Library branch management

[geoffturk]

There are two continuously open branches in the Library repo:

• main (the production/live branch)
• test (the testing branch)

Unlike with a code repo where we would aim to push all changes from test to main so that the production branch is 100% consistent with the testing branch, it is helpful to "cherry pick" from test the fields and schemas that eventually end up on main.

For example, we do not want to pollute the production branch with schemas used only for testing.

Also, we may want to phase out fields and schemas over time, so this can first be done on the testing branch before ultimately removing them from the production branch. If we are simply pushing all changes from test to main, there is no way to time the implementation of phased out fields/schemas in production alongside the introduction of new/updated fields/schemas.

To make clear when a schema is only meant to live in the testing branch, we can add the test- prefix to the id inside the schema: