I found the man who is responsible for breaking things that worked.
"If it ain't broke, don't fix it!"

https://ffmpeg.org/pipermail/ffmpeg-cvs ... 99152.html

So he is the one person responsible for this giant codecpar disgrace that people around the world face.
He did not provide anyone with instructions on how to handle this issue with incompatible programs such as KDEnlive.

Of course, KDEnlive developers also have failed resolving this since two years, but FFmpeg is also used elsewhere, and Khirnov is potentially toxic to every program that utilizes FFmpeg.

Anton Khirnov

He was unable to maintain compatibility, and is the person who is responsible for all the conflicts regarding codecpar.

I suggest the developer team to get rid of him for this felony. Immediately.

If I search the word "deprecated", I get 821 results!

And look at this: https://ffmpeg.org/pipermail/libav-user ... 09800.html .


Yes, Anton should have used a few more brain cells before messing around with FFmpeg.

…then guess what: don't fix it, Mr. Khirlov!

Please never scapegoat anybody under any condition.
and refrain from blaming even more when you don't really understand what you are talking about!
The "deprecated" warning is... a warning, not an error! Compatibility is still maintained for long, it just tells downstream developers that the old approach was discovered to cause problems, solved with the new API (with a very informative commit message).

The problem now is on MLT side, but I won't blame them for willing to maintain compatibility with older FFmpeg (shipped with older distros), we are just waiting for someone writing the code compatible with both new and old FFmpeg (maybe me someday, really no time for now).
No emergency: for the moment, everything works, this warning is very not probably the cause of your problems.

vpinon wrote:Please never scapegoat anybody under any condition.
and refrain from blaming even more when you don't really understand what you are talking about

That's not my job, but I agree.
But “everything works” is not really true since 2016. If it did, rendering would not crash.

However, I appreciate your response. I was a bit heated when I wrote this above.