After receiving requests for a self-hosted version of Dolt sharing hub DoltHub, the team behind the Git for data project came to see that sharing private data isn’t everyone’s cup of tea and pushed a self-managed flavour of the product into the open. Roughly one month in, DoltLab is now making the jump to version 0.2, quenching users’ thirst for more UI features.
Developers can now edit Dolt databases on the web or merge pull requests through the UI, which should make the project as a whole a little more accessible to those not too familiar with the command line. Collaborator management is also said to have become easier, and admins will be interested to hear that the tool learned to use established SMTP servers to send emails for account management tasks. Other enhancements include the option for users to upload files, just as they can with GitHub.
The last two changes mean that organisations that have been using DoltLab need to adjust their setup slightly. Amongst other things they need to open another port so that the file server can be accessed through the browser, and set additional environment variables supplying DoltLab with the right credentials for email exchanges.
SQL database Dolt has been around since 2019 and means to provide a literal form of database version control by allowing users to perform command line operations they know from Git on table rows instead of files. To round off the experience, Dolt databases can be shared via a database management web GUI called DoltHub, which is similar to GitHub. DoltLab is its self-managed equivalent.
The database space seems to see its fair share of new projects progressing these days, as for instance relational database project EdgeDB just recently tried to win teams over with its 1.0 release. Its makers describe EdgeDB as a relational database with an object-oriented data model meant to help users build safe software and, while DoltLab seems to be quite happy with the query language, “do better than SQL”.
The project tries to achieve this by featuring a protocol designed to minimise the number of server roundtrips, a new query language with a rich type system and declarative schema, and using PostgreSQL under the hood.
Other aspects hoped to make EdgeDB stand out are the integration of data relationships as a first class concept, the fact that everything is treated as expressions with just one class of values, and the promise of “straightforward navigation and manipulation of non-trivial graphs of data”.
With the bump to v1.0, the EdgeDB creators mainly aim to signal the stability of their project to potential uses. Moving forward, releases are meant to become available in shorter cycles and also in the form of a cloud service, to make the project easier to integrate.
If any of the new releases will be able to help to get more users onto these database projects remains to be seen, especially since jumping ship on databases makes for a lot of migration work not everyone has the capacity to muster. EdgeDB and DoltLab might however be worth considering as an option for new projects, as for example Dolt’s versioning and diffing options could come in quite handy for debugging purposes amongst other things.