platform/upstream/gst-plugins-good.git
13 years agogst-indent
Thomas Vander Stichele [Sun, 14 Mar 2004 22:34:33 +0000 (22:34 +0000)]
gst-indent

Original commit message from CVS:
gst-indent

13 years agoRemove all usage of gst_pad_get_caps(), and replace it with gst_pad_get_allowed_caps...
David Schleef [Mon, 12 Jan 2004 03:40:18 +0000 (03:40 +0000)]
Remove all usage of gst_pad_get_caps(), and replace it with gst_pad_get_allowed_caps() or gst_pad_get_negotiated_cap().

Original commit message from CVS:
Remove all usage of gst_pad_get_caps(), and replace it with
gst_pad_get_allowed_caps() or gst_pad_get_negotiated_cap().

13 years agoMerge CAPS branch
David Schleef [Mon, 22 Dec 2003 01:47:09 +0000 (01:47 +0000)]
Merge CAPS branch

Original commit message from CVS:
Merge CAPS branch

13 years agoremoved GST_*_CAST. Disabling of type checking is done in glib.
Benjamin Otte [Sat, 13 Dec 2003 16:59:51 +0000 (16:59 +0000)]
removed GST_*_CAST. Disabling of type checking is done in glib.

Original commit message from CVS:
removed GST_*_CAST. Disabling of type checking is done in glib.

13 years agoremove copyright field from plugins
Andy Wingo [Thu, 4 Dec 2003 10:37:38 +0000 (10:37 +0000)]
remove copyright field from plugins

Original commit message from CVS:
remove copyright field from plugins

13 years agoRemove all config.h includes from header files, add it to each source file and remove...
Ronald S. Bultje [Fri, 7 Nov 2003 12:47:02 +0000 (12:47 +0000)]
Remove all config.h includes from header files, add it to each source file and remove duplicate config.h includes fro...

Original commit message from CVS:
Remove all config.h includes from header files, add it to each source file and remove duplicate config.h includes from several source files

13 years agoJack fixed too
Iain Holmes [Sat, 1 Nov 2003 23:43:13 +0000 (23:43 +0000)]
Jack fixed too

Original commit message from CVS:
Jack fixed too

13 years agochange gst/bytestream.h to gst/bytestream/bytestream.h
David Schleef [Wed, 29 Oct 2003 03:15:55 +0000 (03:15 +0000)]
change gst/bytestream.h to gst/bytestream/bytestream.h

Original commit message from CVS:
change gst/bytestream.h to gst/bytestream/bytestream.h

13 years agomerge TYPEFIND branch. Major changes:
Benjamin Otte [Tue, 28 Oct 2003 20:52:41 +0000 (20:52 +0000)]
merge TYPEFIND branch. Major changes:

Original commit message from CVS:
merge TYPEFIND branch. Major changes:
- totally reworked type(find) system
- all typefind functions are in gst/typefind now
- more typefind functions then before
- some plugins might fail to compile now because I don't have them installed and they
a) require bytestream or
b) haven't had their typefind fixed.
Please fix those plugins and put the typefind functions into gst/typefind if they don't have dependencies

13 years ago/GstBuffer/GstData/ in the API where you can pass events. Fix the plugins to deal...
Andy Wingo [Wed, 8 Oct 2003 16:08:19 +0000 (16:08 +0000)]
/GstBuffer/GstData/ in the API where you can pass events. Fix the plugins to deal with that. Fixes #113488.

Original commit message from CVS:
/GstBuffer/GstData/ in the API where you can pass events. Fix the plugins to deal with that. Fixes #113488.

13 years agoNew typefind system: bytestream is now part of the core all plugins have been modifie...
Ronald S. Bultje [Wed, 1 Oct 2003 13:14:50 +0000 (13:14 +0000)]
New typefind system: bytestream is now part of the core all plugins have been modified to use this new typefind syste...

Original commit message from CVS:
New typefind system:
* bytestream is now part of the core
* all plugins have been modified to use this new typefind system
* asf typefinding added
* mpeg video stream typefiding removed because it's broken
* duplicate typefind entries removed
* extra id3 typefinding added, because we've seen 4 types of files
(riff/wav, flac, vorbis, mp3) with id3 headers and each of these needs
to work. Instead, I've added an id3 element and let it redo typefiding
after the id3 header. this needs a hack because spider only typefinds
once. We can remove this hack once spider supports multiple typefinds.
* with all this, mp3 typefinding is semi-rewritten
* id3 typefinding in flac/vorbis is removed, it's no longer needed
* fixed spider and gst-typefind to use this, too.
* Other general cleanups

13 years agoconform to the buffer-frames props entry -- much nicer now...
Andy Wingo [Tue, 30 Sep 2003 12:56:27 +0000 (12:56 +0000)]
conform to the buffer-frames props entry -- much nicer now...

Original commit message from CVS:
conform to the buffer-frames props entry -- much nicer now...

13 years agoRemove redundant plugindir definition
David Schleef [Sun, 10 Aug 2003 00:01:58 +0000 (00:01 +0000)]
Remove redundant plugindir definition

Original commit message from CVS:
Remove redundant plugindir definition

13 years ago+ changes for new float caps without slope/intercept + some category changes for...
Leif Johnson [Sat, 19 Jul 2003 23:25:25 +0000 (23:25 +0000)]
+ changes for new float caps without slope/intercept + some category changes for plugins

Original commit message from CVS:
+ changes for new float caps without slope/intercept
+ some category changes for plugins

13 years agoNew mimetypes gone into effect today - this commit changes all old mimetypes over...
Ronald S. Bultje [Sun, 6 Jul 2003 20:49:52 +0000 (20:49 +0000)]
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri...

Original commit message from CVS:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs

13 years agofix type punning
David Schleef [Tue, 1 Jul 2003 02:27:06 +0000 (02:27 +0000)]
fix type punning

Original commit message from CVS:
fix type punning

13 years agocompatibility fix for new GST_DEBUG stuff.
Benjamin Otte [Sun, 29 Jun 2003 19:46:13 +0000 (19:46 +0000)]
compatibility fix for new GST_DEBUG stuff.

Original commit message from CVS:
compatibility fix for new GST_DEBUG stuff.
Includes fixes for missing includes for config.h and unistd.h

I only ensured for plugins I can build that they work, so if some of them are still broken, you gotta fix them yourselves unfortunately.

13 years agoRemoved ugly caps fixed flag hack, will be done automatically in core soon
Wim Taymans [Fri, 13 Jun 2003 21:21:17 +0000 (21:21 +0000)]
Removed ugly caps fixed flag hack, will be done automatically in core soon

Original commit message from CVS:
Removed ugly caps fixed flag hack, will be done automatically in
core soon

13 years agoupdate for the latest jack cvs and non-cothreaded gst scheduler
Andy Wingo [Tue, 4 Mar 2003 15:34:20 +0000 (15:34 +0000)]
update for the latest jack cvs and non-cothreaded gst scheduler

Original commit message from CVS:
update for the latest jack cvs and non-cothreaded gst scheduler

13 years agoChanged caps->fixed to use FLAG_SET
Jan Schmidt [Wed, 5 Feb 2003 20:38:41 +0000 (20:38 +0000)]
Changed caps->fixed to use FLAG_SET

Original commit message from CVS:
Changed caps->fixed to use FLAG_SET

13 years agoPadConnect -> PadLink
Thomas Vander Stichele [Fri, 10 Jan 2003 13:38:32 +0000 (13:38 +0000)]
PadConnect -> PadLink

Original commit message from CVS:
PadConnect -> PadLink

13 years agoanother batch of connect->link fixes please let me know about issues and please refra...
Thomas Vander Stichele [Fri, 10 Jan 2003 10:22:25 +0000 (10:22 +0000)]
another batch of connect->link fixes please let me know about issues and please refrain of making them yourself, so t...

Original commit message from CVS:
another batch of connect->link fixes
please let me know about issues
and please refrain of making them yourself, so that I don't spend double
the time resolving conflicts

13 years agoparallel install fixes
Thomas Vander Stichele [Sun, 8 Dec 2002 14:50:10 +0000 (14:50 +0000)]
parallel install fixes

Original commit message from CVS:
parallel install fixes

13 years agolicenses again
Andy Wingo [Sun, 29 Sep 2002 18:12:18 +0000 (18:12 +0000)]
licenses again

Original commit message from CVS:
licenses again

13 years agoplugins part of license field patch
Christian Schaller [Wed, 18 Sep 2002 19:02:52 +0000 (19:02 +0000)]
plugins part of license field patch

Original commit message from CVS:
plugins part of license field patch

13 years agoThis updates all plugins to the new API for gst_pad_try_set_caps
Ronald S. Bultje [Tue, 10 Sep 2002 09:31:40 +0000 (09:31 +0000)]
This updates all plugins to the new API for gst_pad_try_set_caps

Original commit message from CVS:
This updates all plugins to the new API for gst_pad_try_set_caps

13 years agoremoving warnings as approved by wim
Thomas Vander Stichele [Mon, 9 Sep 2002 23:27:38 +0000 (23:27 +0000)]
removing warnings as approved by wim

Original commit message from CVS:
removing warnings as approved by wim

13 years agofix jack input port connection
Andy Wingo [Fri, 23 Aug 2002 04:04:11 +0000 (04:04 +0000)]
fix jack input port connection

Original commit message from CVS:
fix jack input port connection

13 years agocompile fixen, and prepare to move MAINTAINER_MODE to as-version.m4
Andy Wingo [Tue, 9 Jul 2002 17:39:17 +0000 (17:39 +0000)]
compile fixen, and prepare to move MAINTAINER_MODE to as-version.m4

Original commit message from CVS:
compile fixen, and prepare to move MAINTAINER_MODE to as-version.m4

13 years agomake jack work in all its full duplex glory
Andy Wingo [Tue, 2 Jul 2002 23:35:07 +0000 (23:35 +0000)]
make jack work in all its full duplex glory

Original commit message from CVS:
make jack work in all its full duplex glory

13 years agoworking jack elements (fixed a problem in upstream jack) random other fixen...
Andy Wingo [Wed, 12 Jun 2002 03:32:02 +0000 (03:32 +0000)]
working jack elements (fixed a problem in upstream jack) random other fixen...

Original commit message from CVS:
* working jack elements (fixed a problem in upstream jack)
* random other fixen...

13 years agouse new bytestream api
Steve Baker [Wed, 15 May 2002 19:08:49 +0000 (19:08 +0000)]
use new bytestream api

Original commit message from CVS:
use new bytestream api

13 years agoupdate to new jack api
Andy Wingo [Mon, 13 May 2002 18:08:33 +0000 (18:08 +0000)]
update to new jack api

Original commit message from CVS:
update to new jack api

13 years agoadd some includes
Andy Wingo [Sun, 5 May 2002 19:39:17 +0000 (19:39 +0000)]
add some includes

Original commit message from CVS:
add some includes

13 years agobetter initialization. it doesn't work over here, though.
Andy Wingo [Sun, 5 May 2002 01:08:05 +0000 (01:08 +0000)]
better initialization. it doesn't work over here, though.

Original commit message from CVS:
better initialization. it doesn't work over here, though.

13 years agoa commit so that jack will build without errors on Uraeus's system ;)
Andy Wingo [Sat, 4 May 2002 21:38:56 +0000 (21:38 +0000)]
a commit so that jack will build without errors on Uraeus's system ;)

Original commit message from CVS:
a commit so that jack will build without errors on Uraeus's system ;)

13 years agoset caps once we know the sample rate of the system
Andy Wingo [Sat, 4 May 2002 20:53:35 +0000 (20:53 +0000)]
set caps once we know the sample rate of the system

Original commit message from CVS:
set caps once we know the sample rate of the system

13 years agosome jack fixes, alsa touchups, and add rtp by default to the build if there are...
Andy Wingo [Sat, 4 May 2002 18:57:44 +0000 (18:57 +0000)]
some jack fixes, alsa touchups, and add rtp by default to the build if there are any problems building rtp, we're mov...

Original commit message from CVS:
some jack fixes, alsa touchups, and add rtp by default to the build

if there are any problems building rtp, we're moving it back to experimental ;)

13 years agoa hack to work around intltool's brokenness a current check for mpeg2dec details...
Andy Wingo [Sat, 20 Apr 2002 21:42:51 +0000 (21:42 +0000)]
a hack to work around intltool's brokenness a current check for mpeg2dec details->klass reorganizations an element br...

Original commit message from CVS:
* a hack to work around intltool's brokenness
* a current check for mpeg2dec
* details->klass reorganizations
* an element browser that uses details->klass
* separated cdxa parse out from the avi directory

13 years agoFinally we're on to a proper jack setup, with a specialized bin and elements that...
Andy Wingo [Tue, 16 Apr 2002 17:14:05 +0000 (17:14 +0000)]
Finally we're on to a proper jack setup, with a specialized bin and elements that can only go in a jack bin. I had to...

Original commit message from CVS:
Finally we're on to a proper jack setup, with a specialized bin and elements
that can only go in a jack bin. I had to fix the parser first to do this, but
to run it, the syntax is like so:

gst-launch jackbin.( filesrc ! mad ! jacksink )

But of course it's not fully functional yet. Sigh.

13 years agoGstPadTemplate <-> gst_pad_template <-> GST_PAD_TEMPLATE same with *factory and typefind.
Andy Wingo [Thu, 11 Apr 2002 20:42:26 +0000 (20:42 +0000)]
GstPadTemplate <-> gst_pad_template <-> GST_PAD_TEMPLATE same with *factory and typefind.

Original commit message from CVS:
GstPadTemplate <-> gst_pad_template <-> GST_PAD_TEMPLATE

same with *factory and typefind.

also, some -Werror fixes.

13 years agoalphabetization fixen a jack caps fix
Andy Wingo [Sat, 30 Mar 2002 21:07:51 +0000 (21:07 +0000)]
alphabetization fixen a jack caps fix

Original commit message from CVS:
* alphabetization fixen
* a jack caps fix

13 years agoadd notify back to filesrc, it's needed for MVC applications remove notify printouts...
Andy Wingo [Sat, 30 Mar 2002 19:31:13 +0000 (19:31 +0000)]
add notify back to filesrc, it's needed for MVC applications remove notify printouts from gst-launch cleanup in gst-p...

Original commit message from CVS:
* add notify back to filesrc, it's needed for MVC applications
* remove notify printouts from gst-launch
* cleanup in gst-plugins configure.ac
* some jack updates
* remove SELF_ITERATING flag in favor of SEF_SCHEDULABLE (not a clear name,
but it's what we have for the moment)
* improve parsing of request pad names, no more sscanf
* fixes to the fastscheduler Makefile.am

13 years agos/Gnome-Streamer/GStreamer/
Andy Wingo [Wed, 20 Mar 2002 21:45:04 +0000 (21:45 +0000)]
s/Gnome-Streamer/GStreamer/

Original commit message from CVS:
s/Gnome-Streamer/GStreamer/

13 years agoremoval of //-style comments don't link plugins to core libs -- the versioning is...
Andy Wingo [Tue, 19 Mar 2002 04:10:06 +0000 (04:10 +0000)]
removal of //-style comments don't link plugins to core libs -- the versioning is done internally to the plugins with...

Original commit message from CVS:
* removal of //-style comments
* don't link plugins to core libs -- the versioning is done internally to the plugins with the plugin_info struct,
and symbol resolution is lazy, so we can always know if a plugin can be loaded by the plugin_info data. in theory.

13 years agos/@GST_PLUGIN_LDFLAGS@/$(GST_PLUGIN_LDFLAGS)/ @-substitued variables variables are...
Andy Wingo [Tue, 19 Mar 2002 01:39:43 +0000 (01:39 +0000)]
s/@GST_PLUGIN_LDFLAGS@/$(GST_PLUGIN_LDFLAGS)/ @-substitued variables variables are defined as make variables automagi...

Original commit message from CVS:
s/@GST_PLUGIN_LDFLAGS@/$(GST_PLUGIN_LDFLAGS)/

@-substitued variables variables are defined as make variables automagically,
and this gives the user the freedom to say make GST_PLUGIN_LDFLAGS=-myflag

13 years agos/gst_element_install_std_props/gst_element_class_install_std_props/ -- it just makes...
Andy Wingo [Mon, 18 Mar 2002 04:41:35 +0000 (04:41 +0000)]
s/gst_element_install_std_props/gst_element_class_install_std_props/ -- it just makes more sense that way added jack ...

Original commit message from CVS:
* s/gst_element_install_std_props/gst_element_class_install_std_props/ -- it just makes more sense that way
* added jack element, doesn't quite work right yet but i didn't want to lose the work -- it does build, register,
and attempt to run though
* imposed some restrictions on the naming of request pads to better allow for reverse parsing
* added '%s' to reverse parsing
* added new bin flag to indicate that it is self-iterating, and some lame code in gst-launch to test it out
* fixen on launch-gui
* added pkg-config stuff for the editor's libs

13 years agov4l2: mark v4l2sink as experimental and build only if --enable-experimental is passed
Tim-Philipp Müller [Sun, 2 Jan 2011 11:37:14 +0000 (11:37 +0000)]
v4l2: mark v4l2sink as experimental and build only if --enable-experimental is passed

It's not really of 'good' quality yet, but there's a lot of
code shared with v4l2src, so not so easy to move it elswhere.

https://bugzilla.gnome.org/show_bug.cgi?id=612244

13 years agoRevert "v4l2: add norm property"
Tim-Philipp Müller [Sun, 2 Jan 2011 01:24:21 +0000 (01:24 +0000)]
Revert "v4l2: add norm property"

This reverts commit 9e1d419d07337e6db2cc3936472be205ce927e54.

Reverting this since it adds unreviewed and bad API to v4l2src
(property of type enum, with seemingly random and unsorted values).

13 years agotools: remove unused left-over directory
Tim-Philipp Müller [Sat, 1 Jan 2011 23:26:33 +0000 (23:26 +0000)]
tools: remove unused left-over directory

These are all in -base/tools.

13 years agomp4adepay: improve timestamps on outgoing packets
Wim Taymans [Fri, 31 Dec 2010 12:57:05 +0000 (13:57 +0100)]
mp4adepay: improve timestamps on outgoing packets

Improve parsing of the samplerate.
Parse the framelen so that we can calculate timestamps.
When interpollate the incomming timestamp on outgoing buffers when there are
multiple subframes.

fixes #625825

13 years agocairo, videofilter: use gst/math-compat.h header for rint
Tim-Philipp Müller [Fri, 31 Dec 2010 02:16:54 +0000 (02:16 +0000)]
cairo, videofilter: use gst/math-compat.h header for rint

13 years agovideobalance: Check for HAVE_RINT instead
David Schleef [Thu, 30 Dec 2010 22:30:27 +0000 (14:30 -0800)]
videobalance: Check for HAVE_RINT instead

Also change M_PI to G_PI for giggles.

13 years agocairo: Don't use #ifdefs inside macros
David Schleef [Thu, 30 Dec 2010 22:21:37 +0000 (14:21 -0800)]
cairo: Don't use #ifdefs inside macros

13 years agoChange M_PI to G_PI
David Schleef [Thu, 30 Dec 2010 22:20:52 +0000 (14:20 -0800)]
Change M_PI to G_PI

13 years agovideobalance: use G_OS_WIN32 for windows check
David Schleef [Thu, 30 Dec 2010 20:07:52 +0000 (12:07 -0800)]
videobalance: use G_OS_WIN32 for windows check

13 years agomp4adepay: fix timestamps on buffers
Wim Taymans [Thu, 30 Dec 2010 15:24:16 +0000 (16:24 +0100)]
mp4adepay: fix timestamps on buffers

13 years agompvpay: fix flushing and discont
Wim Taymans [Thu, 30 Dec 2010 15:22:48 +0000 (16:22 +0100)]
mpvpay: fix flushing and discont

Fix flushing and disconts.
Clean up in state changes.

13 years agomatroska-demux: increase allowed max. block size for push mode from 10M to 15M
Tim-Philipp Müller [Wed, 29 Dec 2010 23:38:18 +0000 (23:38 +0000)]
matroska-demux: increase allowed max. block size for push mode from 10M to 15M

It was an arbitrary limit from the start, meant as a basic sanity check,
so may just as well increase it a little. Would be good to provide
progress reporting while completing the block in any case..

https://bugzilla.gnome.org/show_bug.cgi?id=637060

13 years agomatroska-demux: assume matroska if no doctype is specified
Tim-Philipp Müller [Wed, 29 Dec 2010 23:09:04 +0000 (23:09 +0000)]
matroska-demux: assume matroska if no doctype is specified

https://bugzilla.gnome.org/show_bug.cgi?id=638019

13 years agov4l2: add interlaced support
Rob Clark [Sat, 4 Dec 2010 19:43:11 +0000 (13:43 -0600)]
v4l2: add interlaced support

13 years agov4l2sink: add navigation support
Rob Clark [Sat, 2 Oct 2010 19:45:14 +0000 (14:45 -0500)]
v4l2sink: add navigation support

13 years agov4l2: add norm property
Rob Clark [Sun, 4 Apr 2010 11:43:41 +0000 (06:43 -0500)]
v4l2: add norm property

Based on a patch by Guennadi Liakhovetski.

13 years agov4l2: cleanup get/set input/output
Rob Clark [Tue, 13 Jul 2010 15:03:51 +0000 (10:03 -0500)]
v4l2: cleanup get/set input/output

output devices should use get/set output, and in either case we should
not print a warning message if the ioctl fails but the device does not
claim to support the tuner interface

13 years agov4l2xoverlay: add support to create window
Rob Clark [Thu, 10 Jun 2010 16:15:46 +0000 (11:15 -0500)]
v4l2xoverlay: add support to create window

If xoverlay is available, v4l2sink should create a window for the overlay to
display in.

The window automatically tries to make itself as large as possible.
This works well on a small screen, but perhaps should first attempt to use
the size of the video that is played (no scaling).

13 years agov4l2sink: special handling for cases gst_buffer_make_metadata_writable()
Rob Clark [Sun, 4 Apr 2010 11:41:28 +0000 (06:41 -0500)]
v4l2sink: special handling for cases gst_buffer_make_metadata_writable()

Special case check for sub-buffers:  In certain cases, places like
GstBaseTransform, which might check that the buffer is writable before copying
metadata, timestamp, and such, will find that the buffer has more than one
reference to it.  In these cases, they will create a sub-buffer with an offset=0
and length equal to the original buffer size.

This could happen in two scenarios: (1) a tee in the pipeline, and (2) because
the refcnt is incremented in gst_mini_object_free() before the finalize function
is called, and decremented after it returns..  but returning this buffer to the
buffer pool in the finalize function, could wake up a thread blocked in
_buffer_alloc() which could run and get a buffer w/ refcnt==2 before the thread
originally unref'ing the buffer returns from finalize function and decrements
the refcnt back to 1!

This is related to issue #545501

13 years agov4l2: fix race condition
Rob Clark [Sun, 4 Apr 2010 11:39:52 +0000 (06:39 -0500)]
v4l2: fix race condition

The size of the buffer would be zero'd out in gst_v4l2_buffer_finalize()
after the buffer is qbuf'd or pushed onto the queue of available buffers..
leaving a race condition where the thread waiting for the buffer could awake
and set back a valid size before the finalizing thread zeros out the length.
This would result that the newly allocated buffer has length of zero.

13 years agov4l2sink: add properties to control crop
Rob Clark [Sun, 4 Apr 2010 11:39:08 +0000 (06:39 -0500)]
v4l2sink: add properties to control crop

13 years agov4l2: re-enable x-overlay support
Rob Clark [Sun, 4 Apr 2010 11:37:16 +0000 (06:37 -0500)]
v4l2: re-enable x-overlay support

13 years agov4l2sink: fix for PAUSED->READY->PAUSED state transitions
Rob Clark [Sat, 25 Dec 2010 17:52:36 +0000 (11:52 -0600)]
v4l2sink: fix for PAUSED->READY->PAUSED state transitions

When v4l2sink goes to PAUSED->READY it only stops streaming, so the state
should be set to STATE_PENDING_STREAMON in case the element transitions
back to PLAYING.

13 years agov4l2sink: add "min-queued-bufs" property
Rob Clark [Sun, 4 Apr 2010 11:28:51 +0000 (06:28 -0500)]
v4l2sink: add "min-queued-bufs" property

13 years agov4l2sink: Add support for blocking dequeue.
Rob Clark [Sun, 4 Apr 2010 11:26:50 +0000 (06:26 -0500)]
v4l2sink: Add support for blocking dequeue.

We'd prefer to throttle the decoder if we run out of buffers, to keep a bound
on memory usage.  Also, for OMAP4 it is a requirement of the decoder to not
alternate between memory alloced by the display driver and malloc'd userspace
memory.

13 years agov4l2: clear flags before reusing buffer from buffer pool
Rob Clark [Sun, 4 Apr 2010 11:24:41 +0000 (06:24 -0500)]
v4l2: clear flags before reusing buffer from buffer pool

note: this really only affects v4l2sink since gst_v4l2_buffer_pool_get() is
only called once per buffer in the v4l2src case (in
gst_v4l2src_buffer_pool_activate())

13 years agov4l2sink: don't render preroll buffers
Rob Clark [Sun, 4 Apr 2010 11:23:31 +0000 (06:23 -0500)]
v4l2sink: don't render preroll buffers

Most v4l2 drivers will get upset when you queue the same buffer twice in a
row without first dequeueing it.

Rendering of pre-roll buffers can be re-introduced later, but will require
tracking the state of the buffer, and avoiding to re-QBUF if the buffer has
already been passed to the driver.

13 years agov4l2sink: Improve behavior for shared buffers.
Rob Clark [Sun, 4 Apr 2010 11:22:43 +0000 (06:22 -0500)]
v4l2sink: Improve behavior for shared buffers.

When the decoder is using pad_alloc(), v4l2sink would behave badly if
the number of buffers ('queue-size' property) was not high enough to
account for all the buffers needed by the decoder, and other elements
(such as queues) between the decoder and v4l2sink.  This patch
slightly increases the default number of buffers, and changes v4l2sink
to drop frames rather than return an error in case the number of
buffers is not high enough.

13 years agoadd "client" property
Andy Wingo [Mon, 15 Nov 2010 14:58:28 +0000 (15:58 +0100)]
add "client" property

* ext/pulse/pulsesrc.c (gst_pulsesrc_class_init, gst_pulsesrc_init)
  (gst_pulsesrc_set_property, gst_pulsesrc_get_property)
  (gst_pulsesrc_open): Add a "client" property, as in pulsesink.

Fixes #634914

13 years agortspsrc: serialise/deserialise floats without changing locale
Tim-Philipp Müller [Wed, 29 Dec 2010 15:54:46 +0000 (15:54 +0000)]
rtspsrc: serialise/deserialise floats without changing locale

Use g_ascii_dtostr() and g_ascii_strtod() to serialise/deserialise
floating point numbers, instead of ugly hacks that switch locale
before and after calling libc functions (which is not a good idea
in a multi-threaded application).

13 years agortpjpegdepay: fix framerate parsing for locales that use a comma as floating point
Tim-Philipp Müller [Wed, 29 Dec 2010 14:40:05 +0000 (14:40 +0000)]
rtpjpegdepay: fix framerate parsing for locales that use a comma as floating point

atof() converts strings according to the current locale, but the
framerate string will likely always use a dot as floating point
separator, so use g_ascii_strtod() instead (but also canonicalise
the string before, so we can handle both formats as input).

13 years agortpsource: use the right variable
Wim Taymans [Mon, 27 Dec 2010 12:11:59 +0000 (13:11 +0100)]
rtpsource: use the right variable

Use the right variable for specifying that we sent a receiver report.

13 years agov4l2: fix typo
Rob Clark [Thu, 23 Dec 2010 22:42:29 +0000 (16:42 -0600)]
v4l2: fix typo

13 years agomatroska-demux: add stream-format and alignment properties for h264
Rob Clark [Thu, 23 Dec 2010 22:03:00 +0000 (16:03 -0600)]
matroska-demux: add stream-format and alignment properties for h264

13 years agogstpay: fix klass, add RTP as a use case
Wim Taymans [Wed, 22 Dec 2010 10:41:59 +0000 (11:41 +0100)]
gstpay: fix klass, add RTP as a use case

13 years agogstdepay: cleanup the cache
Wim Taymans [Sun, 12 Dec 2010 14:10:47 +0000 (15:10 +0100)]
gstdepay: cleanup the cache

13 years agogstpay/depay: add generic gstreamer payloader
Wim Taymans [Sun, 12 Dec 2010 04:10:01 +0000 (05:10 +0100)]
gstpay/depay: add generic gstreamer payloader

Add the beginnings of a generic GStreamer buffers payloader.

13 years agomp4gpay: reset state on flush-stop
Wim Taymans [Thu, 23 Dec 2010 16:06:58 +0000 (17:06 +0100)]
mp4gpay: reset state on flush-stop

13 years agomp4gdepay: flush state on flush-stop
Wim Taymans [Thu, 23 Dec 2010 15:26:07 +0000 (16:26 +0100)]
mp4gdepay: flush state on flush-stop

13 years agortspsrc: on-npt-stop is a manager signal
Wim Taymans [Thu, 23 Dec 2010 15:25:15 +0000 (16:25 +0100)]
rtspsrc: on-npt-stop is a manager signal

13 years agortspsrc: improve RTP session handling
Wim Taymans [Thu, 23 Dec 2010 14:24:29 +0000 (15:24 +0100)]
rtspsrc: improve RTP session handling

Store the RTP session in the stream so that we can more efficiently
perform actions on the stream based on RTP signals.

13 years agortpsource: include last send RB block
Wim Taymans [Thu, 23 Dec 2010 12:55:31 +0000 (13:55 +0100)]
rtpsource: include last send RB block

Only report RB values for non-internal sources.
Report not only the RB blocks we last received from but also the last RB
block we sent to a source.

13 years agortpsession: remember last sent RB values.
Wim Taymans [Thu, 23 Dec 2010 12:52:57 +0000 (13:52 +0100)]
rtpsession: remember last sent RB values.

13 years agortpsource: include all stats and document
Wim Taymans [Thu, 23 Dec 2010 12:00:49 +0000 (13:00 +0100)]
rtpsource: include all stats and document

Include all possible stats of a source in the stats structure because we might
be interested in what happened in the past.
Document the stats property and the fields.

13 years agoexamples: add example RTP stats
Wim Taymans [Thu, 23 Dec 2010 11:59:59 +0000 (12:59 +0100)]
examples: add example RTP stats

Add some more RTP examples for how to retrieve RTP stats in a receiver.

13 years agortpsession: also emit RTCP activity on SR
Wim Taymans [Thu, 23 Dec 2010 11:58:05 +0000 (12:58 +0100)]
rtpsession: also emit RTCP activity on SR

Also emit RTCP activity signals when we receive an SR packet without RB blocks,
such as from a sender that is not receiving anything.

13 years agodocs: add some more gstrtpbin docs
Wim Taymans [Thu, 23 Dec 2010 10:10:55 +0000 (11:10 +0100)]
docs: add some more gstrtpbin docs

13 years agoximagesrc: remote is a boolean (and not uint) property
Edward Hervey [Wed, 22 Dec 2010 20:27:11 +0000 (21:27 +0100)]
ximagesrc: remote is a boolean (and not uint) property

13 years agomatroskademux: Don't use gst_pad_alloc_buffer()
Sebastian Dröge [Wed, 22 Dec 2010 18:58:21 +0000 (19:58 +0100)]
matroskademux: Don't use gst_pad_alloc_buffer()

Using this in a demuxer will cause deadlocks if there's
a pad with a pending pad-block downstream, no matter if
there is a queue between the pad or not. Queues pass
bufferalloc downstream from the same thread and only
act as a thread boundary for events and buffers.

13 years agomatroskamux: fix subtitle pad template, we only handle kate for now
Tim-Philipp Müller [Wed, 22 Dec 2010 14:14:08 +0000 (14:14 +0000)]
matroskamux: fix subtitle pad template, we only handle kate for now

13 years agodocs: update rtspsrc docs, rtpbin is not in -bad any more
Tim-Philipp Müller [Thu, 16 Dec 2010 11:44:44 +0000 (11:44 +0000)]
docs: update rtspsrc docs, rtpbin is not in -bad any more

13 years agortpsession: unlock before emitting signals
Wim Taymans [Wed, 22 Dec 2010 10:42:31 +0000 (11:42 +0100)]
rtpsession: unlock before emitting signals

13 years agortpac3pay: add AC3 payloader
Wim Taymans [Tue, 21 Dec 2010 21:34:49 +0000 (22:34 +0100)]
rtpac3pay: add AC3 payloader