Original commit message from CVS:
* gst/interleave/Makefile.am:
* gst/interleave/deinterleave.h:
* gst/interleave/interleave.h:
* gst/interleave/plugin.h:
Split definitions into separate header files for better documentation
generation.
* gst/interleave/deinterleave.c: (gst_deinterleave_base_init),
(gst_deinterleave_class_init), (gst_deinterleave_sink_setcaps),
(gst_deinterleave_process):
Don't use alloca, allow caps changes as long as the number of channels
does not change, don't use g_warning, return NOT_NEGOTIATED as early
as possible and some other cleanup.
* gst/interleave/interleave.c: (gst_interleave_base_init),
(gst_interleave_class_init):
Do some random cleanup.
* tests/check/Makefile.am:
* tests/check/elements/deinterleave.c: (GST_START_TEST),
(deinterleave_chain_func), (deinterleave_pad_added),
(deinterleave_suite):
Add unit tests for the deinterleave element.
Original commit message from CVS:
* gst/avi/gstavimux.c: (gst_avi_mux_start_file):
Send an initial BYTE segment to inform downstream of later seeking,
and to forego sync attempts.
Original commit message from CVS:
* gst/matroska/matroska-demux.c:
(gst_matroska_demux_push_dvd_clut_change_event):
Convert subtitle palette info in VobSub private data from VobSub's
(buggy) RGB to YUV.
Original commit message from CVS:
* gst/avi/gstavimux.c: (gst_avi_mux_pad_reset):
Do not leave fourcc stream header field empty upon reset.
Fixes#519301.
Original commit message from CVS:
Based on patch by: Wouter Cloetens <wouter at mind be>
* gst/rtsp/gstrtspsrc.c: (gst_rtspsrc_init),
(gst_rtspsrc_skip_lws), (gst_rtspsrc_unskip_lws),
(gst_rtspsrc_skip_commas), (gst_rtspsrc_skip_item),
(gst_rtsp_decode_quoted_string),
(gst_rtspsrc_parse_digest_challenge), (gst_rtspsrc_parse_auth_hdr),
(gst_rtspsrc_setup_auth):
Support Digest authentication. Fixes#532065.
Original commit message from CVS:
* gst/level/gstlevel.c:
Also support 32bit (e.g. whe having it after 'mad'). Add more notes
about whats needed for liboil acceleration. Simplify docs a bit.
Original commit message from CVS:
Patch by: Sjoerd Simons <sjoerd at luon dot net>
* gst/matroska/matroska-mux.c: (gst_matroska_mux_collected):
Update the track duration if the old one was invalid.
Fixes bug #532117.
Original commit message from CVS:
* gst/rtp/gstrtph264pay.c (gst_rtp_h264_pay_parse_sps_pps):
Use GST_STR_NULL when trying to print sps and pps strings that could
be NULL, as this might crash on some platforms.
Original commit message from CVS:
patch by: Haakon Sporsheim <haakon.sporsheim at tandberg com>
* sys/directdraw/gstdirectdrawsink.c (gst_directdraw_sink_setup_ddraw):
Do IDirectDrawClipper_SetHWnd() if the window ID has already been
set after creating the clipper.
Original commit message from CVS:
patch by: Haakon Sporsheim <haakon.sporsheim at tandberg com>
* sys/directdraw/gstdirectdrawsink.c (gst_directdraw_sink_show_frame):
Added checking of surface lost case after an unsuccessful
IDirectDrawSurface7_Lock() call.
If surface is lost, return GST_FLOW_OK.
Original commit message from CVS:
patch by: Haakon Sporsheim <haakon.sporsheim at tandberg com>
* sys/directdraw/gstdirectdrawsink.c (gst_directdraw_sink_show_frame,
WndProc, gst_directdraw_sink_window_thread):
Improved Windows message loop and fixed window destruction issue.
When the window which DirectDraw is rendering to is destroyed, the
render/show_frame function will return GST_FLOW_ERROR.
Partially fixes#520885.
Original commit message from CVS:
patch by: Haakon Sporsheim <haakon.sporsheim at tandberg com>
* sys/directdraw/gstdirectdrawsink.c (gst_directdraw_sink_set_caps):
Fixed mid stream resolution change bug, the offscreen surface is now
released when set_caps is called.
Partially fixes#520885.
Original commit message from CVS:
* sys/directdraw/gstdirectdrawsink.c
(gst_directdraw_sink_buffer_alloc):
Make it so that gst_directdraw_sink_buffer_alloc uses the right
width/height.
Especially when looking through the pool of buffers, make sure that
the width/height of caps is used instead of the already negotiated
dimensions.
For example if a buffer with different caps is requested, i.e.
higher resolution, the caller would get a buffer with the old
dimensions and thus corrupt the heap.
Original commit message from CVS:
* sys/directdraw/gstdirectdrawsink.c
(gst_directdraw_sink_buffer_alloc):
Clear the flags on recycled buffers from buffer_alloc.
Partially fixes#520885.
The right fix this time.
Original commit message from CVS:
* sys/directdraw/gstdirectdrawsink.c
(gst_directdraw_sink_buffer_alloc):
Reverting previous commit, it had it all mixed up, was for a different
patch (major automation screw-up). Sorry!
Original commit message from CVS:
* sys/directdraw/gstdirectdrawsink.c
(gst_directdraw_sink_buffer_alloc):
Clear the flags on recycled buffers from buffer_alloc.
Partially fixes#520885.
Original commit message from CVS:
* gst/rtp/gstrtpilbcpay.c:
Added missing stdlib.h include for strtol(), and made include ordering and
style consistent with the corresponding depayloader.
Original commit message from CVS:
* gst/rtp/gstrtpilbcpay.c:
Added missing stdlib.h include for strtol(), and made include ordering and
style consistent with the corresponding depayloader.
Original commit message from CVS:
* ext/gconf/gstgconfaudiosrc.c: (gst_gconf_audio_src_reset),
(gst_gconf_audio_src_change_state):
* ext/gconf/gstgconfvideosink.c: (gst_gconf_video_sink_reset),
(gst_gconf_video_sink_change_state):
* ext/gconf/gstgconfvideosrc.c: (gst_gconf_video_src_reset),
(gst_gconf_video_src_change_state):
* ext/gconf/gstswitchsink.c: (gst_switch_sink_reset),
(gst_switch_commit_new_kid), (gst_switch_sink_change_state):
When we can't create a fakesink/fakesrc complain instead of unreffing
NULL pointers and crashing later. See bug #530535.
Original commit message from CVS:
* gst/rtp/gstrtph263pdepay.c: (gst_rtp_h263p_depay_process):
Add some more debug info and guard against small payloads.
* gst/rtp/gstrtppcmudepay.c: (gst_rtp_pcmu_depay_process):
Set duration on outgoing buffers because we can.
Original commit message from CVS:
Patch by: Olivier Crete <tester at tester dot ca>
* ext/speex/gstspeexenc.c: (gst_speex_enc_sink_getcaps),
(gst_speex_enc_init), (gst_speex_enc_chain):
Add negotiation for the speex channels and rate. Fixes#465146.
Original commit message from CVS:
Patch by: Olivier Crete <tester at tester dot ca>
* gst/rtp/gstrtpspeexpay.c: (gst_rtp_speex_pay_class_init),
(gst_rtp_speex_pay_getcaps):
Add negotiation for the speec channels and rate. See #465146.
Original commit message from CVS:
Patch by: Olivier Crete <tester at tester dot ca>
* gst/rtp/gstrtpilbcpay.c: (gst_rtpilbcpay_class_init),
(gst_rtpilbcpay_sink_setcaps), (gst_rtpilbcpay_sink_getcaps):
Add negotiation for the ILBC mode. See #465146.
Original commit message from CVS:
Patch by: Youness Alaoui <youness.alaoui at collabora co uk>
* gst/udp/gstudpsrc.c: (gst_udpsrc_create):
Don't error out if we get an ICMP destination-unreachable
message when trying to read packets on win32 (#529454).
Original commit message from CVS:
* configure.ac:
* gst/qtdemux/qtdemux.c: (qtdemux_parse_trak):
Use new error code for encrypted streams (which requires core CVS).
Original commit message from CVS:
* gst/qtdemux/qtdemux.c: (gst_qtdemux_videosrc_template),
(gst_qtdemux_audiosrc_template):
Fix swapped pad template names, spotted by Thiago Sousa Santos.
Original commit message from CVS:
* ext/speex/gstspeexdec.c: (speex_dec_sink_event),
(speex_dec_chain_parse_data):
Produce concealment data when time progresses in a segment update.
Original commit message from CVS:
* ext/speex/gstspeexdec.c: (speex_dec_chain_parse_data),
(speex_dec_chain):
Try to preserve input timestamps when we can.
Do beginnings of error concealment.
Original commit message from CVS:
2008-04-28 Julien Moutte <julien@fluendo.com>
* gst/debug/rndbuffersize.c: (gst_rnd_buffer_size_loop): Fix printf
format to pacify Mac OSX's gcc.
Original commit message from CVS:
* gst/debug/rndbuffersize.c: (DEFAULT_SEED), (DEFAULT_MIN),
(DEFAULT_MAX), (src_template), (sink_template),
(gst_rnd_buffer_size_base_init), (gst_rnd_buffer_size_class_init),
(gst_rnd_buffer_size_init), (gst_rnd_buffer_size_activate),
(gst_rnd_buffer_size_loop), (gst_rnd_buffer_size_plugin_init):
Bring rndbuffersize element into a state that doesn't require us
to move it to -bad immediately. For one, fix up default min/max
values so that the element actuall works using the default values.
Also, don't ignore flow return values and do some kind of minimal
eos logic. Allow min=max to pull fixed-sized buffers. Bunch of
other gratuitious clean-ups.
Original commit message from CVS:
* gst/rtp/gstasteriskh263.c: (gst_asteriskh263_chain):
* gst/udp/gstmultiudpsink.c: (gst_multiudpsink_add_internal):
* gst/udp/gstudpsrc.c: (gst_udpsrc_start):
Use GLib versions of htonl, htons, ntohl and ntohs in order
to avoid problems on win32 (#529707).
Original commit message from CVS:
Patch by: Jesús Corrius <jesus at softcatala org>
* gst/goom/filters.c: (zoomVector):
* gst/goom/goom_core.c: (init_buffers):
Fix build with mingw32: use rand() instead of random() and
replace bzero() with memset(). Fixes#529692.
Original commit message from CVS:
* tests/check/elements/multifile.c:
Include stdlib.h and unistd.h for mkdtemp. Some platforms have it
declared in the former, some have it declared in the latter.
