117 replies to this topic

[littlesat]

Here we go again. The problem is clear we need to store somehow the prefered audio channel and subtitles from web streams in enigma2. But without any discussion, agreement, design and agreement how to do this the best clean lean and good suddenly there is existing code…. Now it is somehow take it or leave it…. But the outcome is a giant code change that feels like a huge work a round (except for the author or it)…. So now it is take it or leave it. Better next time first discuss the problem and think together what could be the best solution before creating something…..

[DimitarCC]

Here we go again. OK discuss it longer but who will implement it? Do you have someone that willing to do that extensions in enigma2? Even if you find really the greatest, cleanest solution if there is no one that willing to implement it everything is pointless in my opinion.

 

And the main problem is not that i or somebody else is disappointed by single denial. 

Here everything got denied, rejected or delayed for ages.

 

When working in a team is clear not every your suggestion will be accepted but when almost all your suggestions got rejected then something is wrong and that developer got feeling like he is not part of the team. He have no purpose there.

 

@WanWizard That strategy that code owners decides who to live and who to die is more like a dictatorship and its definitely not the meaning of the opensource and the idea of contributing to it.

[DimitarCC]

Now it is somehow take it or leave it…. 

 

Umm no. Now is more like "Do it in our way or not do it at all"

 

There is not actually discussion here. All i hear is how the code is not right, how it slows down things (without even to be tested) and nobody offers improvements or direction to where to go with the code. 

Nor anyone point to the actual places in the code where improvements have to be done.

You dont accept my code because it is not in the way you want it and seems dont want to accept any of my arguments why it should be implemented.

 

as i said in other post there have to be compromises from both sides. Otherwise there can not be collaboration at all.

[Dimitrij]

My opinion...

This is a normal Python solution.

First, it needs to be checked and some adjustments(simplify the code) can be made if necessary.

 

 

 

DimitarCC

There is an alternative, if the patch is not accepted, add this code to the plugin(e.g "Automatic Volume Adjustment").

I've had to use workarounds like this before.

Edited by Dimitrij, 5 October 2023 - 05:22.

[DimitarCC]

I do not force anyone to use this. If it is not accepted ok. Users can install this manually if they want. I dont want to put this in auto volume adjustment plugin since that plugin do stuff i dont like. But i wanted to avoid the hack that plugin will do so to hook to audio window....
Anyway lets see what will happen.

[littlesat]

Indeed good suggestion. Made a plug-in from it (when possible)…

[DimitarCC]

That plugin have to "steal" functions from AudioSelection screen however. That i wanted to avoid because it can cause issues once the AudioSelection screen got changes in base code.

It is required when somebody select a track in audioselection that selection to be stored. That means i have to redefine keyOK function of AudioSelection screen in my plugin

Once you change the behavior of keyOK in base code the plugin will stop to work or will cause issues. 

Edited by DimitarCC, 5 October 2023 - 08:01.

[Dimitrij]

That plugin have to "steal" functions from AudioSelection screen however. That i wanted to avoid because it can cause issues once the AudioSelection screen got changes in base code.

It is required when somebody select a track in audioselection that selection to be stored. That means i have to redefine keyOK function of AudioSelection screen in my plugin

Once you change the behavior of keyOK in base code the plugin will stop to work or will cause issues. 

Simple.

Call plugin in AudioSelection menu --> open plugin("Assign audio/subtitle track") same as keyOK function of AudioSelection screen

[DimitarCC]

 

That plugin have to "steal" functions from AudioSelection screen however. That i wanted to avoid because it can cause issues once the AudioSelection screen got changes in base code.

It is required when somebody select a track in audioselection that selection to be stored. That means i have to redefine keyOK function of AudioSelection screen in my plugin

Once you change the behavior of keyOK in base code the plugin will stop to work or will cause issues. 

Simple.

Call plugin in AudioSelection menu --> open plugin("Assign audio/subtitle track") same as keyOK function of AudioSelection screen

 

 

I dont really get it.... So on OK press in AudioSelection Screen there should be custom logic to take the selected track and current service and store them in the file. Also currently payed track if differs from the stored value then the value have to be updated.

How all that will be fixed by calling open plugin("....")?

And where to call that? On selectionChange in AudioSelection list? Or where?

 

Maybe i dont understand your idea....

[Dimitrij]

https://github.com/O...lection.py#L175 -->open plugin("Assign audio/subtitle track")

[DimitarCC]

Hmm an how that will be integrated without to extend AudioSelection.py? 

Isn't it that the main reason to be a separate plugin? To not extend PLi base files?

If i have to extend AudioSelection.py anyway then why not integrate all the code in it?

 

If you think to use it as a separate AUDIOMENU plugin...then to store the selected audio track for example have to call that menu separately just to store it... which is opposite to the whole idea.

Edited by DimitarCC, 5 October 2023 - 13:25.

[Huevos]

https://github.com/O...lection.py#L175 -->open plugin("Assign audio/subtitle track")

So are you saying that instead of controlling audio track selection from LanguageSelection.py, we should create a plugin that hijacks enigma core functions? So a super hack to avoid a hack? Or am I not understanding you properly?

[Dimitrij]

 

Or am I not understanding you properly?

This is just a discussion.

[Huevos]

This thread gets more and more bizarre.

 

Selecting and remembering an audio track should be a core function. You shouldn't need a plugin to do it. Current code is broken because you can select a track but not save it. The patch fixes that.

 

OpenPLi team would prefer its users to have a broken image rather than take this patch. And no one from PLi is going to fix it because as WanWizard says "there is not the will".

Edited by Huevos, 5 October 2023 - 13:36.

[luisteraar]

Default answer from pli is "its in C code we can do nothing with it' 

missing knowledge.

In lamedb you can add a field url for iptv streams.

sql command.

The field url is then the reference for iptv streams.

Edited by luisteraar, 5 October 2023 - 13:52.

[DimitarCC]

 

In lamedb you can add a field url for iptv streams.

sql command.

The field url is then the reference for iptv streams.

 

And what will do if the url is dynamic like in some of IPTV providers?

You can not use the url for identifier. It must have fixed service reference so to uniquely identify the channel.

 

The problem is not the knowlege... The problem is that messing with lamedb will break everything that works with it now (channel editors, plugins, etc.)

 

But that is another topic. Here we talking about remembering the user selected audio and subtitle track.

[luisteraar]

you're right

I mus take a good look at db.cpp

station name is also unique

[littlesat]

Also station name is not unique…

the biggest issue in our community is that there is a problem and out of the blue someone creates a solution…. Without discussing what method and what way the solution can best be resolved… there is nothing that intends to work together… and when someone remarks a solution is not that optimal is forbidden thing to say… 

What are no the bases what is stored in that pickle file to get the service and the audio/sub channels?

we have lamedb since the beginning what is lame…. So this was never finished at all and optimal and every time it was tried (also with db5) the project stops because we cannot work together…. We get what works and tell we are better that we can do something (and do not tell it is created on an ughly way) and take what is after all the better smooth and streamlined solution…

 

maybe this time we can better discuss how we can make this more streamlined….

Edited by littlesat, 5 October 2023 - 14:54.

[DimitarCC]

 

What are no the bases what is stored in that pickle file to get the service and the audio/sub channels?

 

 

If you are asking how i identify the channel in my solution then in the pickle storage file i save a key-value pairs.

The key is the service reference.

The value is the subtitle and/or audio track dumped data.

 

When i say that the key is service reference maybe it is not what you imagine. So let me explain....

 

Lets assume you have a IPTV stream in userbouquet. There normally it is defined like that

#SERVICE 4097:0:19:8A0:16:70:10:0:0:0:http%3a//xxxxxxxx/index.m3u8:Channel Name 1 HD

So to make EPG to work with such a channel it is required the first part of the definition (in the example it is 4097:0:19:8A0:16:70:10:0:0:0) to be unique.

So i have the same requirement for my implementation. in the key i write 4097:0:19:8A0:16:70:10:0:0:0. And that must be unique so to identify the channel.

 

Ofcource my solution will not work with m3u importers that generate random srefs everytime. Also it will not work with third party plugins that have their own infobars and channellists.

 

My solution is mainly for the storing of IPTV services in userbouquets.

[luisteraar]

 

 

 

 

I

 

Ofcource my solution will not work with m3u importers that generate random srefs everytime. Also it will not work with third party plugins that have their own infobars and channellists.

 

My solution is mainly for the storing of IPTV services in userbouquets.

 

A lot of restrictions.

maybe is url more suitable.