Closes #21 (partly, at least), #4288 and #4304

Description
===================
Fixed bugs in AVI mux to support creation of AVI files with following popular video/audio combinations
(a) MPEG4 video (aka XVID) with MP3/AC3/PCM Audio
(b) MJPEG video with PCM audio.
AVI as such is not used for storing other formats like H.264, AAC etc

Bug fix details
===================
(a) 'hdrl' list size was wrong. It incorrectly included some parts of MOVI list, size taken by hdrl LIST tag and its size and Size of 'RIFF', File size and 'AVI ' type.
(b) Block Align was set to 0 in case of MP3/AC3. It has to be non-zero. It has to been set to 1.
(c) XVID video decoders expect header to present as part of first frame. Added code to append header present at end of BITMAPINFOHEADER to beginning of the first video frame in case of XVID video.
(d) nAvgBytesPerSec of Waveformatex was not properly being set in case of PCM audio.
(e) Last index entry value was set to garbage due to incorrect index size calculation.

Test PC configuration
===================
Tested with Windows Media Player 11 installed on Windows XP Service pack 3 machine with following directshow filters installed
(a) XVID MPEG4 video decoder filter for MPEG4 video decoding
(b) LAME MP3 codec for MP3 audio decoding

Combinations tested
====================
(a) MPEG4/MP3 transcode
vlc -I dummy "Mr_MrsSmith-h264_aac.mp4" :sout="#transcode{vcodec=mp4v,vb=1024,fps=25,acodec=mpga,ab=128}:standard{mux=avi,access=file,dst=out.avi}" vlc://quit

(b) MPEG4/AC3 transcode
vlc -I dummy "Mr_MrsSmith-h264_aac.mp4" :sout="#transcode{vcodec=mp4v,vb=1024,fps=25,acodec=a52, ab=128}:standard{mux=avi,access=file,dst=out.avi}" vlc://quit

(c) MPEG4/PCM(16bit) transcode
vlc -I dummy "Mr_MrsSmith-h264_aac.mp4" :sout="#transcode{vcodec=mp4v,vb=1024,fps=25,acodec=s16l}:standard{mux=avi,access=file,dst=out.avi}" vlc://quit

(d) MPEG4/PCM(24 bit) transcode
vlc -I dummy "Mr_MrsSmith-h264_aac.mp4" :sout="#transcode{vcodec=mp4v,vb=1024,fps=25,acodec=s24l}:standard{mux=avi,access=file,dst=out.avi}" vlc://quit

(e) MPEG4/PCM(32bit) transcode
vlc -I dummy "Mr_MrsSmith-h264_aac.mp4" :sout="#transcode{vcodec=mp4v,vb=1024,fps=25,acodec=s32l}:standard{mux=avi,access=file,dst=out.avi}" vlc://quit

(f) MJPEG/PCM(32bit) transcode
vlc -I dummy "Mr_MrsSmith-h264_aac.mp4" :sout="#transcode{vcodec=mjpeg,vb=1024,fps=25,acodec=s32l}:standard{mux=avi,access=file,dst=out.avi}" vlc://quitSigned-off-by: Jean-Baptiste Kempf <jb@videolan.org>
(cherry picked from commit 1e6d1f3cb3d0ee4b36a939b17eb3a1de2131d34d)
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

(cherry picked from commit 73cadb239d790835d81a83e53069fbe768a2d260)
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Win32 spawns its own console so this is irrelevant
(cherry picked from commit f1433011c2cae9e25cfcb56d64cc92b2f523df4e)
Signed-off-by: Pierre Ynard <linkfanel@yahoo.fr>

(cherry picked from commit 8ac466488795c555fb32978509eb62df40ad57f0)

(cherry picked from commit 2562696be3f12bf080dea6fa6a0b690bbd761e30)

(cherry picked from commit da59205e44a1d105ad020cd9a94f8e37e6a012c7)

Blind commit

Signed-off-by: Christophe Mutricy <xtophe@chewa.net>

(See also previous commit)
(cherry picked from commit cf190075bda0c39e2a203104d9095c450e283def)

As far as I understand, this breaks C++ object scoping. qtu() creates
a temporary anonymous QByteArray from its QString parameter. That array
is destroyed at the end of the expression, thus the return value of
qtu() becomes a dangling pointer.
(cherry picked from commit 41c3dd9b229c44cb313786f96cc2089a9d5197cc)

(cherry picked from commit 00eb305f2afebf28dd649d507aca4901d0ad1e8a)

Signed-off-by: Christophe Mutricy <xtophe@chewa.net>

This caused duplciate strings in PO

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>
Signed-off-by: Felix Paul Kühne <fkuehne@videolan.org>

Pointed-out-by: Fabian Greffrath <fabian@greffrath.com> (Debian BTS)
(cherry picked from commit c5e166c05a49cacbabc52a32ee612533e909e882)

Conflicts:

	share/vlc.desktop

macosx: fixed disappearing repeat button images introduced in [8f79caa3]

Closes #4886

Unlike _endthread and ExitThread, _endthreadex doesn't close the handle.
It is therefore up to the caller to close it. Hence no need for an extra
handle duplication (that mostly amounted to a memory leak when the normal
 path was used, i.e normal return from thread with implicit _endthreadex)
(cherry picked from commit afe774e0b19c7dcad8d0ae248480df1b0660a380)
Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>

According to msdn doc, _endthreadex(0) is to be used when thread is
 launched with _beginthreadex().

Note that when returning from the normal path, an implicit _endthreadex is
issued anyway.  It is important to be consistent whatever the path, since
those final functions (_endthread, ExitThread, _endthreadex) don't have
the same deallocation policy.
(cherry picked from commit 99f71e5a62fdd14b031fdc89f452d873e3509c74)
Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>
(cherry picked from commit 4fc3bd15a3844e41a75760fa8e100feecea969a5)
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Signed-off-by: Christophe Mutricy <xtophe@chewa.net>

Not working

This reverts commit f7a1db99.

macosx: fixed volume normalizer prefs and crash when saving the Growl option in combination with some config sets

(cherry picked from commit b38b9c24c20313bfa3fa76fe7dffa6177edd270e)