Datalayer 0.4.0 Spitfire
Datalayer 0.4.0 Spitfire is released on January 17, 2024 and builds on top of the
Remote Kernels capabilities introduced in the 0'Malley release.
- Users receive applications to connect to any (local and remote) Kernels: REPL, Exec, Functions, Apps, Viewer and Dashboard. Those applications are specifically tailored for intensive AI data cases and aim to cover most of the Users' needs while analysing datasets with Jupyter.
- Devops receive more management capabilities: Traitlets, Manager and Operator. Those management user interfaces or libraries give more control and speed-up the Jupyter stack maintenance.
Supermarine Spitfire, is a British single-seat fighter aircraft used by the Royal Air Force and other Allied countries before, during, and after World War II. Many variants of the Spitfire were built, from the Mk 1 to the Rolls-Royce Griffon-engined Mk 24 using several wing configurations and guns. It was the only British fighter produced continuously throughout the war. The Spitfire remains popular among enthusiasts; around 70 remain airworthy, and many more are static exhibits in aviation museums throughout the world.
Jupyter Functions are serverless functions that Users define and call from any Jupyter Notebook. This release introduces "Deployed" Jupyter Functions which are cheap and lightweigh stateless variant of a statefull Kernel.
"Ephemeral" (a.k.a. "Temporary", "Spot", "On-demand") Functions are planned for the next Datalayer release and will allow to offload only predefined methods of Users entire code base.
Jupyter RTC implements collaborative rooms for notebooks and files. Users connect to rooms and have live editing capabilities with shared cursors, as well as commenting and reviews like Google Docs.
It also ensures to fine-grained access control to the data and enable collaborative access to Kernels.
Jupyter REPL allows Users to Read, Evaluate, Print and Loop code in Jupyter Kernel land. It is a drop-in replacement to the well-known IPython REPL, with remote Kernel execution capabilities.
Jupyter Exec allows Users to execute code in Jupyter Kernel land. Users can run code files locally or on any remote Kernel.
Jupyter Run allows Users to connect their Notebooks, Python and Markdown files to a remote Kernel or Function while leveraing RTC functionality.
Jupyter Apps allows Users to create various variant of application. IPyWidgets, Panel variants and supported, Solara variant is being planned for the next release.
Once created, Users can expose those applications to the world with a public facing URL.
Jupyter Dashboard allow Users to create and publish an interactive dashboard from Jupyter with drag-and-drop capabilities.
This is usable by technical but also business Users that can just reuse a Notebook prepared by a technical User.
Jupyter Traitlets allows a Developer to create a user intefarce from a Jupyter configuration (a.k.a. Trailets).
That user interface can be read-only or can also be used as for real management features to update the Jupyter component (server or lab extension, JupyterHub...) configurations.
Jupyter Manager is a user interface to manage the Jupyter platform in a more visual way instead of relying to shell script of configuration files.
Jupyter Operator maintains the state of the Jupyter system on Kubernetes, ensuring Users take control of the Identites, Access, Content and Kernels.