2.5.0

Use relative paths for bookstore.

2.4.0

Bookstore now supports cloning from a local or network attached file system.

This feature is disabled by default.

To enable this filesystem (fs) cloning, set BookstoreSettings.fs_cloning_basedir to the root directory from which you want people to be able to clone.

Adding fs cloning offers users more flexibility to clone notebooks from attached filesystems, like EFS. For more information about the motivation and design of this endpoint, please see #154.

Release 2.3.1

Patch release to fix bug around issue where cloning would hang in certain circumstances.

2.3.0

This bookstore release adds a number of new features (e.g., cloning) and massively improves both the unit-testing frameworks and the documentation.

Major Changes

Cloning

As of 2.3.0 cloning from S3 is now enabled by default.

Cloning allows you to bring files from different S3 buckets locally and then opens them in an interactive Jupyter context. To use them, you will need to set up your configuration for any such bucket.

Massive Testing improvements

We have built out a framework for unit-testing Tornado handlers. In addition, we have added a collection of unit tests that bring us to a coverage level in non-experimental code of well over 80%.

Features and Version endpoint

You can identify which features have been enabled and which version of bookstore is available by using the /api/bookstore endpoint.

New Publishing endpoint

Previously our publishing endpoint was /api/bookstore/published, it is now /api/bookstore/publish.

REST API Documentation

All APIs are now documented at our REST API docs using the OpenAPI spec.

Experimental

Clients (subject to change in future releases)

To enable access to bookstore publishing and cloning from within a notebook, we have created a Notebook and Bookstore clients.

bookstore 2.2.0

This bookstore release builds up the 2.0 release by adding additional refinements to storage of notebooks to S3 and more integration tests.

Major changes

• Refactored the archive completely to use aiobotocore [@mpacer]
• Added the actual publish endpoint implementation and tests [@mpacer]
• Fixed the autoenabling configuration [@mpacer]
• Streamlined release process by removing versioneer [@willingc]

bookstore 2.0.0

This begins the new life of bookstore, which will provide tooling and workflow recommendations for storing, scheduling, and publishing notebooks.

This first release includes automatic archiving of notebooks to S3 on every save. We've got some performance enhancements slated for minor releases to check hashes prior. For now, this gets us going.

Wait, why 2.0.0?

Pre 2.0.0 was an old outdated module for OpenStack swift storage of notebooks. If you still need that, pin back to the 1.0.0 release of bookstore.