JupyterLab is the next-generation version of Jupyter Notebook, which provides more powerful functions and a more flexible user interface. On June 6, the official release of JupyterLab 4.0 is said to be the next major version.
The major improvements in JupyterLab are:
-
User Interface : Jupyter Notebook uses a single-document interface that displays open notebooks on a tab-by-tab basis. Each tab corresponds to a notebook. JupyterLab, on the other hand, provides a more flexible multi-document interface that can open multiple notebooks, terminals, text files, and other plug-ins in the same window at the same time.
-
Layout : The layout of Jupyter Notebook is fixed and the user can only scroll vertically between each cell. JupyterLab allows users to freely drag and drop and rearrange notebooks, code editors, output windows, and other panels in one window, making the layout more customizable and flexible.
-
File browser : JupyterLab has a built-in sidebar file browser for users to manage files and folders. This functionality is implemented in Jupyter Notebook by operating on the command line.
-
Extensibility : JupyterLab's architecture is more modular and extensible, enabling users to add custom plugins and extend functionality. This means that developers can add new features and tools according to their needs, making JupyterLab more suitable for specific workflows.
-
Terminal : JupyterLab has a built-in terminal feature that allows you to run command-line commands directly in the interface without opening an additional terminal window. This is very convenient for users who need to perform command-line tasks in an interactive computing environment.
For JupyterLab 4.0, the biggest update is faster, through CSS rule optimization, CodeMirror 6, MathJax 3 and other improvements. JupyterLab 4 is much more efficient than JupyterLab 3 when dealing with large notebooks.
To optimize performance, real-time collaboration (RTC) was moved to a separate package jupyter_collaboration, version 1.0.0 of which is now available. In this way, if we use it stand-alone, we don't need to install these unnecessary content. If you want to use RTC in JupyterLab 4, you need to install the jupyter_collaboration package.
In JupyterLab 4, a new extension manager is also included, so that it can be installed directly from PyPI without compiling locally, which is much more convenient for us to install.
The above is a brief summary of JupyterLab 4.0, the complete official release is here: https://blog.jupyter.org/jupyterlab-4-0-is-here-388d05e03442
Overall, JupyterLab provides richer features and a more flexible user interface, enabling users to better organize and manage notebooks and other related tools. But for the current AI-assisted programming, including ChatGPT and Copilot, JupyterLab has fallen behind too much. If you just need to simply run and share notebooks, using VS code+Copilot to run Notebook is still a good choice.