We’re building a solution that uses the Enhanza data in BQ as input and we need to decide what dataset to start building from and how to handle schema changes.
How often can we expect schema changes for the different datasets that are made available to use in BQ? Do these get announced ahead of time, or what is the process for announcing & releasing schema changes? Are changes additive (e.g. new tables or columns), or can changes also be breaking?
Knowing this will help us decide whether, for example, the API dataset is a better source for us to develop from than one of other ones.
Thanks in advance!
Those are very good questions. As you might know, the dataset fortnox_api contains the raw data, directly from fortnox API. Those will not change as long as Fortnox doesn’t change their API, which happens very rarely.
Howevery, best practices to start building queries is from fortnox_bi. We see them as building blocks and normally start with them ourself. They are the next step after fortnox_api.
If you want even more prepared data, you can use fortnox_flat. Those are build on fortnox_bi.
But, and this might be your strategy; we are fully transparent with all our queries in Bigquery. You might want to setup your own Bigquery project, if you haven’t done so already, and keep all queries there? Then you will have fully control over them.
Could that be your way forward? Please let me know if you want to discuss it in more details.