On 24/01/2023 03:28, Takashi Yano via Cygwin-apps wrote:
Thanks!

For a long time, ffmpeg was not included in Fedora, due to concerns
about codec patents, but those issues seem to have been resolved.

Can you please investigate what configuration is used in Fedora, and
what the effects of matching that would be?

Thanks. I will check the configuration of fedora.

I have checked the configuration of ffmpeg in fedora.
It seems that most of codecs are disabled in fedora
by default unless 'all_codecs' flag is specified.
https://src.fedoraproject.org/rpms/ffmpeg/blob/f37/f/ffmpeg.spec

I have built a ffmpeg package using the configuration
almost imitated fedora.
https://tyan0.yr32.net/cygwin/x86_64/release/ffmpeg-free/

You would find this ffmpeg of the above configuration is
almost unuseful...
aac, h264, hevc, mpeg4, wma, wmv, etc. are not supported.

I also imitated the configuration with 'all_codecs' flag
and have built another ffmpeg package.
https://tyan0.yr32.net/cygwin/x86_64/release/ffmpeg-all_codecs/

This is almost common with the major ports of ffmpeg.

Could you please review them?

Thanks for looking into this. The packaging looks fine from a technical standpoint.

The concern here is that we have an informal policy to only accept packages which is would be allowed in Fedora (by it's policies on content and being free of legal encumbrances (e.g. license, patent and trademark issues))

After some discussions, it seems that policy should be formal. I've amended [1] to state that.


I'm sorry to cause you more trouble, but given that, can you package this based on the codec set in Fedora's ffmpeg-free?


If you don't think the package is useful under those constraints, don't let this discourage you from offering either the cygport for interested people to build it themselves, or the packages via an overlay package server [2]

[1] https://cygwin.com/packaging-contributors-guide.html#submitting
[2] https://cygwin.com/package-server.html#overlay

(Note for the peanut gallery: none of this should be construed as reflecting my personal opinions on the virtues and validity of software patents generally, or in this specific case. This is not the place for discussion of such issues)

Reply via email to