128 replies to this topic

[christophecvr]

ok no I see what you changed. Next time just show a diff output of the versions and if you think that is not enoug, show the original version AND the diff.

Showing the before and after code makes it very hard to compare even if you mention line numbers (but fail to show the line numbers in the examples)

 

pushed your fix for it

Ok thanks.

 

If You wish I also will make a patch for the bsod.cpp self to just correct the line : 87

if (!key.compare(0, size, line) && line[size] == '=') {

to :

if (!line.compare(0, size, key) && line[size] == '=') {

That has nothing to do with debug level and level at start self.  But well with autocrash recovery after e2 restart it will start with the users selected skin and not with the pli defaults skin then. That was one off the goals which did not worked out due to a small human error.

[christophecvr]

 

ok no I see what you changed. Next time just show a diff output of the versions and if you think that is not enoug, show the original version AND the diff.

Showing the before and after code makes it very hard to compare even if you mention line numbers (but fail to show the line numbers in the examples)

 

pushed your fix for it

Ok thanks.

 

If You wish I also will make a patch for the bsod.cpp self to just correct the line : 87

if (!key.compare(0, size, line) && line[size] == '=') {

to :

if (!line.compare(0, size, key) && line[size] == '=') {

That has nothing to do with debug level and level at start self.  But well with autocrash recovery after e2 restart it will start with the users selected skin and not with the pli defaults skin then. That was one off the goals which did not worked out due to a small human error.

 

 Oh I saw that the error into bsod.cpp has already been patched on commit :

 

https://github.com/O...aaff2e1d32f4da1

 

But what about the really evolution to a user possible selection of debug level patch I'll post this patch again here. ?

Attached Files

•  e2-DebugLevel-Configurable-trough-user-interface_4.patch   3.83KB   17 downloads

[mirakels]

not acceptable. you copy a function that is already available elsewhere. And you you still need to restart to make it effective. You did not read/listen to what is said in the comments of the debuglevel patch and what littlesat pointed you at.

[christophecvr]

not acceptable. you copy a function that is already available elsewhere. And you you still need to restart to make it effective. You did not read/listen to what is said in the comments of the debuglevel patch and what littlesat pointed you at.

- 100

 

The biggest bullshit ever written for me over and out bye 

[littlesat]

I'm afraid you did not understand how this should be made... 

And telling us bullshit and over and out bye does not help either....

 

You can discover nice thinks - we appreciate - but this will not mean you can force us to commit everything you did... And this is not only valid for you... even for me...!

Edited by littlesat, 10 February 2016 - 08:07.

[TheVoice]

Sorry to barge in here not being a developer - but I do have a question.

 

Reading the several threads about this - it looks like this is the bug that's been bothering me on our DM8000 for a few years already - watching movies and noticing that sometimes subtitles cause issues, ending up in completely freezing the machine.

 

If this is basically caused by something filling up because of excessive debug information - Is there currently a way in the latest OpenPli image to disable debugging? Or - has this been fixed/applied recently?

 

Apologies if the answer was in the threads already - but it's difficult for me to understand what's part of coding speak and what's part of standard configuration / commands.

[betacentauri]

Subtitles from mkv files? If yes, then this doesn't solve the problem. There is a known bug in gstreamer 0.10 regarding this.

[kalehrl]

watching movies and noticing that sometimes subtitles cause issues, ending up in completely freezing the machine.

 

In my case, it only happens if you have both embedded and external subtitles.

Just remux files to remove embedded subtitles and use only external ones.

[Erik Slagter]

Or the other way around, which is preferable.

[MastaG]

Hmm, I also have MKV movies with internal subtitle tracks.

But I always add an external .srt file with dutch subtitles as well.

Then in the audio selection menu I can select between the ones which are embedded and the external one as well.

Works pretty good here.

Edited by MastaG, 12 February 2016 - 12:51.

[Erik Slagter]

I have the same experience, no problems there.

[MastaG]

Oh wait, now I see betacentauri said it was a known bug in gstreamer 0.10.

And I'm using my own gstreamer 1.7 image.

I guess the only thing I used to get is the spinner that sometimes comes up when it's loading new subtitle text to display.

It happens randomly.. where the movie continues playing (with a spinner) and the subtitles are halting for like 10 seconds.

Then they come back and everything runs smoothly.

Commenting out the eDebug's in servicemp3.cpp fixed this for me.

 

PLi5 will be awesome!

Edited by MastaG, 12 February 2016 - 13:41.

[littlesat]

gstreamer 1.7 is not stable....
http://gstreamer.freedesktop.org/

Edited by littlesat, 12 February 2016 - 16:07.

[MastaG]

Yes I know.

But it works great here so far.

[betacentauri]

I answered TheVoice. And afaik the debug problem causes no crash of the box. "Only" spinners for a little timespan. So I think the new debugging systems won't solve his problems.

[littlesat]

It solves more... previously it might happen streams did not start...

[betacentauri]

Ok, didn't know. But also box crashes?

[littlesat]

No "infinitive" spinners and not starting streams....

[TheVoice]

Thanks for answering guys. I actually had it happen to me again - MKV with (indeed) embedded subs.

 

Also I was surprised by another crash (spinner) while watching a recording from TV . It's terrible.

[christophecvr]

Actually a lot of things are causing bugs. That is normal in a live project.

 

However Yes the use off eDebug is currently a major rather a critical e2 bug.

 

The current so called expert(s)(they call them self experts??)    do not agree. ok nice

 

I do not care any more. I compile my own image adapt to my needs and that’s it .

 

I'm even happy with it