Mailing List Archive

On Thu, 2021-12-02 at 18:58 +0100, Klaas de Waal wrote:
> A better strategy is to contribute new?features?to FFmpeg and then to
> make use of these features in MythTV.

One possible compromise approach might be to insist that there has at
least been a reasonable initial submission of the patches to ffmpeg
upstream which hasn't received a hard "no way we will ever take this" -
- even if there might be significant rework+time required for it to
actually be merged and released -- before considering it acceptable to
apply to the myth downstream.

That way at least you know mythtv isn't taking on anything which is
never going to be upstreamed i.e. going to be an eternal downstream
burden, plus it nudges developers towards the desired "upstream first"
mentality without holding downstream completely to the upstream
development timeframes (assuming there is a disconnect there).

Ian.

_______________________________________________
mythtv-dev mailing list
mythtv-dev@mythtv.org
http://lists.mythtv.org/mailman/listinfo/mythtv-dev
http://wiki.mythtv.org/Mailing_List_etiquette
MythTV Forums: https://forum.mythtv.org

> Wiadomo?? napisana przez Ian Campbell <ijc@hellion.org.uk> w dniu 02.12.2021, o godz. 22:50:
>
> On Thu, 2021-12-02 at 18:58 +0100, Klaas de Waal wrote:
>> A better strategy is to contribute new features to FFmpeg and then to
>> make use of these features in MythTV.
>
> One possible compromise approach might be to insist that there has at
> least been a reasonable initial submission of the patches to ffmpeg
> upstream which hasn't received a hard "no way we will ever take this" -

kernel side:
kernel uAPI example: https://lkml.org/lkml/2020/11/12/487
it is required as ffmpeg is conusmer & will never - even look - at code using private kernel API.
So far public are: mpg2, h264, vp8. very soon will be vp9 & hevc

ffmpeg side:
h264 example https://patchwork.ffmpeg.org/project/ffmpeg/list/?series=2898
Jonas is currently much less active (covid or new work?)

v4l2_request hw decode is official committed long-tern strategy for linux kernel hw.accell access on embedded/SoC ecosystem.
so - from future point of view - it is rather question of _when?_ ffmpeg will officially support it - not question of _will support?_

Guys,

I see there is mood not going with v4l2_request code at the moment.

No problem - we can hold this now and see in future....

br

_______________________________________________
mythtv-dev mailing list
mythtv-dev@mythtv.org
http://lists.mythtv.org/mailman/listinfo/mythtv-dev
http://wiki.mythtv.org/Mailing_List_etiquette
MythTV Forums: https://forum.mythtv.org