The first two of an ongoing series of official Datasette tutorials are now available: Exploring a database with Datasette introduces the Datasette web interface and shows how it can be used to explore a new database, and Learn SQL with Datasette provides an introduction to SQL using Datasette as a learning environment.

Datasette 0.60 adds a new filters_from_request plugin hook, new internal methods for writing to the database, better performance and various faceting improvements. See also the annotated release notes.

Observable notebooks recently added a SQL cell type, allowing SQL queries to be executed as part of an interactive notebook workflow. Alex Garcia built a Datasette Client for these which allows you to excute queries against any Datasette instance and explore and visualize the results using JavaScript code running in a notebook.

Datasette 0.59 adds column descriptions in metadata, a new register_command plugin hook, enhanced --cors support and a bunch of other fixes and documentation improvements. See also the annotated release notes.

Datasette Desktop is a new macOS desktop application version of Datasette, which supports opening SQLite files on your computer, importing CSV files and installing plugins. I wrote more about how it works in Datasette Desktop—a macOS desktop application for Datasette.

The Baked Data architectural pattern describes a pattern commonly used with Datasette where the content for a site is bundled inside a SQLite database file and included alongside templates and application code in a deployment to a serverless hosting provider.

Datasette 0.58 has new plugin hooks, a huge performance improvement for faceting, support for Unix domain sockets and several other improvements. Read the annotated release notes for extra background and context on the release.

Datasette 0.57 is out with an important security patch plus a number of new features and bug fixes. Datasette 0.56.1, also out today, provides the security patch for users who are not yet ready to upgrade to the latest version.

Django SQL Dashboard is a new tool that brings a useful authenticated subset of Datasette to Django projects that are built on top of PostgreSQL.

Datasette 0.56 has bug fixes and documentation improvements, plus some new documented internal APIs for plugin authors and SpatiaLite 5 bundled with the official Datasette Docker container.

Datasette 0.55 adds support for cross-database SQL queries. You can now run datasette --crossdb one.db two.db and then run queries that join data from tables in both of those database files - see cross-database queries in the documentation for more details.

sqlite-utils 3.6 adds similar features: a db.attach(alias, filepath) Python API method and --attach alias filepath.db command-line option, both for attaching additional databases in order to execute cross-database queries.

This new Video introduction to Datasette and sqlite-utils provides a full introduction to both Datasette and sqlite-utils in 17 minutes, including a live demo of creating a database from a CSV file and publishing it to Google Cloud Run.

Serving map tiles from SQLite with MBTiles and datasette-tiles. datasette-tiles is a new plugin that adds a tile server to Datasette, serving map tiles from databases that conform to the MBTiles specification. download-tiles is a tool for building these databases, and datasette-basemap is a plugin that bundles a 22MB SQLite database with OpenStreetMap tiles covering zoom levels 0-6 for the entire world.

Datasette 0.54 is out today. Highlights include the new _internal in-memory database exposing details of connected tables, plus support for JavaScript modules in plugins and add-on scripts. More commentary on this release is available in the annotated release notes.

Drawing shapes on a map to query a SpatiaLite database introduces the new datasette-leaflet-freedraw plugin, which adds support for drawing shapes on a map to specify a GeoJSON MultiPolygon that can be used to query SpatiaLite databases.

• New --properties option for writing GeoJSON properties to a JSON column called properties - or use --properties=props to write to a column called props. Thanks, Chris Amico. #25

• New hash_id_columns= parameter for creating a primary key that's a hash of the content of specific columns - see Setting an ID based on the hash of the row contents for details. (#343)
• New db.sqlite_version property, returning a tuple of integers representing the version of SQLite, for example (3, 38, 0).
• Fixed a bug where register_function(deterministic=True) caused errors on versions of SQLite prior to 3.8.3. (#408)
• New documented hash_record(record, keys=...) function.

Mark function as deterministic on Python 3.8 or later.

What's Changed

• Require click >= 7.0 to hide commands by @1-Byte in https://github.com/eyeseast/geocode-sqlite/pull/20
• Fix error with rowid tables by @eyeseast in https://github.com/eyeseast/geocode-sqlite/pull/27

Full Changelog: https://github.com/eyeseast/geocode-sqlite/compare/v0.5.1...v0.5.2

• The dateutil_parse(), dateutil_parse_fuzzy(), dateutil_parse_dayfirst() and dateutil_parse_fuzzy_dayfirst() functions all now accept an optional second argument specifying a "default" datetime to consider if some of the details are missing. #5

• Fixed bug where potential URLs inside of <pre> links were being turned into links. #11

• Initial release. #1

• README now includes a description of the file format and a link to a demo. #5

• Redesigned the schema to reflect that Google Drive files can only have one, not multiple owners. #34
• Removed drive_folders_owners and drive_files_owners tables.
• drive_files and drive_folders now have a _owner column that is a foreign key to drive_users.
• New google-drive-to-sqlite export format file_id command for exporting Google Docs files as different formats such as pdf. #21
• New options to files: --apps for all Google Docs files of all types, or --docs, --sheets, --presentations and --drawings for Google Docs files of specific types.
• google-drive-to-sqlite files and get commands now have a --verbose option showing what the tool is doing in detail.
• Automatically retries up to twice on HTTP transport errors, with a 2 second delay. #18
• The auth.json file now defaults to 0600 permissions, meaning only the user can read that file on a shared system. #37

• New design for the database schema: #9
• drive_files contains files, with a _parent foreign key to the parent folder
• drive_folders contains folders, with a self-referential _parent foreign key
• drive_users contains users who may own or have modified files
• drive_folders_owners relates folders to their owners
• drive_files_owners relates files to their owners
• The full schema is now included in the documentation.
• New --import-json and --import-nl options to the files command for creating a database using JSON data retrieved previously using the --nl and --json options. #20
• --starred, --trashed and --shared-with-me options for files as shortcuts for constructing more advanced search queries. #25
• Documentation now has a quickstart section. #24