Jump to content


Photo

Stream stop


  • Please log in to reply
119 replies to this topic

Re: Stream stop #21 Siki78

  • Senior Member
  • 55 posts

+2
Neutral

Posted 5 December 2015 - 23:13

@mx3L

Thanks for all!!



Re: Stream stop #22 ozzsurf

  • Senior Member
  • 131 posts

+3
Neutral

Posted 6 December 2015 - 00:28

#SERVICE 4097:0:1:0:0:0:0:0:0:0:http%3a//151.252.32.234%3a8080/stream/?channel=nettvinfo&stream=1mb&sp=nettvplus
#DESCRIPTION music

working for me
with no User-Agent
and deleted ?u=USER

from the url

 



Re: Stream stop #23 ozzsurf

  • Senior Member
  • 131 posts

+3
Neutral

Posted 6 December 2015 - 00:57

when playing on pc

can sniff real stream url

which is

#SERVICE 4097:0:1:0:0:0:0:0:0:0:http%3a//151.252.32.226%3a8080/stream/?fw=true&route=ch3_2&rule=null&sp=nettvplus&channel=nettvinfoint&stream=1mb&adaptive=null&b=3&player=http:music
#DESCRIPTION music

and works in vu as above


Edited by ozzsurf, 6 December 2015 - 00:58.


Re: Stream stop #24 Siki78

  • Senior Member
  • 55 posts

+2
Neutral

Posted 6 December 2015 - 06:45

You can not delete u=USER we need valid user for the other channel that is promo. What you write about link it is a same this link or my link they follow useragent. And with today update works with 4097 but with 1:0:1 i think it dosent work. I gone report you after

Re: Stream stop #25 Siki78

  • Senior Member
  • 55 posts

+2
Neutral

Posted 6 December 2015 - 07:32

My friend say to me it works but it is not stable volume and bild doesnt works properetly.

Re: Stream stop #26 ian1095

  • Senior Member
  • 462 posts

+6
Neutral

Posted 6 December 2015 - 08:12

THIS UPDATE IS A JOKE !

 

https://github.com/O...c64d537c17bfb48

 

This breaks many third party plugins and for what ?

 

90% of users dont even know what alternate headers and user agents are let alone know how to use and add them to their bouquet list. Whilst its true that its down to plugin writers to fix,all you have achieved here is to create massive amounts of work for the plugin writers. As each and every scraper will need new code added to it.

 

You forget that most plugins cater for other images than just Pli, and whilst adding the fix to the final link in the plugin code will make the plugin work once more with Pli,with this fix,it then also stops the plugin from working on other images such as BH

 

This is an example of the fix now working in Pli but after it being applied it no longer works in BH after this was added to the final link in a scraper.

https://s3.amazonaws.com/otmar-misc-tv/beverlyhillbillies-01-clampettsstrikeoil.mp4#Host=s3.amazonaws.com&User-Agent=Mozilla/5.0 (Windows NT 5.1; rv:42.0) Gecko/20100101 Firefox/42.0&Accept=text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8&Accept-Language=en-US,en;q=0.5&Referer=http://oldtimemoviesandradio.net/otmar-misc-tv/beverlyhillbillies-01-clampettsstrikeoil/&Connection=keep-alive

This seriously needs reverting or at the very least patching properly to allow both methods of adding alternate user agents and headers.

 

Ian.


Edited by ian1095, 6 December 2015 - 08:16.


Re: Stream stop #27 littlesat

  • PLi® Core member
  • 57,186 posts

+699
Excellent

Posted 6 December 2015 - 09:44

The joke is that bh can adapt.... And not only bh, also other images as our sources are open....
I think we should not stop development for (closed source) third party plugins... These can also adapt...

Edited by littlesat, 6 December 2015 - 09:49.

WaveFrontier 28.2E | 23.5E | 19.2E | 16E | 13E | 10/9E | 7E | 5E | 1W | 4/5W | 15W


Re: Stream stop #28 mdshd

  • Senior Member
  • 140 posts

+3
Neutral

Posted 6 December 2015 - 10:09

How to know which UserAgent? What is the UserAgent exactly for? For the provider to know from which device we are watching or for what?

 

For me now works all fine as before with 1:0:1 on ET10000.

 

Thanks!


Edited by mdshd, 6 December 2015 - 10:09.

ET10000


Re: Stream stop #29 ian1095

  • Senior Member
  • 462 posts

+6
Neutral

Posted 6 December 2015 - 10:16

Why do you wish to make life harder for people ?

 

Only Pli and Pli based images will have these changes,other images such as BH and VTI will not.

 

So as a result,they simply cannot understand the final link with the headers and user agents applied to that final link in any scraper,not just closed source plugins.

 

What have you achieved by doing this ?

 

As stated very few users even know what a header or user agent is,and this is proven by the post above as the user is asking what this even is,so this feature will be used by final users very rarely if at all,but as a result you have created loads of work for all plugin writers and plugins than at the moment work on all images,will cease to do so.

 

I cannot see nor understand the logic in this or whats to be gained from it.

 

Do you not agree that at least patching it to allow both methods would be a far better idea and make life so much better and easier for everyone ?

 

Ian.


Edited by ian1095, 6 December 2015 - 10:18.


Re: Stream stop #30 littlesat

  • PLi® Core member
  • 57,186 posts

+699
Excellent

Posted 6 December 2015 - 10:54

As we are opensource BH and VTI can simply adapt... So please tell them they should adapt and stop bothering and arguing here.

I think we should not limit our progression to (as far I know) closed source images like BH and VTI... This will be really the opposite world....


Edited by littlesat, 6 December 2015 - 11:07.

WaveFrontier 28.2E | 23.5E | 19.2E | 16E | 13E | 10/9E | 7E | 5E | 1W | 4/5W | 15W


Re: Stream stop #31 Siki78

  • Senior Member
  • 55 posts

+2
Neutral

Posted 6 December 2015 - 10:59

No works like sharm sorry for bad info. Thanks to @mx3L and openpli team

Re: Stream stop #32 ian1095

  • Senior Member
  • 462 posts

+6
Neutral

Posted 6 December 2015 - 11:09

Why would they adapt ?

 

Its a useless update that very few users will ever make use of so I doubt very much they will follow.

 

So as it stands right now,if I were to write a plugin today that requires alternate headers and user agents,I would have to decide if I wanted it to work on either Pli or BH but not both.

 

I'm not arguing with you, I'm just pointing out that this update has made life much harder. I thought that the whole idea of open source was that people could contribute as they saw fit and up to their skill levels and you listened to peoples ideas and views if that's all they were able to offer.

 

I can see though that I'm not getting through to you and you are not going to patch this to allow both methods,so as you say its not worth posting further on this matter.

This is a shame,but hey it is what it is, you do what you want and we as users have to get over it and deal with it.

 

Ian.



Re: Stream stop #33 mx3L

  • Senior Member
  • 616 posts

+79
Good

Posted 6 December 2015 - 12:16

We specifically choosed '#' as delimiter because it's normally not send to the server by client and is only used by client. Gstreamer uses SOUP as http client and it doesn't send anything after '#' delimiter:

client part:

gst-launch-1.0 playbin uri="http://127.0.0.1:7777/path#fdsakfjdls;afjkdlsa;jfsdkl;fjasdklfjkl"


server part:

import SimpleHTTPServer
import SocketServer
import logging
PORT = 7777
class ServerHandler(SimpleHTTPServer.SimpleHTTPRequestHandler):

    def do_GET(self):
        logging.error("request line: %s"%self.raw_requestline)        
        logging.error("headers: %s", self.headers)
        # whatever else you would like to log here            
        SimpleHTTPServer.SimpleHTTPRequestHandler.do_GET(self)
                       
Handler = ServerHandler
                                                   
httpd = SocketServer.TCPServer(("", PORT), Handler)
                             
print "serving at port", PORT
httpd.serve_forever()


result:

ERROR:root:request line: GET /path HTTP/1.1
ERROR:root:headers: Host: 127.0.0.1:7777
Connection: close
icy-metadata: 1
User-Agent: GStreamer souphttpsrc libsoup/2.44.2

So you can clearly see that it doesn't break url at all, in following case everything including '#' is not send, i.e "#fdsakfjdls;afjkdlsa;jfsdkl;fjasdklfjkl".

 

I removed from servicemp3 parsing of extra headers from service reference path string and add old way of passing headers via config.

Your example as expected works even with added '#'.

https://s3.amazonaws.com/otmar-misc-tv/beverlyhillbillies-01-clampettsstrikeoil.mp4#Host=s3.amazonaws.com&User-Agent=Mozilla/5.0 (Windows NT 5.1; rv:42.0) Gecko/20100101 Firefox/42.0&Accept=text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8&Accept-Language=en-US,en;q=0.5&Referer=http://oldtimemoviesandradio.net/otmar-misc-tv/beverlyhillbillies-01-clampettsstrikeoil/&Connection=keep-alive

So this statement is certainly wrong:

 

This is an example of the fix now working in Pli but after it being applied it no longer works in BH after this was added to the final link in a scraper.

 

 

Its a useless update that very few users will ever make use of so I doubt very much they will follow.

+ http headers can be set individually for every service

+ http headers are no longer globally set for every service

+ User-Agent http header is not treated as special case, but as any http header

- plugins which used UserAgent and extraHeaders via config, need to slightly update they playback source code

 

So as it stands right now,if I were to write a plugin today that requires alternate headers and user agents,I would have to decide if I wanted it to work on either Pli or BH but not both.

Not true as stated in the begining

 

I can see though that I'm not getting through to you and you are not going to patch this to allow both methods,so as you say its not worth posting further on this matter.

This is a shame,but hey it is what it is, you do what you want and we as users have to get over it and deal with it.

I can provide patch so also setting http headers old way is still possible, but when plugin sets extraHeaders or userAgent config and don't unset it after usage, then for next playurl (for gstreamer) will be these headers sent! (if not overriden by headers set in service reference path string)



Re: Stream stop #34 athoik

  • PLi® Core member
  • 8,458 posts

+327
Excellent

Posted 6 December 2015 - 13:24

I think we choose wise the #. Maybe we can have global setting too!

So it will be like this: user agent in url > user agent global config > default user agent.

I suggest to make one global setting for both servicedvb and servicemp3. Global user agent must be configurable from UI.
Wavefield T90: 0.8W - 1.9E - 4.8E - 13E - 16E - 19.2E - 23.5E - 26E - 33E - 39E - 42E - 45E on EMP Centauri DiseqC 16/1
Unamed: 13E Quattro - 9E Quattro on IKUSI MS-0916

Re: Stream stop #35 ian1095

  • Senior Member
  • 462 posts

+6
Neutral

Posted 6 December 2015 - 13:38

Yes I can see its not without its merits now and it will allow access to host sites such as mail.ru without having to go through wget

 

But, without backwards compatibility, in the form of a global setting,then if plugin writers adjust their plugins to use this method alone, and as its on the final link,I can see no way of making a plugin use both both the old and new methods together,then images that have not added last weeks code such as BH and VTI will not be able to use these plugins, but if plugin writers dont do this then these plugins wont work on Pli (catch 22)

 

As Littlesat says,Its not your fault that the like of BH and VTI have not "yet" made use of your new code, but until they do, if they ever do,then any plugin that requires alternate user agents and headers cannot work on all images.

 

However this may still be possible, its just we have not seen a way of doing it yet.

 

Whichever way you look at it though it is quite a bit of work for plugin writers to make things work once more as things stand right now.

 

Ian.


Edited by ian1095, 6 December 2015 - 13:42.


Re: Stream stop #36 ozzsurf

  • Senior Member
  • 131 posts

+3
Neutral

Posted 6 December 2015 - 14:49

You can not delete u=USER we need valid user for the other channel that is promo. What you write about link it is a same this link or my link they follow useragent. And with today update works with 4097 but with 1:0:1 i think it dosent work. I gone report you after

You can not delete u=USER

i did and it works

no user agent and no user

on original link and sniffed link

post other channel that is promo



Re: Stream stop #37 ArgonX

  • Member
  • 17 posts

0
Neutral

Posted 6 December 2015 - 22:13

Hi Ozz..

 

that is only channel that is promo and dont need to have u=USER.

 

All other ones we have need to have u=User which identified the user as paid customer.

 

After latest update on OPENPLI i can see that 1:0:1 reference work as charm again !!

 

PS. Openpli Team : thank you !



Re: Stream stop #38 bacicciosat

  • Senior Member
  • 540 posts

+100
Excellent

Posted 6 December 2015 - 23:08

Don't worry about bh.

There is openbh that it is open and pli based and have only to be updated. And bh will adapt itself.

But the problem is not for bh but for vu+ official and derivates. It will be alway more difficoult to survive for Vu+ derivates in this way. And always with not nice words and comments.

But you think you are always in the right and good. ok. let's go.



Re: Stream stop #39 ArgonX

  • Member
  • 17 posts

0
Neutral

Posted 7 December 2015 - 21:39

Don't worry about bh.

There is openbh that it is open and pli based and have only to be updated. And bh will adapt itself.

But the problem is not for bh but for vu+ official and derivates. It will be alway more difficoult to survive for Vu+ derivates in this way. And always with not nice words and comments.

But you think you are always in the right and good. ok. let's go.

Lets hope they will do it.. I am happy that za VU+  I can use openpli since day and not having single issue. Plus I love OPENPLI team which is open on suggestions and really here to help end users with some issues.

 

I know there is so many of my friends who have DM 800 HD or so and still doesnt work for them..

 

Pitty is there that there is no openpli 4.0 for dreambox 800HD..



Re: Stream stop #40 littlesat

  • PLi® Core member
  • 57,186 posts

+699
Excellent

Posted 7 December 2015 - 23:43

I think this kind of discussion will never stop as long vu stays stuck on their fork of the dmm image...

Edited by littlesat, 7 December 2015 - 23:45.

WaveFrontier 28.2E | 23.5E | 19.2E | 16E | 13E | 10/9E | 7E | 5E | 1W | 4/5W | 15W



1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users