Is there any difference between opening a jupytext file vs. paired ipython notebook in jupyter?

Sorry to pollute your issue tracker (this isn’t a bug, but I can’t find the answer in the documentation). In jupyter, I can open both jupytext python scripts and ipython notebook files as “notebooks”. For example (note the file extensions):

If I open a jupytext .py script and save it in jupyter, a paired .ipynb file is created – but the .py file name still appears at the head of the tab rather than the .ipynb file.

Is there any functional difference between “opening” the jupytext .py script vs. the paired .ipynb notebook file in jupyter? Would “opening” the actual .ipynb file result in better performance for any reason?

[mwouts]

Oh sorry @lukelbd I missed your question. Thanks for the reminder!

If you are working with an unpaired .py file, since there is no .ipynb file, where is the cell output data physically stored?

I think it is not stored anywhere (other than in the memory of your web browser). I think the outputs are also transmitted back to the server over the network (so you should not expect performance improvements there), but the server ignores them.

And under what circumstances can jupyter “forget” the cell output data?

In my experience, the cell outputs are lost as soon as you reload the notebook.

[mwouts]

Great idea! I’ll do that.