On Fri, Jul 27, 2012 at 11:29:02AM +0100, Måns Rullgård wrote:
> Diego Biurrun <[email protected]> writes:
> > ---
> > configure | 16 ++++++++--------
> > 1 files changed, 8 insertions(+), 8 deletions(-)
>
> That's a weird description. Those options are currently user-selectable
> by virtue of being in CONFIG_LIST.
configure: Move parts that should not be user-selectable to CONFIG_EXTRA
> > --- a/configure
> > +++ b/configure
> > @@ -945,12 +944,10 @@ CONFIG_LIST="
> > doc
> > - dwt
>
> I'm uncertain about this one.
>
> > @@ -979,21 +976,16 @@ CONFIG_LIST="
> > libxvid
> > - lsp
>
> This one, like dwt, controls a tool used by multiple codecs. I can
> imagine someone wanting to disable these with a single option instead of
> chasing down all the codecs that use them. Although probably not the
> case for these two, such tools are sometimes optional in a codec (zlib
> is used that way), so disabling the feature can make sense if you know
> you don't need it and want to reduce the size.
Let's see what we currently have:
CONFIG_LIST:
bzlib
dct
fft
mdct
rdft
zlib
CONFIG_EXTRA:
aandcttables
ac3dsp
golomb
h264chroma
h264dsp
h264pred
h264qpel
huffman
lpc
mpegaudiodsp
sinewin
vp3dsp
Quite possibly dwt and lsp make more sense as user-selectable options.
But then what about aandcttables, golomb, huffman, lpc, sinewin?
These would appear to also make sense as user-selectable options.
> > pic
> > rdft
> > - rtpdec
>
> What does this option do?
Apparently it's a shorthand introduced in 44b70ce in order not to repeat
a list of dependencies in
rtpdec_select="asf_demuxer rm_demuxer rtp_protocol mpegts_demuxer mov_demuxer"
rtsp_demuxer_select="http_protocol rtpdec"
sdp_demuxer_select="rtpdec"
Now this does not need a command-line selectable component or any
component at all. We could either introduce a shell variable at
the place where it is used or live with the small duplication.
I don't have a preference in either direction.
Diego
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel