Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Previously nut used the fallback of estimation by bitrate.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Possibly improve readability.

Links can be set up to group samples into buffers of
specified minimum and maximum size.

It broke twice in the last three months due to other changes.

Identical to commit 4f6a1c97 by Reimar Döffinger,
that was lost after the merge of 19ad5673.

For consistency only.

* qatar/master:
  file: Only include unistd.h if it exists
  random_seed: Only read /dev/*random if we have unistd.h
  doc: Indicate that RTMPT is natively implemented in libavformat
  rtpdec: Don't explicitly include unistd.h any longer

Conflicts:
	configure
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Possibly improve readability.

Possibly improve readability.

Avoid confusing and pointless double assignment of variable
resampled_data_size.

Libav did not exist in 2008 thus this file cannot have originated from there
Reviewed-by: Paul B Mahol <onemda@gmail.com>

Libav did not exist in 2010 thus the file cannot originate from there
Reviewed-by: Paul B Mahol <onemda@gmail.com>

common wma code existed long before Libav
Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

This is needed for SAMI and RealText demuxers.

The "Default" style written in the header is ignored unless you explicit
it in the Dialogue events (it was valid, just ignored). This requires an
update of the SubRip test since the ASS output obviously changes.

Consistent with the rest of the file, less confusing.

This simplifies usage for segment streaming formats with no global
headers, tipically MPEG 2 transport stream "ts" files.

The seg class duplication is required in order to avoid an infinite loop
in libavformat/utils.c:format_child_next_class().

Consistent with av_find_input_format().

Patch is addition to my previous patch
(https://lists.ffmpeg.org/pipermail/ffmpeg-cvslog/2012-June/051590.html)
and disables seeking on FIFOs/named pipes by setting
URLContext::is_streamed (same as pipe: protocol does for stdin/stdout pipes)

Fixes Ticket986
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>