Skip to content

Release cycles⚓︎

The Vidjil project tries to keep agile development practices while producing certified, qualified healthcare software for clinical needs. See also our roadmap.

  • We target continuous delivery on the public server app.vidjil.org. As of 2021,

    • the client is intended to be semi-automatically deployed through Gitlab,
    • the server is manually deployed,
    • the analysis engine vidjil-algo, which follows his own release cycles, is manually deployed. New releases with incremental changes may thus be deployed on-the-fly, but a notification is issued on the server. However, we announce the most significant features impacting data analysis with a 2-week notice.

  • The healthcare servers get a major release of all their components (client, server, analysis engine) twice a year, in June and December. Any deployment on the healthcare server should have been running on the public server for at least one month before. Such a release is further qualified through interactions with members of the VidjilNet consortium on their qualification datasets.

Critical bug fixes can be deployed on both public and healthcare servers at any time. To help the development move forward towards such stable healthcare releases, several freeze windows are enforced.

Summer release ☀️ Winter release ❄️
Meeting on release objectives mid-February early-September
Dev: Main Freeze
Any new feature merge must be discussed.		 30 March 30 September
Dev: Hard Freeze
No more feature merge, only bug fixes.
Polish strings, documentation, release notes, and announcements.		 15 April 15 October
Deploy on the public server
Continuous deployment on the public server is frozen, except for critical bug fixes.		 before 30 April before 30 October
Deploy stable release on healthcare servers and qualification
Continuous deployment on the public servers resumes.		 June December

For server maintainers, the latest stable release of the complete Vidjil platform is available on DockerHub. For bioinformaticians, the latest stable release of the vidjil-algo analysis engine is available directly on Vidjil's website.