Currently GCP project ID is a part of config.json, and is typically interpolated from secrets.env. However, this project ID is already present in the JSON key file, gcp.json. This is probaly implemented this way to be compatible with potential alternative GCP credential schemes. However, it creates a significant user flow difference between AWS and GCP deployments, and we won't be supporting alternative GCP credential schemes any time soon (and probably ever). This field should be removed from config.json, and the deployer should parse this field from gcp.json.
Currently GCP project ID is a part of
config.json, and is typically interpolated fromsecrets.env. However, this project ID is already present in the JSON key file,gcp.json. This is probaly implemented this way to be compatible with potential alternative GCP credential schemes. However, it creates a significant user flow difference between AWS and GCP deployments, and we won't be supporting alternative GCP credential schemes any time soon (and probably ever). This field should be removed fromconfig.json, and the deployer should parse this field fromgcp.json.