Welcome to JupyterLab Real-Time collaboration documentation!#
From JupyterLab v4, file documents and notebooks have collaborative
editing using the Yjs shared editing framework.
Editors are not collaborative by default; to activate it, install the extension
jupyter_collaboration.
Installation using mamba/conda:
mamba install -c conda-forge jupyter-collaboration
Installation using pip:
pip install jupyter-collaboration
The new collaborative editing feature enables collaboration in real-time between multiple clients without user roles. When sharing the URL of a document to other users, they will have access to the same environment you are working on (they can e.g. write and execute the cells of a notebook).
Moreover, you can see the cursors from other users with an anonymous username, a username that will disappear in a few seconds to make room for what is essential, the document’s content.
A nice improvement from Real Time Collaboration (RTC) is that you don’t need to worry about saving a document anymore. It is automatically taken care of: each change made by any user to a document is saved after one second by default. You can see it with the dirty indicator being set after a change, and cleared after saving.
If a file changes outside the shared session, for example through a third-party editor or a Git branch switch, each user can choose independently to:
Open original file: replace the current shared session with the disk version.
Save As…: enter a new filename, save the current shared content there, and continue with this session.
Close tab: close the current shared document.
Something you need to be aware of is that not all editors in JupyterLab support RTC
synchronization. Additionally, opening the same underlying document using different editor
types currently results in a different type of synchronization.
For example, in JupyterLab, you can open a Notebook using the Notebook
editor or a plain text editor, the so-called Editor. Those editors are
not synchronized through RTC because, under the hood, they use a different model to
represent the document’s content, what we call DocumentModel. If you
modify and save a Notebook with one editor, the other session detects an external change
and prompts its users to choose how to proceed as described above.
Edits within the same collaborative session synchronize automatically. External changes are handled separately so they cannot silently replace the content collaborators are viewing.