97 replies to this topic

[theparasol]

Well I honestly dont think Chris wants to be recognized as "the best", his bad word choice and bold language was just to trigger people to get attention at last.

He was just disappointed that the found serious issue that solves most if not all weird thing he was facing with gstreamer development wasn't embraced by others like he thought it would be.

He merely is asking in his own not so nice way for more attention he thinks the debug locking matter is needing.

If you work long enough on a part of a project and you get a breakthrough you want at least some positive feedback. Instead there was too much silence.

Than it seems no one cares, no one realizes the impact? Whats going on? The locking is not the problem? Different people have different views on issues and how to solve them.

 

Lets look forward and hope some devs have time to burn and  discus and address this matter to solve it in a proper and structural  way.

 

Still wishing I could contribute more in code than posting words on the forum

[WanWizard]

I don't think anyone said it is not a problem.

 

There is a difference in perception on how much of a problem it is, and how to address it (i.e. hack it vs structurally solve it).

 

My personal opinion is that it is not much of a problem. The reason the spinners appear is because insane amounts of debug log entries are produced. And the reason this is a personal problem for Chris is that he's debugging gstreamer. The reason it is a public problem is that these debug calls are enabled for end-user images while they shouldn't.

 

So there are two problems to solve:

- how to make debugging easier

- how to get rid of debugging (at this level) in end-user images

 

The first point might be adressed by what Chris has proposed, and what Mirakels has committed. The second point requires more complex changes, and therefore thought and discussion on how to address this problem, without introducing others (like not being able to get debugging info from end-users anymore).

 

I haven't written any C(++) in over 25 years, so I can't comment in detail about code, and I don't. But I can read it very well. I do write PHP on a daily basis, and as a software architect I try to keep a helicopter view of things, and think and plan more for the long term. So perhaps talking on different "planes" might also contribute to the difference of opinion here.

[mirakels]

@theparasol: it reason why this specific line fix was not recognised as th issues was that it was hidden around lots of ranting about the eDbug system and the fact that  this topic states it is dealing with a 'eDebugNoLock' function.

So as said in one of my previous post here: try to be specific and to the point. Don't hide the issue in a fairytale.

[theparasol]

Yep, I know, its hard to read all what is said especially when a user posts not 1 but 2,3,4,5 or even 6 sequentially follow ups in same topic. Sometimes even contradicting what said previously. You can easily think: let me just read the very last one, that should be enough!

For that reason the 5 minute reply change time is just too short. I wonder if the forum engine supports something like: you may change your posting until somebody else places a follow up.

[WanWizard]

For that reason the 5 minute reply change time is just too short. I wonder if the forum engine supports something like: you may change your posting until somebody else places a follow up.

 

No, unfortunately it doesn't. That would make things a lot easier. There is a upgrade on the todo list, not sure what that has as improvements. And otherwise we can always try a feature request...

[christophecvr]

Well fun    or eighter sad   

 

Looks like a big number off pli core members do really not understand the problem.  If there is a problem they do not want to here about it. 

I've been yelled on and insulted repetitively for just telling the truth .

 

I've been testing working since more the 2 years now on specific very severe problems of e2 . Especially for the base e2 gst player which is servicemp3.cpp.

But also for real live playing dvd support and cd support.

 

On so what all this points I since start always had an extreme counteracting of whatever I wanted to  improve or add.

 

This topic is really the final prove of it.

 

Collection :

http://forums.openpl...ndpost&p=524728

 

I wanted to add more collections. But persons who are really interested should read this topic from start to end. I just did it

 

And sorry WanWizard : .... The only and only who had and has a bad attitude is mirakels , ... He's the guy who refrains any improvement and turned  this topic from what I intended :

 

"A Real positf constructif item with a intelligent debate about some core issues like the eDebug"  TO :

 

"THE MOST IDIOT CONVERSATION"

 

 

I do not think that more is needed .

 

So if You want keep on dreaming that  the problems will solve themselve. ..... ? Just do and follow the KING mirakels I'm shure next year all will be solved after all he told that he's THE EXPERT KNOWS ALL ABOUT ENIGMA2 AND WE ARE ALL IDIOTS .

[athoik]

??

[Taapat]

@christophecvr I am very sorry.
I am really not the person which could to point you something.
But I see that you have exactly the same conflict in gstreamer bug repport: https://bugzilla.gno...g.cgi?id=760477
Do you really think that with your attitude everything is ok?
I am sure that if you simply, without emotion, explain the problems (which you've already done here), then the error correction progress would be much faster.

[christophecvr]

??

Well nice reflection cause that's mine after the answers from some 'PLI@ Core members'

 

Let them come with a descent reason why they ignore severe problems. ?????

[athoik]

It was asked from your christopher to create patches with a proper description.

Now you are coming back (again and again) insulting people.

Really what is your problem?

[christophecvr]

@christophecvr I am very sorry.
I am really not the person which could to point you something.
But I see that you have exactly the same conflict in gstreamer bug repport: https://bugzilla.gno...g.cgi?id=760477
Do you really think that with your attitude everything is ok?
I am sure that if you simply, without emotion, explain the problems (which you've already done here), then the error correction progress would be much faster.

Write and wrong,

 

If I did not insist or pushed like I did nothing would have been done . Now not only it's done but also the problem was understand by the guys.  At the end it's solved (and only because I pushed) . Perhaps if I did pushed it would came out much much later and then ....

[malakudi]

@christophecvr: The problem is a. your attitude b. the way you right posts: adding a post after another - sometimes contradicting what you wrote before. It is difficult for someone to follow.

 

Indeed, you have found an issue that needs to be fixed. And it will be fixed. Telling people "this needs to be fixed yesterday" doesn't help anything. Nothing needs to be fixed yesterday. Even if all images are failing and not working, still nothing needs to be fixed yesterday. There is no warranty, either expressed or implied, that the OpenPLi images will work. There is no warranty that they will not brick your STB, or put a fire on your house. This is how it was and how it will be. So you either play along and try to push your fixes by being polite and understand that this is a hobby and not a real work, or you stop developing with OpenPLI as your base and choose a different tree of development. Thank God, there are many of them! (OE-Alliance, VU+ codebase and others).

[christophecvr]

It was asked from your christopher to create patches with a proper description.

Now you are coming back (again and again) insulting people.

Really what is your problem?

Sorry athoik but only one and single patch is already used with the right clear comments, and very very limited to just only servicemp3.cpp. With some very very few commenting acts . That's all and off course never applied cause You do do not (eighter just not want) to understand it.

 

And the patch was included here :

 

Really good commented :

Attached Files

•  e2-servicemp3-some-eDebug-commenting.patch   5.38KB   8 downloads

Edited by christophecvr, 18 January 2016 - 20:00.

[christophecvr]

@christophecvr: The problem is a. your attitude b. the way you right posts: adding a post after another - sometimes contradicting what you wrote before. It is difficult for someone to follow.

 

Indeed, you have found an issue that needs to be fixed. And it will be fixed. Telling people "this needs to be fixed yesterday" doesn't help anything. Nothing needs to be fixed yesterday. Even if all images are failing and not working, still nothing needs to be fixed yesterday. There is no warranty, either expressed or implied, that the OpenPLi images will work. There is no warranty that they will not brick your STB, or put a fire on your house. This is how it was and how it will be. So you either play along and try to push your fixes by being polite and understand that this is a hobby and not a real work, or you stop developing with OpenPLI as your base and choose a different tree of development. Thank God, there are many of them! (OE-Alliance, VU+ codebase and others).

Very nice , But I want to go and not stagnate.  You are only want to stagnate ok .

 

If pli want's to stop to be at the each ok, did you know that all other images at last almost all are pli based and just modified left or right ?

[athoik]

Really good commented :

What that semicolon does? Copy paste error?

-		eDebug("[eServiceMP3] TOC entry from source %s not used", GST_MESSAGE_SRC_NAME(msg));
+		//eDebug("[eServiceMP3] TOC entry from source %s not used", GST_MESSAGE_SRC_NAME(msg));
+		;

Please create again the patch against current head.

[christophecvr]

 

Really good commented :

What that semicolon does? Copy paste error?

-		eDebug("[eServiceMP3] TOC entry from source %s not used", GST_MESSAGE_SRC_NAME(msg));
+		//eDebug("[eServiceMP3] TOC entry from source %s not used", GST_MESSAGE_SRC_NAME(msg));
+		;

Please create again the patch against current head.

 

No at start and developement I needed I check from where the chapters info into the media came. By the fact that we are using gstreamer looks like we had the chapter info for all we had into the base stream,

 

This means :

 

Every audio stream

Every sub stream

Every video stream.

 

That's why off course I needed the use off that info to ONE element and that's the video stream that I choose for that , .... if i'm wrong or was wrong tell it, I may learn a lot of it.

[christophecvr]

 

Really good commented :

What that semicolon does? Copy paste error?

-		eDebug("[eServiceMP3] TOC entry from source %s not used", GST_MESSAGE_SRC_NAME(msg));
+		//eDebug("[eServiceMP3] TOC entry from source %s not used", GST_MESSAGE_SRC_NAME(msg));
+		;

Please create again the patch against current head.

 

It's in for if or then  that's needed cause  a ruleneeds to be  terminated by ; since by commenting it would have been a if  or else beind it does not compile or ... that's why the <;>

Edited by christophecvr, 18 January 2016 - 20:48.

[athoik]

It's in for if or then  that's needed cause  a ruleneeds to be  terminated by ; since by commenting it would have been a if  or else beind it does not compile or ... that's why the <;>

Then remove the else and recreate the patch. BTW the else without body causes no issues since {} exists.

$ cat test1.c
#include <stdlib.h>
#include <stdio.h>

int main()
{
        if(1)
        {
                printf("ok\n");
        }
        else
        {
                //printf("whatever");
        }
        return 0;
}

$ gcc -Wall test1.c
$ ./a.out
ok

[christophecvr]

Oh bloody hell I can't change my responce above.

 

So @athoik

 

Yes the ";"  is in after a line into a if then has been commented.

 

This means that is correctt in the patch.

[christophecvr]

 

It's in for if or then  that's needed cause  a ruleneeds to be  terminated by ; since by commenting it would have been a if  or else beind it does not compile or ... that's why the <;>

Then remove the else and recreate the patch. BTW the else without body causes no issues since {} exists.

$ cat test1.c
#include <stdlib.h>
#include <stdio.h>

int main()
{
        if(1)
        {
                printf("ok\n");
        }
        else
        {
                //printf("whatever");
        }
        return 0;
}

$ gcc -Wall test1.c
$ ./a.out
ok

But here is now just the discussion . In this very base patch I only and only and only commented some culprit ,culprit, culprit eDebug lines.

 

If those lines where incuded into a  if(blablbla) {xxxxx;}  or a else {xxxxxxxx;}

 

If by commenting the line {xxxx;} we would be in a stuck if or else operation. Thats why the {;}   Hope You see why. And hope You understand I do not want to take away such commenting eDebug lines but wait on developers constructive discussion