Gigantum Client 1.4.0 is now available. This is a significant release that adds new features and fixes. It is strongly recommended that all users update to this version. The two major changes in this release are improvements to the Authentication workflow and the ability to support multiple servers!
With this latest release and our new self-hosted Team Server option, you can easily switch between servers in a single local Client installation. Each server provides its own Authentication system and remote storage. This lets you to work with Gigantum Hub as you do now, or privately with a server that you control.
A new interface has been added to the Desktop app and a new command in the CLI (
gigantum add-server) to add additional servers to a Client.
Once a server is added, you are able to choose between servers when logging in. To switch servers, simply log out of the Client and select the server you wish to start working with.
To learn more about how to add and manage servers in the Client, check out the latest version of our documentation.
Interested in hosting your own Gigantum Team Server? Visit our website, click on "Contact Us", and fill out the form.
For the power users interested in details, this update also restructures the Gigantum working directory (typically
~/gigantum on you host machine). Data is now organized by "server ID". You'll notice that the first time this version of the client runs, a quick migration process will reorganize your files.
Projects and Datasets are now located in
~/gigantum/servers/<server-id>/<logged in username> where
gigantum-com for Gigantum Hub. As you add more servers, you'll see more folders get created in this location.
- Adds support for multiple servers!
- Client automatically configures using gigantum.com on first start if no server configs are present (#1511)
- On start, user data directories are migrated to new organization based on server id (#1469)
- Sensitive files and dataset file cache directories are reorganized by server ID (#1536)
- Datasets supported in self-hosted servers with configurable server-side encryption (#1452)
- Server configuration is automatically discovered via new .well-known services (#1409)
- Update authentication workflow to improve security and support multiple servers (#1505)
- Added a more useful message on container launch when a sensitive file is missing locally. (#1531)
- Moves api query to QueryRenderer and adds react context to provide and consume context for nested components (#1454)
- Client configuration is now cached in memory, reducing file IO during API calls (#1409)
- User data is now organized by server ID (#1409)
- Fixes publish button not locking when publishing a dataset. (#1521)
- Fixes login attempt not passing server id when tokens are expired. (#1539)
- Fixed an issue causing sync state to be incorrectly set when publish modal was opened (#1540)
- Fixes parsing of error message on the login screen. (#1532)
- Refactors detail records to fix issues with more/less buttons not functioning as expected. (#1507)
- The UI will now poll for an API response if it is not immediately available. (#1525)
- Fixes datasets listing not left aligning, fixes alignment and spacing issues on the select server page. (#1454)
- Fixes login token expiration bug, clears out tokens and identity before attempting login. (#1538)
- Fixes css overflow issue introduced by issue 1489. (#1515)