[FFmpeg-devel] Project orientation

[Marton Balint]

On Sun, 5 Jul 2020, Kieran Kunhya wrote:

>> x264 is practically feature complete, but x262 still miss some things,
>> like 4:2:2 interlaced. Sure, x262 can work well enough for some use cases,
>> but it is still not packaged in e.g. Ubuntu, so users are stuck with the
>> - in some ways - inferior mpeg2 encoder of ffmpeg.
>>
>> The point I am trying to make is that you and some other people made a
>> fast and modern mpeg2 encoder, in some ways superior to existing open
>> source alternatives, but very few people is using it because it was not
>> merged to a bigger/more popular project like x264 or ffmpeg. So it
>> receives no mainteance, no distribution support, no user base and
>> ultimately no further development. Or at least that is how I see it.
>
> People aren't using it because people don't use MPEG-2.
> There is no maintenance to be done on a format that's 25 years old, do
> you want me to randomly change cosmetics to make you feel happy?

If you'd get off your high horse for once, that would make me feel happy.

> I know of people using it 24/7 for many years and have had no issues.
>
> Have you opened bug reports on x264-devel about the issues you see?
> MPEG-2 4:2:2 interlaced users are infinitesimally small and professional users.
> If you want your pet feature to work, lo and behold you have to work
> on it or pay someone to work on it.
> If you want people to work on it out of nowhere then you are part of
> the Free Rider Problem:
> https://en.wikipedia.org/wiki/Free-rider_problem

I don't think I told anybody to implement the missing features for free. 
But I do believe that payed or free development has a higher chance of 
happening if existing code is already available in a popular 
package/project. And yes, I believe that some "maintenance burden" should 
be accepted by the base project to give more chance to further 
advancement, payed or free.

Regards,
Marton