xine-lib/src/xine-engine/buffer.h, branch 1-rc3

Update win32 MSVC build:

2003-12-13T11:35:07+00:00

  * internal names cleanup
  * http input plugin uses http_helper.c
  * fix change in audio_directx
  * MSVC stoped compile xineui because of 'this', mysterious
  * change non-constand array lengths to malloc/free,
    it good for such old compilers as my MSVC 6.0
  * change lprintf macro for MSVC to compile again
    (I don't know, how use variable number of macro arguments  in MSVC...)
  * merge compiling of xine-utils and xine-engine

CVS patchset: 5902
CVS date: 2003/12/13 11:35:07

"Anti audio fifo null" patch ;)

2003-11-20T00:42:14+00:00

Makes things a bit more symetric:
- the audio fifo (demuxer->decoder) can't be NULL
- the video driver can be NULL
A dummy fifo (5 buffers) is used if the audio driver or video driver is NULL.
Audio frontends (rhythmbox, seedeexeen, quark) do not need to instanciate a video driver anymore.

CVS patchset: 5767
CVS date: 2003/11/20 00:42:14

buffer type for 4X ADPCM

2003-11-16T14:29:29+00:00

CVS patchset: 5746
CVS date: 2003/11/16 14:29:29

start xine_stream_t separation (public/private).

2003-11-15T14:54:30+00:00

(we should finish this before rc3.)
- stream_info and meta_info variables are private now.
obs: everything must be recompiled due xine_stream_t changes

CVS patchset: 5733
CVS date: 2003/11/15 14:54:30

rename internal API function (_x_).

2003-11-11T18:44:50+00:00

CVS patchset: 5721
CVS date: 2003/11/11 18:44:50

obsolete a special buffer type (fewer special cases is better)

2003-10-24T02:59:03+00:00

CVS patchset: 5583
CVS date: 2003/10/24 02:59:03

Add a callback to the buffer_pool_alloc function.

2003-10-14T22:16:32+00:00

The goal is to detect in the net_buf_ctrl if there is no more free buffer in a fifo, which indicates there will be a deadlock if the engine is not unpaused.
Currently this detection is done from the "put" callback, it's not a safe solution, nothing prevents a demuxer to allocate 2 buffers before calling put().
The impact is low, only one test per buffer_pool_alloc call if no callback is registered.
It's done exactly the same way as the put and get callbacks.

CVS patchset: 5520
CVS date: 2003/10/14 22:16:32

clarify comment

2003-08-05T15:20:00+00:00

CVS patchset: 5252
CVS date: 2003/08/05 15:20:00

Patch from Jeroen Asselman:

2003-06-16T16:42:51+00:00

- fixes mimetype in realaudio demuxer
- big improvement of v4l input and associated
  demuxer. Including sound capture using alsa
  and a/v sync. Now radio is supported as well.

CVS patchset: 5054
CVS date: 2003/06/16 16:42:51

added placeholder fourccs for On2 VPx video

2003-05-31T02:18:01+00:00

CVS patchset: 4990
CVS date: 2003/05/31 02:18:01