[FFmpeg-devel] [PATCH] Move colorspace enums from avcodec.h to libavcore/colorspace.h

Michael Niedermayer michaelni
Tue Jan 11 01:51:07 CET 2011


On Tue, Jan 11, 2011 at 01:09:18AM +0100, Stefano Sabatini wrote:
> On date Monday 2011-01-10 15:52:29 +0100, Michael Niedermayer encoded:
> > On Sun, Jan 09, 2011 at 09:20:21PM -0800, Larry Robinson wrote:
> > > On 1/9/2011 2:30 PM, Michael Niedermayer wrote:
> > >> On Sun, Jan 09, 2011 at 08:26:25PM +0100, Stefano Sabatini wrote:
> > >>> On date Sunday 2011-01-09 19:42:27 +0100, Michael Niedermayer encoded:
> > >>>> On Sun, Jan 09, 2011 at 12:11:02AM +0100, Stefano Sabatini wrote:
> > >>>>> On date Friday 2011-01-07 11:12:59 -0800, Larry Robinson encoded:
> > [...]
> > >>> compile/install libavfilter without installing libavcodec (configure
> > >>> --enable-libavfilter --disable-libavcodec),
> > >> no problem here
> > >>
> > >>
> > >>> in a distro you could have
> > >>> libavfilter-dev and yet not libavcodec-dev.
> > >> it wouldnt work, we dont have 4 independant configure scripts, and
> > >> quadruplicating the build system across 5 split -dev packages seems like a
> > >> bad idea
> > >> and if you meant the distro simply physically removd libavcodec.
> > >> We should nozt support this castration and if someone still wanted he can
> > >> leave the header there as well if he already manually cuts things randomly
> > >> away
> > >>
> > [...]
> > > 
> > > #ifndef AVCORE_COLORSPACE_H
> > > #define AVCORE_COLORSPACE_H
> > 
> > The enums are not just colorspace related some enums are related to the
> > positions of samples.
> > if at all this could be moved into teh main public header of avcore
> 
> Updated.
> 
> I have a preference for the separate header, I'd prefer to keep
> avcore.h media agnostic (e.g. in an audio application I don't want
> colorspace enums at all).

I have a preferrance of keeping them in avcodec.h. But iam seeing that there
are some philosophic arguments to move them to avcore.h, namely that header
dependancy.
Your patch no longer avoids this header dependancy so there is no argument left
that would favor application of this patch
Also i dont really have the time nor interrest to help you in this bikeshed move
and spliting the enums across several headers is a too heavy maintaince burden
for me considering the lack of any compensating advantage to scatter the closly
related enums


[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Rewriting code that is poorly written but fully understood is good.
Rewriting code that one doesnt understand is a sign that one is less smart
then the original author, trying to rewrite it will not make it better.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20110111/14f257d1/attachment.pgp>



More information about the ffmpeg-devel mailing list