[FFmpeg-devel] [PATCH] Codec lookup: do not use codec_id
Michael Niedermayer
michaelni
Mon Jun 25 14:37:00 CEST 2007
Hi
On Mon, Jun 25, 2007 at 11:27:04AM +0200, Nicolas George wrote:
> Le duodi 2 messidor, an CCXV, Michael Niedermayer a ?crit?:
> > i have a very bad feeling about this, this needs very extensive testing
>
> I agree about the need of extensive testing, although I am pretty sure that
> the particular point we were discussing is ok: if there is a problem, it is
> more likely spread on the whole change of logic.
>
> > stream copy with missing decoder/encoder and multiple input and output files
> > amongth others
>
> I have tested the following:
>
> (a and b are two video files with different codecs)
>
> ./ffmpeg -i $a /tmp/t.avi -newvideo -i $b
>
> -> produces one file with two video streams, mplayer can play both
>
> ./ffmpeg -i $a /tmp/t.avi -newaudio -i $b
>
> -> produces one file with two audio streams, mplayer can play both and
> switch with #.
>
> ./ffmpeg -i $a -acodec mp2 -vcodec mpeg4 /tmp/t.avi \
> -acodec ac3 -vcodec mpeg2video /tmp/t2.avi
>
> -> produces two files with one audio and one video stream each, and with the
> requested codec, mplayer can play both and reports the correct codec
>
> ./ffmpeg -i $b -acodec copy -vcodec copy /tmp/t.avi
>
> -> produces one file very similar to $b
>
> ./ffmpeg -i $a -acodec copy -vcodec copy /tmp/t.avi
>
> -> fails with "Could not write header for output file #0 (incorrect codec
> parameters ?)", but replacing t.avi by t.mov works ($a is a QuickTime
> video), so no problem here.
>
> Are there other tests you think necessary?
yes, does the patch change the md5 of the output in any of the above
cases?
and what happens if you configure --disable-codec-occuring-in-the-file
are the md5s still the same?
steam copy MUST still work even if there is no AVCodec available
do the regression tests pass?
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
Its not that you shouldnt use gotos but rather that you should write
readable code and code with gotos often but not always is less readable
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20070625/3c2ccb2a/attachment.pgp>
More information about the ffmpeg-devel
mailing list