Original commit message from CVS:
Patch by: Antoine Tremblay <hexa00 at gmail dot com>
* gst-libs/gst/rtp/gstbasertpdepayload.c:
(gst_base_rtp_depayload_finalize), (gst_base_rtp_depayload_push):
Fix some memory leaks: on finalize, free buffers left in the queue
before destroying the queue; in _push(), unref rtp_buf even if
the process vfunc returned a NULL buffer as output buffer (#337548);
demote some recuring debug messages to LOG level.
Original commit message from CVS:
* ext/ogg/gstoggdemux.c: (gst_ogg_pad_src_query),
(gst_ogg_demux_chain_peer), (gst_ogg_pad_submit_packet),
(gst_ogg_chain_free), (gst_ogg_demux_sink_event),
(gst_ogg_demux_loop):
More cleanups.
Respect segment stop when emiting EOS or SEGMENT_DONE.
Fixes (#337945).
Original commit message from CVS:
* tests/check/Makefile.am:
* tests/check/gst-plugins-base.supp:
Suppress an old libtheora bug (fixed in more recent versions), so
that FC4 buildslaves can pass.
Original commit message from CVS:
* ext/ogg/gstoggdemux.c: (gst_ogg_pad_src_query),
(gst_ogg_demux_receive_event), (gst_ogg_pad_event),
(gst_ogg_demux_init), (gst_ogg_demux_finalize),
(gst_ogg_demux_sink_event), (gst_ogg_demux_get_data),
(gst_ogg_demux_loop):
Don't leak events.
Remember what error we got when finding chains, if we
were shutdown, that would not be an error.
Original commit message from CVS:
* gst-libs/gst/audio/gstbaseaudiosink.c:
(gst_base_audio_sink_event):
Starting the ringbuffer when we did not acquire it can cause
a deadlock, is pointless and causes nasty things for
subclasses.
Fixes gst-launch audiotestsrc num-buffers=0 ! alsasink.
Original commit message from CVS:
* ext/theora/theoradec.c: (theora_dec_src_event),
(theora_handle_data_packet):
Some more debug info.
* tests/examples/seek/seek.c: (start_seek), (main):
Print element messages too.
Original commit message from CVS:
* gst/audioresample/debug.h:
replace debug macros with variable number of parameters
by a simple alias to gstreamer standard debug macros
(#define RESAMPLE_ERROR GST_ERROR, __VA_ARGS__ is not
supported by MSVC 6.0 and 7.1)
* gst/audioresample/resample.h:
define M_PI and rint for WIN32
* win32/common/libgstaudio.def:
* win32/common/libgstriff.def:
* win32/common/libgsttag.def:
* win32/common/libgstvideo.def:
add new exported functions
* win32/vs6:
update project files
Original commit message from CVS:
* ext/alsa/gstalsasink.c: (set_hwparams), (alsasink_parse_spec):
More debug to trace why my USB headset is not working with gst
Original commit message from CVS:
* gst/playback/gstplaybasebin.c: (group_destroy):
Clean up our group elements properly in the case where it never
got committed - it still got added unconditionally to the bin.
Original commit message from CVS:
* ext/vorbis/vorbisenc.c: (gst_vorbis_enc_set_header_on_caps),
(gst_vorbis_enc_chain):
Remove leaks from vorbisenc.
Mostly minor changes, the only significant one is that now the
buffers we set as 'streamheader' on the caps are copies of the
original buffers, to avoid circular refcounting problems.
Original commit message from CVS:
* gst/playback/gstplaybasebin.c: (mute_stream), (setup_substreams):
Don't remove our mute-probe if someone else already did so.
Don't set a 2nd one if there is already one pending on the pad.
* gst/playback/gstplaybin.c: (gst_play_bin_send_event_to_sink),
(do_playbin_seek):
When a seek fails, ensure that playbin is still set back to playing.
* gst/typefind/gsttypefindfunctions.c: (mpeg_ts_probe_headers),
(mpeg_ts_type_find), (plugin_init):
Add a typefind function for mpeg-ts streams.
Original commit message from CVS:
2006-04-06 Andy Wingo <wingo@pobox.com>
* gst/videorate/gstvideorate.c (gst_video_rate_reset)
(gst_video_rate_init): Caps-related parameters should not be reset
by a flush -- move their inits to the instance init function.
(gst_video_rate_flush_prev): Don't complain if gst_pad_push
is not OK, just return the result.
* gst/audiotestsrc/gstaudiotestsrc.c
(gst_audio_test_src_class_init)
(gst_audio_test_src_get_times): Re-enable is-live=true, as was
broken by Stefan's commit on 24 March.
Original commit message from CVS:
2006-04-06 Andy Wingo <wingo@pobox.com>
* ext/ogg/gstoggmux.c (gst_ogg_mux_push_buffer): Set caps on
buffers being pushed out. Fixes oggmux ! multifdsink.
Original commit message from CVS:
2006-04-04 Andy Wingo <wingo@pobox.com>
* gst/videorate/gstvideorate.c (gst_video_rate_flush_prev):
Whoops, fix bug introduced. Bad hacker!
Original commit message from CVS:
2006-04-04 Andy Wingo <wingo@pobox.com>
* gst/videorate/gstvideorate.c (gst_video_rate_flush_prev):
Properly handle the case where you get EOS before any buffers are
received. Use gst_buffer_make_metadata_writable where appropriate.
Original commit message from CVS:
2006-04-04 Andy Wingo <wingo@pobox.com>
* ext/theora/theoradec.c (theora_handle_data_packet): This value
is often negative -- make it signed so as not to wrap around.
Fixes segfaults introduced on 9 March.
Original commit message from CVS:
* ext/theora/gsttheoradec.h:
* ext/theora/theoradec.c: (theora_dec_src_event):
Don't try to store a gdouble in a gboolean.
Small cleanups.
Original commit message from CVS:
* ext/ogg/gstoggmux.c: (gst_ogg_mux_queue_pads):
Oggmux sucks.
Make it suck slightly less by writing out the final page.
Still can't encode a vorbis-in-ogg file correctly, though.
Original commit message from CVS:
2006-04-03 Andy Wingo <wingo@pobox.com>
* ext/theora/theora.c (plugin_init): Register theoraparse.
* ext/theora/gsttheoraparse.h:
* ext/theora/theoraparse.c: New files implementing a theora
parser. Now we can properly remux ogg/theora+vorbis, yay.
Original commit message from CVS:
2006-03-31 Andy Wingo <wingo@pobox.com>
* ext/vorbis/vorbisparse.c (vorbis_parse_sink_event): Fix
uninitialized variable return that would happen.
Original commit message from CVS:
2006-03-31 Andy Wingo <wingo@pobox.com>
* ext/vorbis/vorbisparse.c (vorbis_parse_drain_queue): Fix
uninitialized variable return that would never happen.
Original commit message from CVS:
2006-03-31 Andy Wingo <wingo@pobox.com>
* ext/vorbis/vorbisparse.c (gst_vorbis_parse_init)
(vorbis_parse_sink_event): Add an event function to flush our
state on a seek, and to drain buffers on a premature EOS.
(vorbis_parse_push_headers, vorbis_parse_clear_queue)
(vorbis_parse_push_buffer, vorbis_parse_drain_queue_prematurely)
(vorbis_parse_chain, vorbis_parse_queue_buffer)
(vorbis_parse_drain_queue): Queue up buffers until we can set
their timestamps and granulepos values.
* ext/vorbis/vorbisparse.h: Include the vorbis decoder headers,
and keep track of data needed for deriving granulepos and
timestamps for buffers.
Original commit message from CVS:
* tests/examples/seek/seek.c: (end_scrub), (seek_cb), (start_seek),
(stop_seek):
Don't let double and tripple clicks mess up our state.
Original commit message from CVS:
* gst/playback/gstplaybin.c: (gen_video_element),
(gen_text_element), (gen_audio_element), (gen_vis_element):
Error out gracefully when we can't create any of the usual
conversion elements for some reason. Also, don't try to
create an audioscale (sic) element that's not used anyway.
Original commit message from CVS:
* gst/playback/gstplaybasebin.c: (setup_source):
Don't post RESOURCE_NOT_FOUND error when we can't find a source
element for a particular protocol, that's confusing for users.
Instead, post a RESOURCE_FAILED error, so that our own error
message is actually shown in totem etc. (#336303).
Original commit message from CVS:
Patch by: Tommi Myöhänen <ext-tommi dot myohanen at nokia dot com>
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnome_vfs_src_finalize),
(gst_gnome_vfs_src_get_icy_metadata):
Fix some minor memory leaks (#336194).
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfs.c:
(gst_gnome_vfs_location_to_uri_string):
* ext/gnomevfs/gstgnomevfs.h:
* ext/gnomevfs/gstgnomevfssink.c:
(gst_gnome_vfs_sink_set_property):
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnome_vfs_src_set_property):
Make gnomevfssink accept filenames as well as URIs for the
"location" property, just like gnomevfssrc does (and
filesrc/filesink do) (#336190).
Original commit message from CVS:
* tests/check/generic/clock-selection.c: (GST_START_TEST):
set to NULL before unreffing, fixes a valgrind leak.
Why was this not triggering the error that an object needs to
be NULL before unreffing ?
* win32/common/config.h:
update
Original commit message from CVS:
* gst/subparse/gstsubparse.c: (convert_encoding),
(gst_sub_parse_change_state):
* gst/subparse/gstsubparse.h:
Text subtitle files may or may not be UTF-8. If it's not, we
don't really want to see '?' characters in place of non-ASCII
characters like accented characters. So let's assume the input
is UTF-8 until we come across text that is clearly not. If it's
not UTF-8, we don't really know what it is, so try the following:
(a) see whether the GST_SUBTITLE_ENCODING environment variable
is set; if not, check (b) if the current locale encoding is
non-UTF-8 and use that if it is, or (c) assume ISO-8859-15 if
the current locale encoding is UTF-8 and the environment variable
was not set to any particular encoding. Not perfect, but better
than nothing (and better than before, I think) (fixes#172848).
Original commit message from CVS:
2006-03-24 Thomas Vander Stichele <thomas at apestaart dot org>
* configure.ac:
update core requirement to 0.10.4.1 because of async_playback
vmethod on GstBaseSink
