GitLab

Original commit message from CVS:
* gst/gstbin.c: (gst_bin_init), (gst_bin_provide_clock_func),
(is_eos), (gst_bin_add_func), (gst_bin_remove_func),
(gst_bin_recalc_state), (gst_bin_change_state_func),
(gst_bin_dispose), (bin_bus_handler):
* gst/gstbin.h:
Prepare to make current EOS message queue more generic.
Fix some typos.

* gst/gstevent.c: (gst_event_new_newsegment),
(gst_event_parse_newsegment):
* gst/gstevent.h:
Rename base to stream_time.

* gst/gstmessage.h:
Fix typo in docs.

Original commit message from CVS:
* gst/gstbin.c: (gst_bin_init), (gst_bin_provide_clock_func),
(gst_bin_add_func), (gst_bin_remove_func), (gst_bin_recalc_state),
(gst_bin_change_state_func), (bin_bus_handler):
* gst/gstbin.h:
Work on proper clock selection.

libs/gst/controller/gstcontroller.*: Added GList* version of _remove_properties() in order to be able to wrap it in b...

Original commit message from CVS:
* libs/gst/controller/gstcontroller.c: (gst_controller_remove_properties_list):
* libs/gst/controller/gstcontroller.h:
Added GList* version of _remove_properties() in order to be able to wrap
it in bindings.

Original commit message from CVS:
* docs/design/part-states.txt:
Some more docs.

* gst/gstbin.c: (gst_bin_set_clock_func), (gst_bin_recalc_state),
(gst_bin_change_state_func), (bin_bus_handler):
Doc updates. Don't distribute the same clock over and over again.

* gst/gstclock.c:
* gst/gstclock.h:
Doc updates.

* gst/gstpad.c: (gst_flow_get_name), (gst_flow_to_quark),
(gst_pad_get_type), (gst_pad_push), (gst_pad_push_event),
(gst_pad_send_event):
* gst/gstpad.h:
Make probe emission threadsafe again.
Register quarks and move _get_name() from utils.
Doc updates.

* gst/gstpipeline.c: (gst_pipeline_class_init),
(gst_pipeline_change_state), (gst_pipeline_provide_clock_func):
Only redistribute the clock of it changed.

* gst/gstsystemclock.h:
Doc updates.

* gst/gstutils.c:
* gst/gstutils.h:
Moved the _flow_get_name() to GstPad.

Original commit message from CVS:
if we log our init, should also log finalize

Original commit message from CVS:
fix more valgrind warnings before turning up the heat

Original commit message from CVS:
don't declare on the proper define

Original commit message from CVS:
unmangle the nesting a little

Original commit message from CVS:

* gst/parse/grammar.y:
some cleanup before the hacking

Original commit message from CVS:

* gst/base/gstbasesrc.c: (gst_base_src_query):
use conversions
* gst/gstutils.c: (gst_guint64_to_gdouble),
(gst_gdouble_to_guint64), (gst_util_uint64_scale):
* gst/gstutils.h:
externalize, basesrc uses it
obviously the implementation needs testing

Original commit message from CVS:
another cast bites the dust

Original commit message from CVS:
another cast bites the dust

Original commit message from CVS:
* tests/sched/Makefile.am:
* tests/sched/sort.c: (make_pipeline1), (make_pipeline2),
(make_pipeline3), (make_pipeline4), (print_elem), (main):

Original commit message from CVS:

* gst/gstutils.c: (guint64_to_gdouble), (gst_util_uint64_scale):
apparently converting from guint64 to double is not implemented
on MSVC

Original commit message from CVS:
* check/Makefile.am:
* check/generic/states.c: (GST_START_TEST):
* check/gst/gstbin.c: (GST_START_TEST):
* check/gst/gstpipeline.c: (GST_START_TEST), (gst_pipeline_suite):
* check/states/sinks.c: (GST_START_TEST):
* check/states/sinks2.c: (GST_START_TEST), (gst_object_suite),
(main):
Check fixes, use API as stated in design docs, remove hacks.

* gst/base/gstbasesink.c: (gst_base_sink_handle_object),
(gst_base_sink_change_state):
Catch stopping our task while we're shutting down.

* gst/gstbin.c: (gst_bin_init), (gst_bin_add_func),
(gst_bin_remove_func), (gst_bin_get_state_func),
(gst_bin_recalc_state), (gst_bin_change_state_func),
(bin_bus_handler):
* gst/gstbin.h:
* gst/gstelement.c: (gst_element_init),
(gst_element_get_state_func), (gst_element_abort_state),
(gst_element_commit_state), (gst_element_lost_state),
(gst_element_set_state), (gst_element_change_state),
(gst_element_change_state_func):
* gst/gstelement.h:
New state change algorithm (see #318116)

* gst/gstpipeline.c: (gst_pipeline_class_init),
(gst_pipeline_init), (gst_pipeline_set_property),
(gst_pipeline_get_property), (do_pipeline_seek),
(gst_pipeline_change_state), (gst_pipeline_provide_clock_func):
* gst/gstpipeline.h:
Remove crude state change hacks.

* gst/gstutils.h:
Remove crude hacks.

* tools/gst-launch.c: (main):
Fixes for state change. Needs some more work to fully use the
new stuff.

Original commit message from CVS:
2005-10-10  Andy Wingo  <wingo@pobox.com>

* tests/Makefile.am (noinst_PROGRAMS): No more init.c.

gst/gst.c (G_OPTION_FLAG_NO_ARG): Apparently GLib 2.8 requires this flag, but it's not even in GLib 2.6. Odd. Hack ar...

Original commit message from CVS:
2005-10-10  Andy Wingo  <wingo@pobox.com>

* gst/gst.c (G_OPTION_FLAG_NO_ARG): Apparently GLib 2.8 requires
this flag, but it's not even in GLib 2.6. Odd. Hack around the
issue.

gst/gstiterator.c: Fix my previous commit: GTypes passed to gst_iterator_new() can be fundamental types.

Original commit message from CVS:
* gst/gstiterator.c: (gst_iterator_new):
Fix my previous commit: GTypes passed to gst_iterator_new()
can be fundamental types.

Original commit message from CVS:
* gst/gstelement.c: (gst_element_iterate_pad_list),
(gst_element_iterate_pads), (gst_element_iterate_src_pads),
(gst_element_iterate_sink_pads):
Use src/sink pads lists for the respective iterators instead
of filtering.

Original commit message from CVS:
2005-10-10  Andy Wingo  <wingo@pobox.com>

Merged in popt removal + GOption addition patch from Ronald, bug
#169772.

* docs/gst/gstreamer-sections.txt: Add STATE_(UN)LOCK_FULL, move
GstElement macros around, remove popt-related symbols, add goption
stuff.

* configure.ac: Remove popt checks, require GLib 2.6 for GOption.

* docs/gst/Makefile.am:
* docs/libs/Makefile.am: No POPT_CFLAGS.

* examples/manual/Makefile.am:
* docs/manual/basics-init.xml: Doc updates with an example.

* gst/gst.c: (gst_init_get_option_group), (gst_init_check),
(gst_init), (parse_one_option), (parse_goption_arg):
* gst/gst.h: Removed gst_init_with_popt_table and friends. Took a
bit of hand merging and debugging to get the GOption stuff working
tho.

* tests/Makefile.am:
* tools/Makefile.am:
* tools/gst-inspect.c: (main):
* tools/gst-launch.c: (main):
* tools/gst-run.c: (main):
* tools/gst-xmlinspect.c: (main): Thanks Ronald!

gst/gstiterator.c: Add assertions to make sure passed GType is likely to really be a GType (as the compiler won't cat...

Original commit message from CVS:
* gst/gstiterator.c: (gst_iterator_new):
Add assertions to make sure passed GType is likely to really
be a GType (as the compiler won't catch it if the size and
GType arguments get mixed up, see #318447).

gst/gstbin.c: Pass GType and size arguments to gst_iterator_new() in the right order (maybe we should make _new() tak...

Original commit message from CVS:
Reviewed by: Tim-Philipp Müller  <tim at centricular dot net>
* gst/gstbin.c: (gst_bin_iterate_sorted):
Pass GType and size arguments to gst_iterator_new() in the right
order (maybe we should make _new() take the GType as first argument
just like _new_list()?) (#318447).

Original commit message from CVS:
* gst/gstelement.c: (gst_element_finalize):
And free the GStaticRecMutex too

Original commit message from CVS:
don't echo path

Original commit message from CVS:
2005-10-10  Andy Wingo  <wingo@pobox.com>

* gst/gstelement.c (gst_element_init, gst_element_finalize):
Allocate and free the mutex properly.

* gst/gstelement.h (GST_STATE_UNLOCK_FULL, GST_STATE_LOCK_FULL):
New macros.
(GstElement): The state_lock is now recursive. Rebuild your
plugins, suckers. Old macros adapted.

Original commit message from CVS:
changelog

Original commit message from CVS:
2005-10-10  Andy Wingo  <wingo@pobox.com>

* docs/gst/gstreamer-sections.txt: Doc updates.

* gst/gstutils.h:
* gst/gstutils.c (g_static_rec_cond_timed_wait)
(g_static_rec_cond_wait): Ported from state changes patch, while
we wait on bug #317802 to be solved in a well-distributed GLib.

Original commit message from CVS:
add more win32 build files

gst/gstelement.c (gst_element_change_state_func): Renamed from gst_element_change_state, variable name changes.

Original commit message from CVS:
2005-10-10  Andy Wingo  <wingo@pobox.com>

* gst/gstelement.c (gst_element_change_state_func): Renamed from
gst_element_change_state, variable name changes.
(gst_element_change_state): Split out of gst_element_set_state in
preparation for the state change merge. Doesn't pay attention to
the 'transition' argument.
(gst_element_set_state): Updates, hopefully purely cosmetic.
(gst_element_sync_state_with_parent): MT-safety. Ported from the
state change patch.
(gst_element_get_state_func): Renamed from get_state, cosmetic
changes.

Original commit message from CVS:

* gst/elements/gstelements.c:
* win32/GStreamer.vcproj:
* win32/config.h:
* win32/dirent.c: (_tseekdir):
* win32/gst-inspect.vcproj:
* win32/gst-launch.vcproj:
* win32/gstconfig.h:
* win32/gstelements.vcproj:
* win32/gstenumtypes.c: (gst_object_flags_get_type):
* win32/gstreamer.def:
* win32/msvc71.sln:
updates for the win32 build (patch from Sebastien Moutte)

gst/gstbin.c (gst_bin_get_state_func): Renamed from gst_bin_get_state, cleaned up (but no logic changes).

Original commit message from CVS:
2005-10-10  Andy Wingo  <wingo@pobox.com>

* gst/gstbin.c (gst_bin_get_state_func): Renamed from
gst_bin_get_state, cleaned up (but no logic changes).
(bin_element_is_sink): Comment updates.
(sink_iterator_filter): Remove needless cast.
(gst_bin_iterate_sinks): Doc update.
(gst_bin_change_state_func): Renamed from gst_bin_change_state,
cleaned up (but no logic changes).

Original commit message from CVS:
2005-10-10  Andy Wingo  <wingo@pobox.com>

* check/states/sinks.c (test_src_sink): Cleanups from the state
change patch.
(test_livesrc_sink): Sync on the state.

Original commit message from CVS:
2005-10-10  Andy Wingo  <wingo@pobox.com>

* check/pipelines/simple_launch_lines.c (run_pipeline): Merge from
the state change patch.

Original commit message from CVS:
2005-10-10  Andy Wingo  <wingo@pobox.com>

* check/gst/gstghostpad.c (test_ghost_pads): Merge from the state
change patch.

Original commit message from CVS:
2005-10-10  Andy Wingo  <wingo@pobox.com>

* check/gst/gstbin.c: Merge in some style fixes and additional
checks from Wim's state change patch.

gst/base/gsttypefindhelper.c: Check whether we have the requested data already in our list of cached buffers before p...

Original commit message from CVS:
* gst/base/gsttypefindhelper.c: (helper_find_peek),
(gst_type_find_helper):
Check whether we have the requested data already in our list of
cached buffers before pulling a new buffer; also make the buffer
list a GSList. Speeds up typefinding by ca. 5-10% altogether.

Original commit message from CVS:

* gst/gstcaps.c:
* gst/gstevent.c:
doc updates
* gst/gstvalue.c: (gst_value_deserialize_int_helper):
don't use long long, it's not portable.  Replacing with
gint64 seems to work; let's hope no skeletons fall out of the closet.

Original commit message from CVS:
2005-10-10  Andy Wingo  <wingo@pobox.com>

* autogen.sh (CONFIGURE_DEF_OPT): No more --plugin-buiddir, yay

Original commit message from CVS:
* docs/gst/gstreamer-sections.txt:
* gst/gstevent.c:
* gst/gstevent.h:
* gst/gstinfo.c:
* gst/gstinfo.h:
* gst/gstmessage.c: (gst_message_parse_state_changed):
* gst/gstpad.c:
* gst/gstpad.h:
more docs, fix compilation

Original commit message from CVS:
2005-10-09  Philippe Khalaf <burger@speedy.org>
* gst/gstmessage.c:
Fixed a few forgotten variables on previous commit