RemyMess
commented
3 years ago 1. Summary of the feature (v.2: improvement): General improvement of v.1: we want to improve the experience of the people watching the past talk recording. We want to capture the timestamp and allow people who view the past talks to see the slides moving at the same time.
2. Fundamental reason it is needed:
- Some talks use the slides and the recording only capture voice and faces if speaker dumps his slides. For such talks, the past recordings alone are dry.
3. Explicit tasks for bounty:
-
Frontend
-
[ ] Add a view for "past talk" (or make the overlay bigger)
-
[ ] If the streaming used slides, display two adjacent screens: one for the faces and the other one for the slides, each of them can be made big (see http://videolectures.net/deeplearning2017_sutton_td_learning/ for some ideas)
-
[ ] If talk is on youtube, iframe the recording
-
Backend
-
[ ] Capture timestamps of slidechanges during the stream
-
[ ] Capture when slides are on/off
1. Summary of the feature (v.1: core): We want to integrate the recording of seminars and upload in the pipeline for the organiser. This version just build the core foundation. Everything is kept as easy as possible.
2. Fundamental reason it is needed:
3. Explicit tasks for bounty:
Frontend
[ ] Record / pause / stop button for the administrator to allow to control the recording during the stream
[ ] Update to the audience that the seminar is being recorded "This seminar is being recorded and might be publically published." (imagine a sort of red button that lights up when its recording, turns grey when its paused or not recording)
[ ] When clicking on "Stop the seminar" (or at the end of it), ask the organiser "Your seminar recording is currently being compiled. Do you want us to automatically upload it once it is ready?)
[ ] Once the admin clicks on the "past talk card" (and assuming he did not choose to automatically upload the recording), add some text "Your recording is being processed and will be ready in a few minutes." or "Your recording is ready.". For the latter, the following options will be presented "View" or "Publish" (later on, for v2, we will allow them to edit the seminars and remove some sections if they want but we dont care about it for now)
Backend
[ ] agora.io API to record streams: here, we focus on admin + speaker only. (Later, for v.3 maybe, we will allow the recording of questions or people from the audience but it is far from a priority for now)
[ ] agora.io API to collect the streams and compile them into one file
[ ] Regarding the upload on agora.stream (i.e. when the organiser publishes the recording), we have a couple options:
[ ] Add an endpoint to delete the recording (Remy can do this if we want to do it on the flask)
[ ] Add an endpoint to publish the recording (Remy can do this if we want to do it on the flask)