[go: up one dir, main page]

Manage:Import Consider being GraphQL-first for new features

As backend team gains more & more experience with adding GraphQL-endpoints during our work on ~"group import", we should have significantly lower friction to build our APIs in GraphQL-first side

As frontend engineer our ~"group import" feature already uses GraphQL internally for local representation and I will be more than happy if we will use GraphQL for building our endpoints. This aligns with overall strategy (https://about.gitlab.com/handbook/engineering/development/#graphql https://about.gitlab.com/handbook/engineering/#balance-refactoring-and-velocity where we name "graphql" a preferred approach and I believe will increase reliability of our code at least in the area of backend-frontend interaction and not having a private APIs which are currently widely used across the group

@gitlab-org/manage/import/backend WDYT? I will be more than happy also to contribute in writing resolvers on backend side, but as usual - as a frontend person I might be missing something and underestimate how much effort will it require for us to build our API's in graphql-first way

/cc @hdelalic @dennis