Open JacobLeach opened 6 years ago
I'm confused. Why do videos have multiple files? Is that just if I have one video split in multiple parts? (I picture back-in-the-day of ripping DVDs, or how "part 1/part 2" used to be a thing on youtube).
With that in mind, is each one really a "video?" Can we make video the abstraction that is the culmination of all files that are part of that logical viewing?
Then a library contains videos, and a video contains files. I think that feels like a pretty clean abstraction.
Is there any benefit to a user only getting one of the files in a video? The only one I can think of is "I don't want to redownload this whole fucking thing and I only need to see the last minute of it" - we can solve that by having the user pass in a timestamp or offset or something to the API, and the server can determine which file(s) to send.
Did what I say make sense? I'm not sure I grokked what you were proposing so I responded to what I thought you were proposing.
I was thinking having multiple encodings of the same video one in low def and one in high def. I wanted people to be able to have both a 4K and a like 720p source file so they can watch one while local and the smaller one while on the road. Thus multiple video files per video and either is correct. I wanted libraries to hold files since a library is basically just a place to go find files (most likely a folder). You might split up your folders in HD and SD so we would need to know which folder had which so we could go stream it for the user.
Videos have files. Video libraries have files. Files have ids. It feels like you need to be able to GET /videos/:videoId/files as well as GET /video_libraries/:videoLibraryId/files but also GET /files/. Should we let people do GET /videos/:videoId/files/:fileId or should we restrict that endpoint (and the one of video libraries) to just get all?
CC: @travisby