[FFmpeg-devel] [PATCH 3/4] ffprobe: stop printing deprecated fields

[Marton Balint]

On Wed, 27 Jan 2021, James Almer wrote:

> On 1/27/2021 7:42 PM, Michael Niedermayer wrote:
>> On Tue, Jan 26, 2021 at 06:01:01PM +0100, Anton Khirnov wrote:
>>> Also drop the sections guarded by #if FF_API*
>>> These macros are private and should not be used by external callers.
>>> ---
>>>   fftools/ffprobe.c                             |  34 --
>>>   .../ref/fate/concat-demuxer-extended-lavf-mxf |   2 +-
>>>   .../fate/concat-demuxer-extended-lavf-mxf_d10 |   2 +-
>>>   .../ref/fate/concat-demuxer-simple1-lavf-mxf  | 248 +++++-----
>>>   .../fate/concat-demuxer-simple1-lavf-mxf_d10  | 144 +++---
>>>   tests/ref/fate/concat-demuxer-simple2-lavf-ts | 302 ++++++------
>>>   tests/ref/fate/ffprobe_compact                |  34 +-
>>>   tests/ref/fate/ffprobe_csv                    |  34 +-
>>>   tests/ref/fate/ffprobe_default                |  42 --
>>>   tests/ref/fate/ffprobe_flat                   |  42 --
>>>   tests/ref/fate/ffprobe_ini                    |  42 --
>>>   tests/ref/fate/ffprobe_json                   |   9 -
>>>   tests/ref/fate/ffprobe_xml                    |   6 +-
>>>   tests/ref/fate/flcl1905                       |  34 +-
>>>   ...hapqa-extract-nosnappy-to-hapalphaonly-mov |   8 -
>>>   .../fate/hapqa-extract-nosnappy-to-hapq-mov   |   8 -
>>>   tests/ref/fate/hls-fmp4_ac3                   |   1 -
>>>   tests/ref/fate/mov-aac-2048-priming           | 434 +++++++++---------
>>>   tests/ref/fate/mov-init-nonkeyframe           | 240 +++++-----
>>>   tests/ref/fate/mov-zombie                     | 134 +++---
>>>   tests/ref/fate/mxf-probe-applehdr10           |  10 -
>>>   tests/ref/fate/mxf-probe-d10                  |   8 -
>>>   tests/ref/fate/mxf-probe-dnxhd                |   9 -
>>>   tests/ref/fate/mxf-probe-dv25                 |  10 -
>>>   24 files changed, 807 insertions(+), 1030 deletions(-)
>> 
>> This decreases the amount of fields regression tested.
>> Iam not sure if thats a good idea. Generally more testing is better
>
> These are all deprecated fields, so they will be gone with the bump.

I thought so too, but on a second look ist->dec_ctx is the codec used at 
avcodec_open, so that still can be used, and the deprecation #ifs were put 
there as a mistake or historically. Overall the way I see it these fields 
can and should still be provided:

   print_int("coded_width",  dec_ctx->coded_width);
   print_int("coded_height", dec_ctx->coded_height);
   print_int("closed_captions", !!(dec_ctx->properties & FF_CODEC_PROPERTY_CLOSED_CAPTIONS));
   print_q("codec_time_base", dec_ctx->time_base, '/');

And if "max_bit_rate" is converted to use dec_ctx instead of stream->codec 
then probably that can stay too.

So only timecode, convergence_duration and the pseudopal flag should be 
removed, because there is no replacement for those.

Regards,
Marton


> This patch is doing what you mentioned was the correct approach 
> regarding deprecation removals, which was removing FF_API defines 
> individually or in small batches per patch, instead of bumping 
> LIB*_VERSION_MAJOR and disabling them all in one go.
>
> If you prefer, what could be done is wait until ffmpeg 4.4 is branched 
> out before applying this patch (and probably also 4/4), which i assume 
> was planned to happen before the bump so distros can get all of last 
> year's development in their current releases (mainly LTSs) simply by 
> dropping in the new libraries.
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel at ffmpeg.org
> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>
> To unsubscribe, visit link above, or email
> ffmpeg-devel-request at ffmpeg.org with subject "unsubscribe".