0.5.1

Mostly bugfixes:

• Fix incorrect reference on json views, making some family empty contents.
• Fix DB docker image on kubernetes that was missing the logs
• Add verification of existing files before file modification operations
• Fix incorrect common refactoring of global and workspace json views subquery
• Fix incorrect metadata retrieval inside a workspace when there is a new
  change on another workspace (it should be irrelevant, but it was incorrectly
  used as new metadata)

0.5.0

• Added API key authentication security scheme.
• Fixed incorrect model update on file delete.
• Added a new query dialect, postgresql_json, for a Postgres SQL view where
  each family is represented as a JSONB column.

0.4.0

• Added k8s deployment through helm.
• Removed k8s nginx in favor of ingress.
• Added certificate management through certbot's k8s application.
• Added automatic database backups using k8s cronjobs.
• Minor security issue fixed on werkzeug dependency [CVE-2019-14806]

0.3.0

• Added GCP cluster auto-scaling.
• Added k8s horizontal scaler for web and worker.
• Added certbot for certificate management.
• Started rewrite and restructuring of documentation.
• Changed upload to systematically save into the workspace data directory.
• Changed commit to copy from workspace data directory to the data directory.
• Implemented file delete.

0.2.0

• Refactored app as a package in quetzal.app.
• Added a file storage backend, an alternative storage that does not use Google
  Cloud Platform.
• Added Let's encrypt SSL certificate management.
• Added deployment on GCP documentation.
• Added Sphinx documentation structure.
• Added file delete endpoint.
• Added file state in base metadata.
• Added a simple.naive implementation to file delete.
• Added a simple/naive workspace commit conflict detection.
• Added global queries.
• Added support for temporary files. They are uploaded on the users' bucket.
• Added path as query parameter when uploading files.
• Fixes unit tests that failed since introduction of roles.

0.1.0

First release of Quetzal API.

• Complete API specification with OASv3.
• Flask server implementation using zalando/connexion.
• Development environment with docker-compose.
• Deployment on Google Cloud Platform with kubernetes.