I'm pretty sure this has something to do with the kernel spec that is saved into the ipynb file, because I'm working on the same file in a dropbox folder synced across different devices: once I save a file and open it on another device I always have to go through this. I guess at startup vscode tries to match the kernel found in the ipynb to the ones found locally but it fails because this file was last saved on another device.
I'm attaching a screen capture which may explain the situation better.
jupyter_vscode_bug.mov
VS Code Version
1.96.1
Jupyter Extension Version
2024.11.0
Jupyter logs
Coding Language and Runtime Version
No response
Language Extension Version (if applicable)
2024.22.0
Anaconda Version (if applicable)
No response
Running Jupyter locally or remotely?
Local
The text was updated successfully, but these errors were encountered: