[FFmpeg-devel] ffmpeg.c/cmdutils.c compilation

[Benoit Fouet]

Michael Niedermayer wrote:
> Hi
>
> On Mon, Jun 25, 2007 at 10:01:52AM -0400, Ronald S. Bultje wrote:
>   
>> Hi,
>>
>> On 6/25/07, Michael Niedermayer <michaelni at gmx.at> wrote:
>>     
>>> On Mon, Jun 25, 2007 at 09:18:23AM -0400, Ronald S. Bultje wrote:
>>>       
>>>> On 6/24/07, Michael Niedermayer <michaelni at gmx.at> wrote:
>>>>         
>>>>> On Sun, Jun 24, 2007 at 11:23:10AM -0400, Ronald S. Bultje wrote:
>>>>>           
>>>>>> On 6/22/07, Michael Niedermayer <michaelni at gmx.at> wrote:
>>>>>>             
>>>>>>> On Fri, Jun 22, 2007 at 09:15:30PM -0400, Ronald S. Bultje wrote:
>>>>>>>               
>>>>>>>> recent changes don't allow those to be compiled w/o
>>>>>>>>                 
>>> HAVE_AV_CONFIG_H
>>>       
>>>>>>>> anymore. Attached patch fixes half of it, although av_abort() and
>>>>>>>> closesocket() need some dirt. Should I no longer try to do them
>>>>>>>>                 
>>> w/o
>>>       
>>>>>>>> HAVE_AV_CONFIG_H? That's bad, since they're examples for
>>>>>>>>                 
>>>>> applications,
>>>>>           
>>>>>>>> and other applications would need to do w/o HAVE_AV_CONFIG_H
>>>>>>>>                 
>>> also.
>>>       
>>>>>>> HAVE_AV_CONFIG_H should not be set when compiling examples, someone
>>>>>>>               
>>>>> should
>>>>>           
>>>>>>> change svn if needed so HAVE_AV_CONFIG_H isnt set for examples ...
>>>>>>>               
>>>>>> So what's the preferred way of doing this?
>>>>>>
>>>>>> - change source code to not use internal av_*() functions (e.g.
>>>>>>             
>>>>> av_abort())
>>>>>           
>>>>>> which are protected by HAVE_AV_CONFIG_H, but use system functions (
>>>>>>             
>>> e.g.
>>>       
>>>>>> abort()) instead
>>>>>>             
>>>> [..]
>>>>
>>>>         
>>>>> s/av_abort/abort/ in these test/example applications
>>>>>           
>>>> Please see attached.
>>>>
>>>> - config.h is needed in ffmpeg.c b/c of FFMPEG_CONFIGURATION and
>>>> ENABLE_SWSCALER macros.
>>>>         
>>> hmm why ENABLE_SWSCALER? it should work fine if s/ENABLE_SWSCALER/1/
>>>       
>> Dunno, but it indeed looks kind of silly. Patch for that attached also.
>>     
>
> looks ok (assuming someone did test it without sws)
>
> [...]
>   
>   

now it's ok :)

Applied

Ben
-- 
Purple Labs S.A.
www.purplelabs.com