2024-08-27, 10:01 PM
Continued post from JF 10.9.10 release thread.
I don't believe what I've suggested for Collections (or possibly playlists as well) should be fundamentally dependent on the full file path as these operations are possible in other database programs without using such.
For example, in Kodi you can create smart playlists where the variables can include multiple types of calls from the main database to return results. Variables include full path but also title, sort title, tags, names, studios, ratings and so on. This is because the smart playlists don't have to look for the exact entry path but instead checks other areas of the database. So from within Kodi I have established my own Categories that have those such variables. For the MCU I simply have a tag in the media metadata NFO for Marvel Cinematic Universe and then the smart playlist for MCU just pulls in all movies with that tag. It's easy and I've built a huge system around that. Other entries look for a specific variable for Director, so say Steven Spielberg, and returns all movies where he is credited in the director metadata.
Now Kodi doesn't build smart playlists, to my knowledge anyways, that incorporates different types of media the way JF does. That is why I like the JF Collections and have built close to 300 all manually. If JF was to handle Collections in a similar manner and be able to add items by library and then title/sorttitle, the Collections would auto populate even if the actual file name path was to change for some reason. So for MCU once you've added Iron Man from your Movies library going forward as long as you have Iron Man in your Movies library it will be pulled to your MCU Collection even if you move the actual file/folder to a different location, rename it, change the version etc. When I upgrade a movie to Atmos audio I don't have to do anything in Kodi to keep that entry in the smart playlist because the required title or tag or director doesn't change for the new library entry. Being able to use sorttitle allows you to keep out titles with the same name that don't fit, say there's another movie called Iron Man staring Dolph Lundgren as a crime fighter that bashes baddies with a clothes iron and came out in 1987.
In the JF database all the information is already* (almost) scraped and held. You can even search by title, names, tags etc. It's just allowing use of that information to populate entries into Collections which then can streamline some collections while also making any library changes not mess up Collection entries. Nothing changes on the behaviour of how JF brings in metadata to the DB. Only how it is accessed after.
I could build a Collection for Academy Award Winners with sub collections for each award category if JF Collections could be built on tags as I've curated all that information into my library metadata NFOs. I'm not going to spend hours and hours going through all of that manually though. But all that data pulls into such categories and collected entries in Kodi (which JF happily pushes forward and has made awards season so incredibly easier to maintain for me now).
What I love about the JF collections is, using the MCU example, I can have all the movies, all the TV shows, all the soundtrack albums, all the theatrical posters and then anything else associated within my JF server collected together in one place. That is something Kodi can't do and I don't use Kodi for music or photos or anything outside movies, tv shows and videos/trailers. I've made hundreds of JF collections for director's and put together their libraries with soundtracks to their films and that's helped me organize and fill holes in my collections.
It's been very useful overall but I believe there is a better ay of JF handling the Collection entry data than just relying on full file and path name. And now seemingly not using the created XML files which further happens the backend organization ability of the end user, especially if an error occurs in the JF database itself which has happened and caused my to rollback JF versions to get things fixed without spending days re-doing everything manually.
If within collections the entries can be separated into their JF library entries so Movies and Movie Trailers show up as such and not just lumped together under videos, that would make Collections even more awesome.
* I said almost because JF doesn't seem to scrape IMDB Top 250 metadata even when it is present in the NFO files. Unsure why this isn't included but hopefully it will be at some point as I like seeing those rankings, even if I don't agree with many of them.
I don't believe what I've suggested for Collections (or possibly playlists as well) should be fundamentally dependent on the full file path as these operations are possible in other database programs without using such.
For example, in Kodi you can create smart playlists where the variables can include multiple types of calls from the main database to return results. Variables include full path but also title, sort title, tags, names, studios, ratings and so on. This is because the smart playlists don't have to look for the exact entry path but instead checks other areas of the database. So from within Kodi I have established my own Categories that have those such variables. For the MCU I simply have a tag in the media metadata NFO for Marvel Cinematic Universe and then the smart playlist for MCU just pulls in all movies with that tag. It's easy and I've built a huge system around that. Other entries look for a specific variable for Director, so say Steven Spielberg, and returns all movies where he is credited in the director metadata.
Now Kodi doesn't build smart playlists, to my knowledge anyways, that incorporates different types of media the way JF does. That is why I like the JF Collections and have built close to 300 all manually. If JF was to handle Collections in a similar manner and be able to add items by library and then title/sorttitle, the Collections would auto populate even if the actual file name path was to change for some reason. So for MCU once you've added Iron Man from your Movies library going forward as long as you have Iron Man in your Movies library it will be pulled to your MCU Collection even if you move the actual file/folder to a different location, rename it, change the version etc. When I upgrade a movie to Atmos audio I don't have to do anything in Kodi to keep that entry in the smart playlist because the required title or tag or director doesn't change for the new library entry. Being able to use sorttitle allows you to keep out titles with the same name that don't fit, say there's another movie called Iron Man staring Dolph Lundgren as a crime fighter that bashes baddies with a clothes iron and came out in 1987.
In the JF database all the information is already* (almost) scraped and held. You can even search by title, names, tags etc. It's just allowing use of that information to populate entries into Collections which then can streamline some collections while also making any library changes not mess up Collection entries. Nothing changes on the behaviour of how JF brings in metadata to the DB. Only how it is accessed after.
I could build a Collection for Academy Award Winners with sub collections for each award category if JF Collections could be built on tags as I've curated all that information into my library metadata NFOs. I'm not going to spend hours and hours going through all of that manually though. But all that data pulls into such categories and collected entries in Kodi (which JF happily pushes forward and has made awards season so incredibly easier to maintain for me now).
What I love about the JF collections is, using the MCU example, I can have all the movies, all the TV shows, all the soundtrack albums, all the theatrical posters and then anything else associated within my JF server collected together in one place. That is something Kodi can't do and I don't use Kodi for music or photos or anything outside movies, tv shows and videos/trailers. I've made hundreds of JF collections for director's and put together their libraries with soundtracks to their films and that's helped me organize and fill holes in my collections.
It's been very useful overall but I believe there is a better ay of JF handling the Collection entry data than just relying on full file and path name. And now seemingly not using the created XML files which further happens the backend organization ability of the end user, especially if an error occurs in the JF database itself which has happened and caused my to rollback JF versions to get things fixed without spending days re-doing everything manually.
If within collections the entries can be separated into their JF library entries so Movies and Movie Trailers show up as such and not just lumped together under videos, that would make Collections even more awesome.
* I said almost because JF doesn't seem to scrape IMDB Top 250 metadata even when it is present in the NFO files. Unsure why this isn't included but hopefully it will be at some point as I like seeing those rankings, even if I don't agree with many of them.
JellyFin Wish List:
IMDb Top250 metadata
Collection content rules - Library-Title/Sorttitle, Tag, Director, Filename/Path Contains
Collection organized by library
Collections scanned to editable XML
Media info show Collections added to
Soundtracks auto link to movie by title/sort + Manual
IMDb Top250 metadata
Collection content rules - Library-Title/Sorttitle, Tag, Director, Filename/Path Contains
Collection organized by library
Collections scanned to editable XML
Media info show Collections added to
Soundtracks auto link to movie by title/sort + Manual