The release of Hasura 2.0 included one of the most exciting features to date - multiple database backends, and in particular, support for our first new backend, Microsoft SQL Server. On the face of it, this probably seems like a relatively straightforward change.

This release contains some large refactorings to the core of the Hasura product, as well as some large new feature additions. I’m going to talk about the software engineering tasks we tackled, challenges we encountered and how we met them.

In this post, we will look at how to programmatically generate or manipulate the Hasura metadata using the Jsonnet data templating language.

In this post, I will cover the following: What are updatable views, How Hasura used updatable views, Key problems with updatable views (from a Hasura point of view) and Moving away from updatable views