[FFmpeg-devel] [PATCH] Allow linking to an external libpostproc

[Michael Niedermayer]

Hi everyone

On Mon, Oct 28, 2024 at 01:24:48AM +0100, Michael Niedermayer wrote:
> Sponsored-by: Sovereign Tech Fund
> Signed-off-by: Michael Niedermayer <michael at niedermayer.cc>
> ---
>  Makefile             |  3 +--
>  configure            | 18 ++++++++----------
>  fftools/ffprobe.c    |  1 -
>  fftools/opt_common.c |  1 -
>  4 files changed, 9 insertions(+), 14 deletions(-)

the other side of this is here:

https://github.com/michaelni/libpostproc

it should build and install and link
but expect bits from ffmpeg remaining in it (most not intentional, some maybe are)

It contains libavutil because after decades of API rewrites, we still
are not capable to write working and usable APIs.
This is not a issue of libpostproc its an issue of libavutil API
which is not treated like a library with a public API, but like a
random subdirectory that randomly gets used by other bits of code.

Ranting a little more, i do have to mention that libavutil and
libavcore or what it was called, did split a general library and the
multimedia parts apart. That was still a better design than now.
where i need to leave a copy of libavutil in the split out libpostproc
and no the system installed libavutil doesnt even build
as it tries to include headers that are not even installed with
HAVE_AV_CONFIG_H

ok one last rant bit,
a library has a public API
and NOTHING outside the library should use anything else
there should not be headers or features enabled in ffmpegs tree
that can be used by other libs. That is wrong.
If its usefull to us it is usefull to others, it should be a
proper public API or it should not be there at all

rant end

Ahh and forks and pull requests of libpostproc are very welcome!

and if you want to improve libpostproc, up to the point of making
it a full replacement of a full filter framework, thats IS welcome

thx

[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Awnsering whenever a program halts or runs forever is
On a turing machine, in general impossible (turings halting problem).
On any real computer, always possible as a real computer has a finite number
of states N, and will either halt in less than N cycles or never halt.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 195 bytes
Desc: not available
URL: <https://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20241028/c26ed58b/attachment.sig>