I’m thinking of creating a metadata provider for Jellyfin (or maybe Emby) which would source metadata from Trakt rather than from underlying sources like TMDB, TVDB, etc… I really like the curation that Trakt does to get things right, and I already use Trakt lists to manage what gets added to my collection so it makes sense to source the metadata from the same curated source…
I am curious though whether anyone knows/thinks this might violate some policies/etc… I think I would have to restrict it to VIP access only as metadata API is kinda nerfed without VIP, but that seems like a reasonable requirement to me…
The Trakt API doesn’t provide images, so that is one limitation. We get info from TMDB/TVDB so those scrapers are usually ok on their own, but Trakt does have some custom things I can see being useful. My main concern is how much traffic this will send to the API. If it’s metadata only that gets cached at Cloudflare so might not make much of an impact, but I can’t say for sure.
Yeah, I saw it doesn’t contain the images themselves, but it does contain pointers to those images (from what I can tell in the API docs, so that would be a secondary call for sure)…
Regarding edge caching and API limits, that’s pretty much why I wanted to ask here first… Definitely don’t want to break anything.
I guess I’ll start the build process and we’ll see what happens.
We don’t include image pointers either in the API. We do in a few specific places like user avatars and cover images, so maybe that is what you saw in the docs?
Hmm, yeah, just noticed I was conflating the IDs to other sites (like TMDB, IMDB, etc) with the id’s to the images themselves on those sites… guess the images piece would take a bit more “doing”, but certainly this is a great start.