119 replies to this topic

[Erik Slagter]

Do we have a bitbake recipe and a public repository with the source?

[Rob van der Does]

https://github.com/o...toBouquetsMaker

Edited by Rob van der Does, 30 September 2015 - 12:42.

[Erik Slagter]

Now yet a bitbake recipe that "fits" into our OE and we can start.

[Rob van der Does]

Yeaah, there must be something left for you to do.....

[peterjo]

You can have a look here:

https://github.com/o...ance-plugins.bb

[Erik Slagter]

Last few times I tried to import an OE-A plugin it appeared to be "more work" than writing the plugin So I am a bit reluctant now. Even more because I don't like this kinds of plugins anyway.

Edited by Erik Slagter, 1 October 2015 - 16:59.

[Rob van der Does]

Last few times I tried to import an OE-A plugin it appeared to be "more work" than writing the plugin...

Well, then I'm very anxious to see your version very soon

 

No idea why building from the OE-A feeds is so difficult? Many image-builders do it.

[Huevos]

Even more because I don't like this kinds of plugins anyway.

You've not even used it. How can you say you don't like it?

[Rob van der Does]

I know why: Erik wants to scan all satellites and make his own bouquets. And TBH: I do the same.

 

But nevertheless: ABM is liked by many users, as it mimics the setup as suggested by the provider, allows for some manipulation of the bouquets and works really fast. And it supports a whole range of satellite and cable providers.

And sometimes I use it temporarily to check a provider versus my bouquets. And ABM allows an easy deleting of all ABM bouquets.

 

 

@Erik: Be honest; we (i.e. every image and plugin maker) don't do all the 'work' for ourselves, but mainly for all kinds of people using an image/plugin.

[Erik Slagter]

Rob is correct.

 

And not everything I do is work I have direct advantage of myself, but channel lists (generators) is where I really draw the line. There are plenty technical people that use channel lists themselves, so volunteers please stand forward!

[Huevos]

 

Rob is correct. [...] channel lists (generators) is where I really draw the line.

 

This isn't just a channel list (generator). That function can be disabled.

 

Can also be used for scanning. Here is an example.

 

PLi network service scan of 28 east: 20 minutes.

AutoBouquetsMaker network service scan of 28 east: 35 seconds.

Number of channels found: identical.

[littlesat]

You cannot compare a service scan with a scan that can be compared with the Fastscan of course...

 

My concerns are more against the way it is implemented. I think it was better that this functionality was added to the Fastcan feature we currently have. 

 

But I'm aware the creators of the ABM plugin need to have more knowledge of c++ to implement it that way... And I understand the really decoding stuff is easier to handle/debug in python by those who develop it...

 

And as separate plugin it should be performing on openpli aswell... So who cares.... 

Edited by littlesat, 2 October 2015 - 09:59.

[peterjo]

It is running fine in PLI. There is only a warning during install of a clib.

Probably we have compiled it with a newer clib.

Beside this small issue, ABM is a dynamic plugin.

Often the provider list changes (more providers) and new functionality is made.

When users are using the plugin it would be nice all these changes are direct in a new image build and updated during image update.

Not every month (or 2) download the ipk and reinstall it again.

[Erik Slagter]

 

Rob is correct. [...] channel lists (generators) is where I really draw the line.

 

This isn't just a channel list (generator). That function can be disabled.

 

Can also be used for scanning. Here is an example.

 

PLi network service scan of 28 east: 20 minutes.

AutoBouquetsMaker network service scan of 28 east: 35 seconds.

Number of channels found: identical.

My 28 east scan doesn't take quite 20 minutes. Of course you should take care to not have feeds in your xml file... Also in theory it would be suffienct to only scan the transponder that carries the NIT for "all" providers (1? 2?) and then the rest would go in a breeze. I've never tried that though. ATM I can't, no dish...

[Huevos]

My 28 east scan doesn't take quite 20 minutes. Of course you should take care to not have feeds in your xml file

We don't have feeds in our XML file. https://github.com/o...commits/master

It was me that suggest their removal from the OpenPLi XML file several years ago.

 

Also in theory it would be suffienct to only scan the transponder that carries the NIT for "all" providers (1? 2?) and then the rest would go in a breeze.

NIT doesn't contain services. That is in the SDT, in descriptor 0x46 for other transponders, on the few transponders that carry it.

 

My concerns are more against the way it is implemented. I think it was better that this functionality was added to the Fastcan feature we currently have.

Fastscan and Cablescan both embed code directly in the E2 binary. How is that a portable plugin if It relies on the image having embeded code? That means it will only run on images that take on that code and can't just be loaded by the user into any image. Also Fastscan is just one type of provider. Most providers do not use that system.

Edited by Huevos, 2 October 2015 - 18:55.

[Erik Slagter]

If you have all the right transponders (either from a 100% correct xml file or from the NIT), a complete scan doesn't have to take a long time, that's my point. I would never trust a mechanism that injects services not directly retrieved from the SDT.

 

Anywyay this is a moot point, I don't say we can't have such a mechanism, only that I won't spend time on it.

[Dimitrij]

If you have all the right transponders (either from a 100% correct xml file or from the NIT), a complete scan doesn't have to take a long time, that's my point. I

Is not a fact.

Example 28.2 e...

I'm in Latvia accept only part of the transponder, so a very long scan.

[Erik Slagter]

Yes, that is true. In your case you'd need to strip the transponders you can't receive.

[peterjo]

Yes, that is true. In your case you'd need to strip the transponders you can't receive.

and that is just functionality ABM has.

When the provider XML is configured well, you will get the correct channel-list of the chosen provider and when the services on the transponder has been changed by the provider, it will auto correct your channel-list.

[littlesat]

As far I understand ABM does not strip/scan the transponders...

 

It reversed engineered the channel list the provider transmits for their own boxes... In fact it only reads the data from these channel lists so in fact it reads only one transponder...

 

It is a pitty that it is a plugin instead of an improvement of the fastscan plugin...