Jump to content

Recommended Posts

Posted
10 minutes ago, Cheesegeezer said:

So I was looking at the TV Show the boys and Amazon Originals call there IntroStart to IntroEnd just "Title Sequence" is this a route we should take? And be done with the 2 flags?

@rbjtech @chef @samuelqwe

 

image.thumb.png.5c3bb5db0ba4b74f35ace3ed9bf098a4.png

This is only my opinion, but like Luke said, chapters don’t seem like a great way to store this information. A database for this specific purpose seems much better, because chapters could be laid out in so many different ways and be named differently, so we likely don’t want to mess too much with them. As an intermediary step, it’s probably fine, but it shouldn’t be the final solution.

  • Agree 1
Cheesegeezer
Posted
2 minutes ago, samuelqwe said:

This is only my opinion, but like Luke said, chapters don’t seem like a great way to store this information. A database for this specific purpose seems much better, because chapters could be laid out in so many different ways and be named differently, so we likely don’t want to mess too much with them. As an intermediary step, it’s probably fine, but it shouldn’t be the final solution.

yep ok.  Really could do with some input from the admins now.  

I've checked my library and i have to say that the detection is great and certainly good enough to use with a high rate of success.

I'm looking at swagger and there is something there, however nothing to post actual results

image.png.203643b4f39001a3a72238c1a577ea72.png

 

Returning Get/EpisodeTitleSequence actually returns a lot of the fingerprint info.

image.png.7d35d27d8aecc272c518bf3950df1b29.png

Posted
6 hours ago, Cheesegeezer said:

Not important right now.. more of a nice to have.

Agree lets move along.

i have some issues with my Westworld episodes and I’m pretty sure it’s down to the encoding that causes the problems and not the plugins detection system.

Are those episodes where the intro is found further along in the stream then 20 mins?

Posted
1 hour ago, Cheesegeezer said:

yep ok.  Really could do with some input from the admins now.  

I've checked my library and i have to say that the detection is great and certainly good enough to use with a high rate of success.

I'm looking at swagger and there is something there, however nothing to post actual results

image.png.203643b4f39001a3a72238c1a577ea72.png

 

Returning Get/EpisodeTitleSequence actually returns a lot of the fingerprint info.

 

Yup, we need to ignore fingerprint data in that endpoint for sure.

Good catch!

Cheesegeezer
Posted
28 minutes ago, chef said:

Are those episodes where the intro is found further along in the stream then 20 mins?

nope, but i'm fairly confident it's down to audio encoding and nothing to do with the plugin.

I've used it on a high power machine and a low power machine.

image.png.9a848be081fa789002a92ff6394e640c.png

Cheesegeezer
Posted

@chef  what's this format? for stand and end...... PT 2min 29s for start time and PT 3min 10sec for end, what does the PT refer to.

image.png.906c6249e9a5625ea2181d34b2bfeac9.png

Posted
4 minutes ago, Cheesegeezer said:

@chef  what's this format? for stand and end...... PT 2min 29s for start time and PT 3min 10sec for end, what does the PT refer to.

image.png.906c6249e9a5625ea2181d34b2bfeac9.png

I believe that’s the duration in ISO 8601 format.

I believe P stands for "period" and T would just be "time", since ISO 8601 also supports dates prefixed before the T instead of the P.

Fairly standard way of storing date & time.

  • Thanks 1
Posted
3 hours ago, samuelqwe said:

This is only my opinion, but like Luke said, chapters don’t seem like a great way to store this information. A database for this specific purpose seems much better, because chapters could be laid out in so many different ways and be named differently, so we likely don’t want to mess too much with them. As an intermediary step, it’s probably fine, but it shouldn’t be the final solution.

It never was planned to be, I just think people are too soon to judge what we are trying to do here. 

The introduction of adding them to chapters was to actually get some use out of the Intro database while the 'powers that be' decide what to do with the data. 

Being pessimistic - they may want to see overall changes in the way this is achieved, they review in 2 years time and we might see it in 2023.  (Remember, this thread started in 2017..)

With my optimistic hat on, it may get reviewed in the next couple of weeks, they agree a mechanism for storing the Intro points in the Core DB and add the 'IntroSkip' to all the clients by the end of the month ..  If this were the case, then we probably could have missed the chapter step altogether.  But lets be realistic here.

Personally, I can use the chapters TODAY while viewing (and more importantly TESTING) - the same way people use the existing ones like @Cheesegeezer mentioned above.  They are in there because they are useful and let you er.. skip the Intro.

Next well be asking for 'Previous/Next Chapter' in the UI - you'know the function in every other media player for the last 20 years ... 🤪

@Cheesegeezer

As it appears we are not planning to use IntroStart/IntroEnd to do the 'Skip Intro' button - then I think there are a couple of things we should probably change on the 'add chapters option' to make it more inline with it's original intention of simply providing a jump to the after the Intro.

a) We do not need to add 'Intro Start' -  as nobody in their right mind is going to skip and want to watch the Intro ;)

b) If there is an existing chapter point within 5 seconds (?) of a planned IntroEnd point - then don't add IntroEnd - as it's highly likely an existing IntroEnd Chapter point anyway.  This will also cater for naming of other Intro's - for example 'Title Sequence' as Amazon call it.

What to you guys think ?

Posted (edited)
3 hours ago, Cheesegeezer said:

@chef  what's this format? for stand and end...... PT 2min 29s for start time and PT 3min 10sec for end, what does the PT refer to.

image.png.906c6249e9a5625ea2181d34b2bfeac9.png

Duration format. 

ISO8601

Edit: Sam answered... Too much puree... 😟

Edited by chef
Posted
30 minutes ago, rbjtech said:

It never was planned to be, I just think people are too soon to judge what we are trying to do here. 

The introduction of adding them to chapters was to actually get some use out of the Intro database while the 'powers that be' decide what to do with the data. 

Being pessimistic - they may want to see overall changes in the way this is achieved, they review in 2 years time and we might see it in 2023.  (Remember, this thread started in 2017..)

With my optimistic hat on, it may get reviewed in the next couple of weeks, they agree a mechanism for storing the Intro points in the Core DB and add the 'IntroSkip' to all the clients by the end of the month ..  If this were the case, then we probably could have missed the chapter step altogether.  But lets be realistic here.

Personally, I can use the chapters TODAY while viewing (and more importantly TESTING) - the same way people use the existing ones like @Cheesegeezer mentioned above.  They are in there because they are useful and let you er.. skip the Intro.

Next well be asking for 'Previous/Next Chapter' in the UI - you'know the function in every other media player for the last 20 years ... 🤪

@Cheesegeezer

As it appears we are not planning to use IntroStart/IntroEnd to do the 'Skip Intro' button - then I think there are a couple of things we should probably change on the 'add chapters option' to make it more inline with it's original intention of simply providing a jump to the after the Intro.

a) We do not need to add 'Intro Start' -  as nobody in their right mind is going to skip and want to watch the Intro ;)

b) If there is an existing chapter point within 5 seconds (?) of a planned IntroEnd point - then don't add IntroEnd - as it's highly likely an existing IntroEnd Chapter point anyway.  This will also cater for naming of other Intro's - for example 'Title Sequence' as Amazon call it.

What to you guys think ?

Intro start is a time to show the button, if there was to be a button.

Don't show a skip button before the intro during recaps, or storyline that happens prior to the title sequence.

It will give the client a time frame to know when to show the button.

Posted
1 hour ago, chef said:

Intro start is a time to show the button, if there was to be a button.

Don't show a skip button before the intro during recaps, or storyline that happens prior to the title sequence.

It will give the client a time frame to know when to show the button.

The IntroStart time for the button (and IntroEnd) is being got from the Intro dB - not the chapter dB.  That is what Luke is alluding to. 

If emby gets it from the IntrodB (or copies it into a Core dB), then there is no need for an IntroStart chapter.  IntroEnd chapter is required only when there is not one already - and even then, it should be optional.

  • Like 1
  • Thanks 1
Posted
1 minute ago, rbjtech said:

The IntroStart time for the button (and IntroEnd) is being got from the Intro dB - not the chapter dB.  That is what Luke is alluding to. 

If emby gets it from the IntrodB (or copies it into a Core dB), then there is no need for an IntroStart chapter.  IntroEnd chapter is required only when there is not one already - and even then, it should be optional.

Ah, I understand what you mean. I misunderstood. Cool.

Posted (edited)

I fixed the response from the API, so it doesn't  return the fingerprint.

I created BaseTitleSequence, which is what the API will respond with.

It is pretty much the same only with out the Fingerprint, or the encoding duration.

I can make an endpoint that responds with the entire object (like it does now) if that is helpful.

Edited by chef
  • Like 2
Posted (edited)

Cheese! I pushed the update.

Check out what I did for Chapters Scheduled Task. I used the new BaseTitleSequence object. It should maybe make things fast not passing around all the fingerprint data.

 

Also check out the Triggers for the task. Now the task will fire when the User presses the task button, but won't make any assumptions to fire on it's own.

This way the user can change their task settings, and it will allow them to control when they want it all to sort their chapters.

The TitleSequence Chapter looks so cool! i love how it grabs the screen from the intro! 

Also, the API will not return the fingerprint object, or the encoding duration in the BaseTitleSequence object, so the API is much better now, and much much faster.

 

The UpdateTitleSequence in the API is a 'GET'. I know it should be a 'POST'. I can fix that.

Also, there POST to scan series can be removed. I will do that before work today.

 

EDIT: on second thought the 'ScanSeries' 'POST' route is hooked up and should work, although there is no button on the configuration to handle it. We might as well leave it, incase we want something in the future.

 

EDIT AGAIN: I pushed to early... I have to fix some things. sorry.

 

 

...FIxed

Edited by chef
  • Like 1
Cheesegeezer
Posted

Thanks very much chef,

Ok so i've sorted out the chapters now.....  Only title sequence is inserted and the next chapter starts in order.

I'll need someone to test this on a small library where they actually have the chapter names.

image.thumb.png.70aee329ca6e17b8c447a49f0ec62bea.png

Cheesegeezer
Posted

Oh and I added in the user options everyone is banging on about

image.png.f97e97cf446bbfc0ada11344be83dba6.png

  • Like 1
Posted
56 minutes ago, Cheesegeezer said:

Oh and I added in the user options everyone is banging on about

image.png.f97e97cf446bbfc0ada11344be83dba6.png

Okay, I've gotta pull that config update.

Cheesegeezer
Posted (edited)
2 minutes ago, chef said:

Okay, I've gotta pull that config update.

LOL, yeah i just noticed that it wasn't in hahaha.  

Lemme know when it's done and you have pushed back to Git and i'll add the push the new chapter code.

 

EDIT: I also notice that the remove all doesn't function at all.  Is this intended?

Edited by Cheesegeezer
Posted
54 minutes ago, Cheesegeezer said:

LOL, yeah i just noticed that it wasn't in hahaha.  

Lemme know when it's done and you have pushed back to Git and i'll add the push the new chapter code.

I think I might be the worst Github worker in the world. 

Looks like I have a learning curve here.

The config is updated, but I don't think I did it the easy way at all. Sorry Dave. 😬

  • Haha 1
Posted (edited)
8 minutes ago, Cheesegeezer said:

LOL, yeah i just noticed that it wasn't in hahaha.  

Lemme know when it's done and you have pushed back to Git and i'll add the push the new chapter code.

 

EDIT: I also notice that the remove all doesn't function at all.  Is this intended?

Remove all use to function. But I hadn't wires it up to the new database yet.

There is a method in the repo interface to handle it, but I don't think the button actually hits an endpoint to trigger 'RemoveAll' method... Yet.

 

Edited by chef
Cheesegeezer
Posted

I'll have to add in lot of my code again lol..... there is quite a bit missing.  Wont take long.  I'll push the commit and we can go from there. 👍

Posted (edited)
1 hour ago, Cheesegeezer said:

I'll have to add in lot of my code again lol..... there is quite a bit missing.  Wont take long.  I'll push the commit and we can go from there. 👍

Yup! the worst Github guy... That's me.

Let me know when I can do... um... a 'fetch'? or a 'pull' 🥴😆

 

EDIT: I think we are ok now. I think it's sync'd with the proper code in the config.

Edited by chef
  • Haha 1
Posted (edited)

@Cheesegeezer I get this in the logs, and emby shuts down.

2021-09-10 15:14:34.518 Error App: Error disposing ChapterManager
	*** Error Report ***
	Version: 4.6.4.0
	Command line: C:\Users\Benjamin\AppData\Roaming\Emby-Server\system\EmbyServer.dll -noautorunwebapp
	Operating system: Microsoft Windows 6.3.9600
	Framework: .NET Core 3.1.13
	OS/Process: x64/x64
	Runtime: C:/Users/Benjamin/AppData/Roaming/Emby-Server/system/System.Private.CoreLib.dll
	Processor count: 4
	Data path: C:\Users\Benjamin\AppData\Roaming\Emby-Server\programdata
	Application path: C:\Users\Benjamin\AppData\Roaming\Emby-Server\system
	System.NotImplementedException: System.NotImplementedException: The method or operation is not implemented.
	   at IntroSkip.Chapters.ChapterManager.Dispose()
	   at Emby.Server.Implementations.ApplicationHost.Dispose(Boolean dispose)
	Source: IntroSkip
	TargetSite: Void Dispose()

Oh wait i see what it is... nevermind.

Edited by chef
Cheesegeezer
Posted

@chef I just pushed a commit, however i had to pull one before I could push.  So I will restart VS and open project, build the solution and test.

We might be ready to throw out another beta if you're up for it?

Not really sure what else we can do at this point to get this moving on.

  • Like 1
Posted

Sopunds good. My emby instance is crashing for some reason. 

I need to see why... hold tight!

Guest
This topic is now closed to further replies.
×
×
  • Create New...