@daeks , per past few posts in this topic, I was saying 2-3 years before capability gets removed entirely (from Plex server.) in response to someone saying ‘several’.
All clients are going first, and one of the first of those started last December 2018.
@elan Will it be possible to continue having plugins in a more diminished capacity? Basically giving us the option to have it available if we choose to, similar to how it’s an optional section in the ios implementations of the Plex app?
Removing the entire plugin management from the front-end of the web/PMS interface is fine, we can still always manually add bundles. We can use management plugins to manage them.
Plex openly being not responsible for the working of plugins will also be fine. We’d understand given you’re moving away from maintaining it.
The plugins tab on the new sidebar can be unticked by default so newer users will not need to know about it.
Making it a PlexPass feature.
This way those are inclined to have it may have it. I don’t think taking it out of the front-end will affect most plugin users since we’re probably old time users of Plex and are comfortable with work-around for the most part. Making if available client side even in the redesign till a new API or some replacement can be produced, is what people want. This can keep people who depend on them happy and earn lots of goodwill from those of us who’ve been around from the start, of what has been an epic journey.
With above, about 95% of all plugins would be gone
And with all of that, the framework would still be outdated, and would req. at least as a bare min. 6 month of dedicated work from a backend dev to update it
this should have been thought about and planned for years ago.
from the start, plugin system should have been completely open and well documented, easily updatable/upgradable, and easily maintainable and sustainable.
all software has a lifespan and needs constant care and updating, if for nothing else security fixes.
plugin framework should be script/language neutral (API based) allowing for just about anyone to implement using whatever
if plex doesn’t want to do the work to move forward with python, then they should at a minimum provide/document the necessary API/hooks in a generic manner for other devs to be able to (re)implement plugins that can leverage the rest of the plex architecture/framework.
Kinda not possible, when the whole engine behind it is dying, as I linked to in my post, and new version is not backwards compatible!
I lied before, since Plex did update the Python engine once, due to a security issue, but that was from Python 2.7.x → 2.7.z, and as such, backwards compatible, so shame on me for not been clear about that in my former post, sorry
It was, and docs was avail, end now Plex wants 3.Party devs to move towards their rest api, using stand alone apps, as stated in their blog post
Above though for now is more or less only for utils, since provider API info is not public. No idea if that’s changing in a near future, since not Plex
plugin python version should have been tracked along with base python version and older versions deprecated as well.
plugins that did not receive the love and attention would have fallen away years ago and the ones that were well used would have been kept updated if not by the original dev, then by others who would either fork or reimplement as necessary.
the same should have been done with the old C library and any other legacy libraries that have been kept around for backwards compatibility.
the mess plex is in is of its own doing by not deprecating old stuff and embracing the latest/current stable versions of necessary tools and libraries.
if that means that people who run 10 year old servers and don’t want to update/upgrade, then they can keep running the last working version of plex.
it would be nice if plex would host a developer wiki themselves instead of people having to reverse engineer and document outside of their control.
True, but, and speaking as a private person here, but since Framework wasn’t updated for a long time, it’s kinda clear, that plugin’s wasn’t their main focus
Partial true, but guess you are solely focused on streaming plugins now, since tools based plugins would hum along all the way
WOW…That would have meant people screaming over and over again for years!
Plex was nice here, and provided backwards compability for as long as they could, IMHO
And with above…You have no problem cutting off folks running old HW, yet you scream when you are hit?
(And when said, granted that folks been cut-off due to old C lib, could invest, but with plugins gone, you have no way to invest your way out of this, besides change the way you watch medias from the web to use native methodes provided by media providers)
Look above
Not so, since PMS and clients are more or less connected, and newer clients req. newer PMS!
not focusing any particular plugin type, plex did not seem to distinguish when they said plugins are going away and I don’t know how streaming plugins would be any different than any other plugins when it came to needing updated for newer python or other libraries.
re: client vs server versions, this is due to failing to make client/server communication more version agnostic, along with many clients being updated outside the user control (so they can’t keep using older clients).
one way would have been to create new apps for new versions of clients that break server compatiblity, or re-release older versions on a different app (like roku plex retro) and make them unsupported but still usable.
(ideally) within same version api, clients and servers should be able to communicate across all minor sub-versions.
like how well implemented client/server api are versioned
re backwards compatibility, I am not saying it is not a tricky or sensitive subject, but it could have been much more smoothly done, deprecating slowly as you go versus hitting a wall at 100mph, and people would have at least been used to it if it had been done that way from the start.
I’ve said this before, in some other thread I now don’t remember and can’t quote/link to. So, I’ll say it again.
If Plex hadn’t waited so many years to update, there may not be such a problem. You Dane22 even left another thread I stated this in because you thought I was trolling when I mentioned knowing of 2 channel developers who stopped coming to the forums because Plex company didn’t care to fix plug-in API bugs they brought up clear issues with, in a constructive manner no less.
Basically, Plex placed pictures of a great beach house in the sales brochure, and allowed the house to break down, become infested with palmetto bugs, and the yard to overgrow. You can’t even reach the beach hardly anymore, let alone see it.
So, the owners of the house (Plex) realize that they let the house break down and get infested with those palmetto bugs. So, they have cleared the yard so that you can see it (new UI) but the house it self, as you described it, is broken down and infested with bugs.
They decide to take it down to the foundation and rebuild it. It will still be a house with rooms and a kitchen and bathrooms and such but the “always open back porch” is now going to be closed in so that the palmetto bugs don’t get back in.
In the end, a newer, more desirable, beach house will be there. Granted, they will have to redo the brochure and your favorite open porch is gone, but it will last a lot longer.
A new UI does not allow me to see or access the beach.
The oceanfront was taken away, and the house was moved 3 blocks inland and larger condos placed on the beach that now even prevent ocean view.
Many, including myself, purchased the property for the advertised oceanfront feature.
It still mentions, perhaps, oceanview, but it’s now extremely vague, and we’ve now payed so many maintenance fees we’ll never get our money back.