I am interested in using Jupyter Lab and Visual Studio Code as tools, with Plotly as a backend, and using hvplot in my Jupyter Notebooks in an offline environment. In order to do this, I would like to be able to set the resource mode to ‘inline’, ‘server’, or ‘cdn’ in the panel package.
If the resource mode is set to ‘inline’, the panel package should read all resources from bundled files and include them inline in the Jupyter Notebook. If the resource mode is set to ‘cdn’, the panel package should include all resources in the form of CDN URLs in the Jupyter Notebook. If the resource mode is set to ‘server’, the panel package should launch the Bokeh server internally and include all resources in the form of the Bokeh server URL in the Jupyter Notebook.
As I understand it, the current panel package allows you to specify whether to use inline resources or not. If inline is set to False, all resources are referenced from CDN URLs. If it is set to True, Bokeh-related resources are read from bundled files and included inline, but the rest of the resources seem to still be referenced by CDN URLs.
I hope this information is helpful in understanding my request. Thank you for your consideration.
It would be very helpful if you can include a minimum, reproducible code example so that the developers would be able to quickly understand the issue or request.
This is is something I have already fully implemented in the Panel branch-1.0 and it seems to be working well so it should be possible when 1.0 is released in late January.
I have been struggling with this problem for the past few months. When using ‘bokeh’ as the backend, I am able to get satisfactory results. However, when I try to use ‘plotly’ as the backend, I have not been able to get the desired results, even after experimenting under various conditions. Sometimes it appeared to be working, but it turned out to be an illusion caused by cached files when I was connected to the internet.
To resolve this issue, I would like to set up a fresh environment with the latest packages, such as panel 0.14.2, and perform further testing. I will report back with more specific details on my findings. Thank you for your attention.