I’m just happy that there is finally some movement on this issue. Whatever the first solution will look like in the end, I’m already really excited about it. And there will certainly be room for improvement, and even if it takes some time, I’m convinced that it will at least be a good first version in the end.
1 Like
Emby seems to simply choose the first episode in the list.
They were just a couple of random scenarios at the top of my head. We are only doing movies right now.
6 Likes
Thank you for sharing this. I would be glad as h*ll if this feature would expand into the classic movie libraries, too. Please…
If by that you mean the older agents, I’m sorry afaik it won’t. One of the big changes that needed to be made is how the GUID is generated in new agent so editions are the same but different. We are not updating old agents anymore with new features.
Sad news.
Could you please explain how the way PMS handles duplicates (ie. different editions of a movie) has something to do with the way how metadata is added? I am pretty sure that your new agent will not provide data about editions. other than interpreting local info from pathes or file names…
This is not the first time that I suggest, that adding the new guid also to older movie libraries (not as main index but as additional field) would be a way to handle things like that.
By offering (a) no way of transfering my metadata to new libraries and (b) offering no programmatic way of adding metadata to new libraries, you nail me down to the cross of your legacy agent… I would be willing to work with you on things like that in order to have a way of getting my data into the new libraries without overwriting my German age ratings, Covers, movie descriptions, several hundreds of collections including collection descriptions, Covers, sortings.
Sorry for this - as it is distraxting from the main thread… I really would be happy to use this feature, but I won’t be able to do so.
1 Like
I’m also curious how editions will be differentiated, especially for things like history and viewstate. I’m assuming the guid itself won’t change.
Will it be as simple as guid + (edition if present)?
Tracking viewstate for editions seems like a complicated data & logic problem - when an edition label is added to a library item, what will happen? Interesting!
What do you mean, and, What do you mean?
I’m curious if you’re describing something relevant to Editions, or if you just don’t like the new agents, or if there’s a genuine issue with the new agents.
I’m worried to ask and open a can of worms. Maybe better in another thread, instead of distracting here?
Any customized metadata should be preserved when updating, and I think the programmatic interfaces for interacting with metadata are the same. (??)
If you like the current data, I wonder if locking more data fields before upgrading would be helpful. That could be done en masse.
That would get messy really fast. Upgrading does a one-time, one-way migration between IDs. But maintaining a two-way mapping, when both sets of IDs can change, is data hell.
1 Like
I’m quite confused by this. When you upgrade a library to the new agent it transfers over. I have unfortunately noticed that there are some issues where some things do in fact get lost, but 95%, or at least 90% worked fine. I did, however, realize that if I were to do it again I would make sure to check all the fields as locked on anything I was concerned about before doing the transition. Plus a backup of course.
Do note that I have had problems after the import with things that are unmatched or have custom artwork loosing artwork when another copy of the same movie/episode is added.
Again this is quite confusing… The way you add metadata is the same in the new agent as it was in the old…
German age ratings: I have a lot of movies that I added custom ratings for. I use TV-10 & TV-12 specifically to add ratings in between TV-PG & TV-14 because some things are okay for them & others are not. The kids on my server’s mother & myself decide if it’s appropriate for all, some, just the oldest, or none. When I transitioned those ratings did transfer just fine. I have a Smart Collection in each library (Rated “X” so it’s only visible to me) that shows all movies/ Shows with No Rating or Unrated or NA or NR or that are blank called zTo Fix so I’m very conscious of ratings. I assure you, at least when I transitioned, which wasn’t when it was 1st available, they did keep.
Covers: This was the category that I had the most problems with. Most transferred over fine, but I did have a number that did not. Sadly Plex does not have a way to export covers for some reason, & where the covers are kept on disk seems to be designed specifically to make them hard to find, I have found a few “Solutions” to this, but none are really satisfactory unfortunately.
movie descriptions: I’m not sure with this, as I do not have many that I have customized
collections including collection descriptions: I know I noticed some issues but I know collections did transfer I’m fairly certain, but the descriptions may well not have.
sortings: With the new agent I have had issues multiple times with Sort Name getting lost, it annoys me as I keep all my collections sort-named " ` Collection Name" so they are at the top of the list. I don’t remember having a problem on the transfer, but I might just not have noticed it at the time, but I do have that problem randomly even for things I have added in the last month.
tl/dr: I’d suggest doing a backup & trying the upgrade library option. Most of the problems you noted are not there anymore or happen rarely
Well, this is not the main topic of this thread and I don’t want to hijack it…
I am using three specifically crafted python metadata agents based on other available sources than Plex does.
From what I fetch there, my agent does a cross check for genres (to harmonize and reduce them), my agent replaces foreign and streaming service age criteria with German “FSK age ratings” (again, to harmonize metadata). I am adding labels for prices and nominations for movies… It’s even grabbing my personal rating for any movie I rate on that web site.
Lots of my documentaries don’t auto-fetch Covers, so I downloaded or even created my own covers for them. I created hundreds of smart collections (750 of them one one of the two servers I am running) - each of them with my own descriptions and with my self-created covers. Almost everything in my very big libraries is German language since I always check metadata for every piece I add to the library. and correct in manually.
I also have a certain library for which the new agents don’t deliver any metadata at all.
All this has been years of work.
With the new agents, I will loose genre and age rating harmonization, I will loose all my covers (catastrophic no-go), I have no ideas if my labels will remain, but labels for new entries won’t be set…
To make a long story short… I will loose most of my work, I will use the way I am using Plex, I will loose data sanity between old movies and new additions.
There’s no way to add third-party agents to the the new agents… if that would be possible, I could try to work on my own data transferral code and even recode my own agents to work with the new solutions.
There’s not much sense in discussiing it.
There has been an interface for third-party-agents with the classic agents. Lost.
There has been a choice of possible sources for metadata with the classic agents. Lost.
There has been a way of including audio books, adult entertainment, comics and Youtube things, using not the main agent but special metadata grabbers. All Lost.
My use case has been almost dropped by Plex. I can still not update PMS and use what I have. This buys me a year or maybe two (till those automatically updated players are not accepting connections from old Plex servers)… and I really hope that Plex adds a new way of API for metadata agents until then - or offers a way of exporting a library including all metadata and let me re-import that on a new server (not using the new guid) - to enable guys like me to get a transition to the new and shiny world of Plex agents…
Again, sorry for hijacking this thread… I originally just asked for a new feature to be available on classic libraries, too.
2 Likes
Gotcha. I didn’t understand your use of additional agents. Thanks for the details and explanation.
It should be possible to preserve all of the current metadata in an upgrade, but that would still mean giving up the additional agents for the future. That doesn’t really sound like it addresses your use case.
I wonder if all of the metadata elements you need are exposed by python-plexapi. It might be possible to refactor from your legacy third-party agents to a different style of metadata enrichment.
Regarding Editions - do those apply to media items that can’t auto-fetch metadata? I’m guessing Editions are more likely to matter for certain popular content.
Just a suggestion: what about not calling them with a single synonymous word like version/edition, but specify the aspect each addresses, so like “technical version” and “edited version”? And of course, this is just for support documents and server managers. Regular users only has to see “720p” or “Director’s Cut” from their interface.
19 posts were split to a new topic: Discussion of Multiple Editions
