[FFmpeg-cvslog] r17905 - trunk/libavformat/utils.c

[Baptiste Coudurier]

On 3/9/2009 8:36 AM, Reimar D?ffinger wrote:
> On Mon, Mar 09, 2009 at 02:50:08PM +0000, M?ns Rullg?rd wrote:
>> reimar <subversion at mplayerhq.hu> writes:
>>
>>> Author: reimar
>>> Date: Mon Mar  9 13:06:31 2009
>>> New Revision: 17905
>>>
>>> Log:
>>> 100l, to start reading from the first frame we must seek to data_offset,
>>> not 0 in av_seek_frame_generic.
>> This broke seektest.
> 
> Hm. Have any of you ever had "make test" say something like "seektest: no such
> file" and then just go on?
> Anyway, the situation is: seeking did not work at all for swf before,
> and it still does not work at all. The difference is that it now creates
> a new audio and video stream each time (before it did not because
> seeking caused it to read nonsense due to that seek to 0).
> swfdec.c seems to have a bit of a protection against creating multiple
> streams:
>> if (tag == TAG_VIDEOSTREAM && !vst) {
> but this only seems to work as long as there is no data packet
> in-between.
> A cleanup job for the maintainer maybe? :-)

I'll have a look.

> Ok to just update the reference values for the regression tests?
> And also, why do the gxf seek regression tests have DOS linebreaks, I
> thought SVN was supposed to avoid that kind of mess? (at least
> Tortoise-SVN always refuses to commit such messups for me...).

That would be my fault, sorry, I thought it would also :/
Using plain svn in msys here. I'll check

Would it be possible to add a pre-commit check for this ?

-- 
Baptiste COUDURIER                              GnuPG Key Id: 0x5C1ABAAA
Key fingerprint                 8D77134D20CC9220201FC5DB0AC9325C5C1ABAAA
checking for life_signs in -lkenny... no
FFmpeg maintainer                                  http://www.ffmpeg.org