1.1.0

Added

• Prepare the app to handle timedtexttracks (full functionality yet to come)
• Add an API schema view
• Extend timed text track language choices to all Django languages
• Create a JWT token for every roles not just instructors (needed to add
  context to xAPI requests)
• Allow timed text track list GET requests on the API (permissions are fixed
  in a separate commit)
• Switch to redux for frontend state management
• Add a link to the dashboard from the video form
• Split Dashboard video matters into a separate component
• Run yarn install in docker node to use node 8

Fixed

• Minor fixes and improvements on features and tests

1.0.0

• Improve settings and their documentation (notably environment variables);
• Add automatic iframe resizing to improve content layout;
• Improve LTI verification and add an origin check;
• Implement an instructor view to allow instructors to change the video on a given resource;
• Refactor "subtitles" into "timed text" to add transcripts in the backend & lambdas;
• Improve admin views for consumer sites & passports.

1.0.0-beta.5

• Add LTIPassport to the Django admin interface
• Logging can now be configured via a environment variable. The string is evaluated using ast.literal_eval and can thus ba a Python dictionary format.

1.0.0-beta.4

Add a Dashboard view, replacing the Upload loader, that shows the video's status through upload and processing. It also allows the instructor to watch the video when it is ready or to upload another one to replace it.

1.0.0-beta.3

This release fixes Marsha to work with Open edX (CMS does not send its hostname and advanced settings don't accept all punctuation characters).

1.0.0-beta.2

Add missing front-end build in the Docker production image.

1.0.0-beta.1

THIS IS A BETA RELEASE

Marsha is a web service meant to host and stream videos in an e-learning context.

It runs as a third party through LTI (or other integrations such as an API) and allows teachers & administrators to upload and manage videos, playlists and subtitles.

Those can then be streamed into any LMS (or in the future in other contexts) using Marsha's player.

Content:

This first release supports the simplest version of the complete workflow.

• Create a consumer site (the first party that will connect to marsha through LTI);
• Add LTI consumers pointing to marsha's lti-video view;
• Upload videos for automated transcoding & further processing;
• Stream those videos through the same LTI consumer integration, with or without adaptive bitrate.

Known issues:

• The LTI view errors out after the upload is complete and until the video is completely processed. It should instead display some helpful information to the user.