[FFmpeg-devel] [PATCH] lavc/rawdec: Remove superfluous AV_PIX_FMT_MONOWHITE code
Mats Peterson
matsp888 at yahoo.com
Sun Jan 31 13:10:27 CET 2016
On 01/31/2016 01:10 PM, wm4 wrote:
> On Sun, 31 Jan 2016 12:35:20 +0100
> Mats Peterson <matsp888-at-yahoo.com at ffmpeg.org> wrote:
>
>> On 01/31/2016 12:31 PM, Mats Peterson wrote:
>>> On 01/31/2016 12:28 PM, Mats Peterson wrote:
>>>> On 01/31/2016 12:27 PM, Mats Peterson wrote:
>>>>> On 01/31/2016 12:25 PM, Mats Peterson wrote:
>>>>>> On 01/31/2016 12:19 PM, Mats Peterson wrote:
>>>>>>> I've obviously missed that one, Michael, but I don't see the reason to
>>>>>>> switch to monow whatsoever. The space saving of using monow rather
>>>>>>> than
>>>>>>> pal8 for 1 bpp data is rather irrelevant nowadays. Your mileage may
>>>>>>> vary, of course. And in order to be consequent, this would have to be
>>>>>>> done for 1 bpp QuickTime Animation (qtrle) as well.
>>>>>>>
>>>>>>> with your patch applied, this one should of course be ignored.
>>>>>>>
>>>>>>> Mats
>>>>>>>
>>>>>>
>>>>>> What's more, the palette side data is retrieved twice with your patch,
>>>>>> at the beginning and at the end of raw_decode().
>>>>>>
>>>>>> Mats
>>>>>>
>>>>>> _______________________________________________
>>>>>> ffmpeg-devel mailing list
>>>>>> ffmpeg-devel at ffmpeg.org
>>>>>> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>>>>>
>>>>> Do you mind reverting your patch? I don't see why monow would have to be
>>>>> used whatseover.
>>>>>
>>>>> Mats
>>>>>
>>>>
>>>> whatsoever, not whatseover.
>>>>
>>>
>>> I still have the old version of rawdec.c before your patch was applied
>>> here, so I can create a patch that restores it if you want.
>>>
>>> Mats
>>>
>>
>> I'll have to regenerate the FATE test files once again as well, obviously.
>
> "git revert" exists.
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>
I'm creating a new patch here, anyway.
Mats
--
Mats Peterson
http://matsp888.no-ip.org/~mats/
More information about the ffmpeg-devel
mailing list