Notebook Cards vs JupyterLab Notebooks

Inside a workspace you can work with two kinds of notebook: notebook cards and JupyterLab notebooks. Both are ordinary .ipynb files, both run on the same kernels, and both expose the same JupyterLab toolbars. The difference is how they are created and where they show up on the platform. Mostly that difference is about portal visibility, not how the notebook itself behaves.

What a notebook card is

A notebook card is a notebook the platform tracks as a managed object. Cards are created from the Syntasa portal — there is no way to create one from inside JupyterLab. Once created, a card lives in the workspace's notebooks/ folder and shows up in the portal alongside apps, dashboards, and other Syntasa artifacts. You can drop cards into Syntasa folders to organize them as part of a larger solution.

What a JupyterLab notebook is

A JupyterLab notebook is exactly what it sounds like: a regular .ipynb file you create from the JupyterLab file browser. It can live anywhere in the workspace — typically inside a team folder you create at the workspace root. JupyterLab notebooks are file-system-only; they do not appear in the portal.

Same format, same infrastructure

Both kinds of notebook are ordinary Jupyter notebook files. They run on the same kernel images, support the same Python and Scala runtimes, and expose the same toolbars: runtime attachment (Runtime Attachment in Notebook Workspaces), init script editor (Init Scripts & Dependencies), kernel logs viewer (Kernel Logs Viewer), Spark UI access (Spark UI Access). Anything you can do in one, you can do in the other.

Visual difference when you open one

Cards open in JupyterLab document mode: there is no left-side file browser and the File menu is hidden. The notebook fills the window. JupyterLab notebooks open in the standard tabbed view with the file browser on the left and the full menu bar at the top. The editing surface itself — cells, toolbars, status bar — is identical in both modes.

Card storage and naming

Cards live at notebooks/{card-folder}/{card-file.ipynb} in the workspace root. The platform blocks renaming or moving anything inside the notebooks/ folder from JupyterLab — attempts to rename a card folder or its file return a 400 error. This is intentional: the link between the portal listing and the on-disk file would otherwise be easy to break. If you need to rename a card, do it from the Syntasa portal.

Sharing inherits from the workspace

Cards do not have their own sharing controls — the Share option is disabled in the card creation UI. A card's audience is always the membership of its workspace; see Notebook Workspaces  for the membership model.

This is the same access model that applies to plain JupyterLab notebooks in the workspace: any workspace member can open them. So, cards do not give your teammates more access than they would have on a JupyterLab notebook in the same workspace — what cards add is portal listing and integration with apps and dashboards, not access.

Notebook Process works on both

Both cards and plain JupyterLab notebooks can be scheduled or triggered as a Notebook Process — see Notebook Process (Jobs). Cards do not have a monopoly on scheduling.

When to use which

A simple decision guide:

• Default to JupyterLab notebooks for new work. Create new notebooks from the JupyterLab file browser and keep them in workspace folders that make sense for your team. Everything in this guide — runtime attachment, init scripts, synutils, logs, security — works the same way for JupyterLab notebooks as it does for cards, and the JupyterLab editing experience is what most users want most of the time.
• Use a notebook card when you have a specific reason to. Cards are kept primarily for backward compatibility with existing platform usage. The one situation where a new notebook is better off as a card is when you want it to live in the Syntasa Workspace alongside related apps, dashboards, and other artifacts as part of a solution — for example, putting the notebook into a Syntasa folder that organizes everything for a particular product or project. If you don't need that organization, a JupyterLab notebook is the simpler choice.

Feature comparison