[FFmpeg-devel] [PATCH] RTSP-MS 15/15: move packet_time_start zero value assignment in asf.c

Michael Niedermayer michaelni
Wed Apr 15 19:07:51 CEST 2009


On Wed, Apr 15, 2009 at 11:26:37AM -0400, Ronald S. Bultje wrote:
> Hi Michael & all,
> 
> On Fri, Apr 3, 2009 at 1:04 PM, Ronald S. Bultje <rsbultje at gmail.com> wrote:
> > OK, ok, I will try to take that approach and send patches for the ASF
> > demuxer that fix it.
> 
> The attached patch is a minimal set required to make use of public API
> possible from the RTP/ASF depacketizer. It contains three hunks (all
> small), which do the following:
> 

> - the third hunk changes the return value on EOF to AVERROR_EOF. This
> isn't necessary, but looks nicer

ok, you can commit that hunk


> - the second hunk omits padsize for asf streams, as I mentioned
> earlier [1] and documented on MSDN [2]

What MSDN says and what your change does are 2 very different things.
first if you feed the demuxer with a header indicating a min_pktsize of
X you MUST pad packet. Or quote some spec / show sample files that
indicate that packets may violate min_pktsize. Or 3rd alternative
dont set a incorrect min_pktsize

also if min_pktsize is a meaningles value then it has no relation to
it being streamed or not

the whole smells a little like h263 in rtp where some people also belive
dumped rtp payloads are a valid h263 stream


> - the first chunk is critical. During ASF packet header parsing, we

the first hunk is possibly redundant if you fix above. If not it is
incorrect, a fix would then involve supporting variable packet size
ASFs which is not achived by an EOF check

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

Dictatorship naturally arises out of democracy, and the most aggravated
form of tyranny and slavery out of the most extreme liberty. -- Plato
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20090415/764f6874/attachment.pgp>



More information about the ffmpeg-devel mailing list