Anton Khirnov <[email protected]> writes:

> On Wed, 12 Oct 2011 21:55:58 +0100, Måns Rullgård <[email protected]> wrote:
>> Anton Khirnov <[email protected]> writes:
>> 
>> > On Wed, 12 Oct 2011 21:41:52 +0100, Måns Rullgård <[email protected]> wrote:
>> >> Jindřich Makovička <[email protected]> writes:
>> >> 
>> >> > Hi,
>> >> >
>> >> > when including avutil/log.h in the code compiled as C++, g++ 4.6
>> >> 
>> >> Libav is C, not C++.
>> >> 
>> >
>> > Which doesn't mean it should be unusable in C++ code.
>> 
>> Will you accept a patch turning the headers into C/Fortran polyglots too?
>
> This is different, keeping the headers compatible with C++ requires
> very little effort.

C/Fortran polyglots are surprisingly easy too.  C/Java is much harder.

>> > At the very least it fixes some warnings when building.
>> 
>> What warnings?
>> 
>
> CC    libavutil/opt.o
> libavutil/opt.c: In function ‘av_opt_child_class_next’:
> libavutil/opt.c:793:9: warning: passing argument 1 of 
> ‘parent->child_class_next’ from incompatible pointer type [enabled by default]
> libavutil/opt.c:793:9: note: expected ‘const struct AVClass *’ but argument 
> is of type ‘const struct AVClass *’
> libavutil/opt.c:793:9: warning: return from incompatible pointer type 
> [enabled by default]
>
> In fact i was just about to send an identical patch myself.

Oh, someone broke it just minutes ago.  Then by all means fix it, but
with an accurate log message.

-- 
Måns Rullgård
[email protected]
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel

Reply via email to