This article about how to manage several graphcool stages was very important. Thank you for writing it.
In the current process of all of our deployments, a schema goes with an app revision. Code @commit revision XYZ needs the schema in revision ABC. With our current approach (used by sequelize/doctrine/dbdeploy/rails …), based on schema migration deltas we have this simplicity. We don’t have in the codebase a production or development schema, but only A single schema represented by a sum of migrations. This schema got eventually deployed in revision ABC in staging, in revision DEF in prod etc.
How would you implement such a workflow with graphcool? Having a single schema, plugged to two different envs? The push command doesn’t allow to specify which env. Would you use a simlink to duplicate the same file?
