platform/upstream/gstreamer.git
7 years agoRelease 1.11.1
Sebastian Dröge [Thu, 12 Jan 2017 14:20:08 +0000 (16:20 +0200)]
Release 1.11.1

7 years agoclip: Make sure that clip start change is notified before children changes
Thibault Saunier [Mon, 9 Jan 2017 15:12:34 +0000 (12:12 -0300)]
clip: Make sure that clip start change is notified before children changes

Fixes https://phabricator.freedesktop.org/T7577

Differential Revision: https://phabricator.freedesktop.org/D1600

7 years agotests_: Check the order of signals when a transition is created
Alexandru Băluț [Wed, 19 Oct 2016 15:36:49 +0000 (15:36 +0000)]
tests_: Check the order of signals when a transition is created

Reviewed-by: Thibault Saunier <thibault.saunier@collabora.com>
Differential Revision: https://phabricator.freedesktop.org/D1391

7 years agoasset: Fix set_proxy to abort when an error happens
Alexandru Băluț [Wed, 21 Dec 2016 11:22:31 +0000 (12:22 +0100)]
asset: Fix set_proxy to abort when an error happens

Differential Revision: https://phabricator.freedesktop.org/D1574

7 years agoasset: Reuse local variable
Alexandru Băluț [Wed, 21 Dec 2016 10:39:12 +0000 (11:39 +0100)]
asset: Reuse local variable

Differential Revision: https://phabricator.freedesktop.org/D1573

7 years agomeson: generate pkg-config -uninstalled pc files
Guillaume Desmottes [Wed, 4 Jan 2017 14:55:36 +0000 (15:55 +0100)]
meson: generate pkg-config -uninstalled pc files

Generating those files is useful for users building the GStreamer stack
using meson and having to link it to another project which is still
using the autotools.

https://bugzilla.gnome.org/show_bug.cgi?id=776810

7 years agoges-launch: Use standard GstEncodingProfile deserialization function
Thibault Saunier [Fri, 23 Dec 2016 18:08:06 +0000 (15:08 -0300)]
ges-launch: Use standard GstEncodingProfile deserialization function

7 years agovalidate: Remove space breaking muting ges-launch
Thibault Saunier [Thu, 22 Dec 2016 13:00:06 +0000 (10:00 -0300)]
validate: Remove space breaking muting ges-launch

7 years agovideo-mixer: Fix the way we release mixer pads
Thibault Saunier [Thu, 22 Dec 2016 12:48:58 +0000 (09:48 -0300)]
video-mixer: Fix the way we release mixer pads

We were using the actual mixer pad to release the smart mixer
pad, which seemed to be on purpose, but was not properly handle,
moreover, it is now forbiden to pass a pad not inside a GstElement
when releasing it.

Also properly remove ghost pads from Smart mixer, we were planly
failling at it.

7 years agomeson: Install ges-launch
Thibault Saunier [Thu, 22 Dec 2016 11:44:07 +0000 (08:44 -0300)]
meson: Install ges-launch

7 years agoRemove generated .spec file
Tim-Philipp Müller [Fri, 16 Dec 2016 17:29:59 +0000 (17:29 +0000)]
Remove generated .spec file

Likely extremely bitrotten, and we should not ship this anyway.

7 years agotests: Remove now meaningless empty testsuite
Thibault Saunier [Fri, 16 Dec 2016 17:04:06 +0000 (14:04 -0300)]
tests: Remove now meaningless empty testsuite

7 years agoges: add a basic unix man page for ges-launch-1.0
Antonio Ospite [Tue, 13 Dec 2016 15:05:17 +0000 (16:05 +0100)]
ges: add a basic unix man page for ges-launch-1.0

Do not list all the possible options in the man page but only the help
options.

This is in order to avoid duplication and prevent the man page from
becoming obsolete in case the options change in the code but do not get
updated in the man page.

https://bugzilla.gnome.org/show_bug.cgi?id=776063

7 years agoges: fix the description of the --help-GES command line option
Antonio Ospite [Tue, 13 Dec 2016 14:10:26 +0000 (15:10 +0100)]
ges: fix the description of the --help-GES command line option

Use "Show GES Options" which is more appropriate and avoids duplication
with --help-gst which already says "Show GStreamer Options".

https://bugzilla.gnome.org/show_bug.cgi?id=776063

7 years agotests: Make sure tests can be listed
Thibault Saunier [Mon, 12 Dec 2016 19:59:08 +0000 (16:59 -0300)]
tests: Make sure tests can be listed

Initializing GstCheck before creating the testsuite

7 years agoges-launch: Make sure GStreamer is always initialized
Thibault Saunier [Wed, 14 Dec 2016 02:26:23 +0000 (23:26 -0300)]
ges-launch: Make sure GStreamer is always initialized

https://bugzilla.gnome.org/show_bug.cgi?id=776064

7 years agomeson: Support building without Gst debug
Thibault Saunier [Fri, 9 Dec 2016 20:50:28 +0000 (17:50 -0300)]
meson: Support building without Gst debug

7 years agoAutomatic update of common submodule
Edward Hervey [Sat, 3 Dec 2016 07:21:07 +0000 (08:21 +0100)]
Automatic update of common submodule

From 1416c16 to 39ac2f5

7 years agoelement: Rework set_child_property_by_pspec
Thibault Saunier [Thu, 1 Dec 2016 20:08:43 +0000 (17:08 -0300)]
element: Rework set_child_property_by_pspec

It was making no sense to loose the information about the pspec itself
to retrieve the child associated to it and was failling when we were
forcing the AssociateType::prop synthax

7 years agotest:nle: Soften check on refcount
Thibault Saunier [Thu, 1 Dec 2016 18:46:51 +0000 (15:46 -0300)]
test:nle: Soften check on refcount

The composition might already have taken a new ref processing the
source.

7 years agonlecomposition: Deactivate current stack in PAUSED_READY state
Thibault Saunier [Wed, 30 Nov 2016 12:53:38 +0000 (09:53 -0300)]
nlecomposition: Deactivate current stack in PAUSED_READY state

To avoid a race when tearing down the composition (PAUSED_TO_READY),
we should make sure to tear down the current stack and let the GstBin
class handle the remaining thing to do during the change state.

We should still ignore any error happening when tearing down the
bin state just in case.

https://bugzilla.gnome.org/show_bug.cgi?id=775051

7 years agotools: Deinit Gst before exiting
Thibault Saunier [Wed, 30 Nov 2016 12:53:21 +0000 (09:53 -0300)]
tools: Deinit Gst before exiting

7 years agomeson: Do not print error logs when building the gir
Thibault Saunier [Tue, 29 Nov 2016 13:37:11 +0000 (10:37 -0300)]
meson: Do not print error logs when building the gir

7 years agoexamples: remove #include <ges-internal.h> from assets.c
Scott D Phillips [Tue, 29 Nov 2016 01:07:39 +0000 (17:07 -0800)]
examples: remove #include <ges-internal.h> from assets.c

It is not needed and pulling it in is causing a link problem with msvc.
Including ges-internal.h sets the default debug category in assets.c to
_ges_debug. Because _ges_debug is marked as DATA in the libges.def, it
will only be linked from libges.dll if it is marked in the source with
dllimport. Instead of messing with that we can just remove this include.

https://bugzilla.gnome.org/show_bug.cgi?id=775295

7 years agocommon: use https protocol for common submodule
Tim-Philipp Müller [Sat, 26 Nov 2016 11:25:41 +0000 (11:25 +0000)]
common: use https protocol for common submodule

https://bugzilla.gnome.org/show_bug.cgi?id=775110

7 years agoges-timeline: Properly calculate absolute diff of two unsigned integers
Sebastian Dröge [Wed, 23 Nov 2016 16:42:27 +0000 (18:42 +0200)]
ges-timeline: Properly calculate absolute diff of two unsigned integers

CID 1394491.

7 years agoges: Add NULL check before dereferencing
Sebastian Dröge [Wed, 23 Nov 2016 16:28:35 +0000 (18:28 +0200)]
ges: Add NULL check before dereferencing

CID 1394494.

7 years agoges-uri-asset: fix compile error 'timeout' may be used uninitialized
Philippe Renon [Sun, 20 Nov 2016 14:34:46 +0000 (15:34 +0100)]
ges-uri-asset: fix compile error 'timeout' may be used uninitialized

https://bugzilla.gnome.org/show_bug.cgi?id=774751

7 years agomake: include common/win32.mak
Scott D Phillips [Fri, 18 Nov 2016 18:21:45 +0000 (10:21 -0800)]
make: include common/win32.mak

With the addition of the .def file for libges we need to make
sure the check-export script from common gets executed so that the
.def stays up to date.

https://bugzilla.gnome.org/show_bug.cgi?id=774641

7 years agoUpdate common submodule
Thibault Saunier [Fri, 18 Nov 2016 19:55:17 +0000 (16:55 -0300)]
Update common submodule

7 years agoEnable building with MSVC
Scott D Phillips [Thu, 17 Nov 2016 18:31:50 +0000 (10:31 -0800)]
Enable building with MSVC

https://bugzilla.gnome.org/show_bug.cgi?id=774641

7 years agoCast away const from GstMetaInfo in *_get_meta_info() functions
Scott D Phillips [Thu, 17 Nov 2016 18:40:05 +0000 (10:40 -0800)]
Cast away const from GstMetaInfo in *_get_meta_info() functions

MSVC warns about the const in the implicit argument conversion in the
calls to g_once_init_{enter,leave}. It's OK so explicitly cast it.

https://bugzilla.gnome.org/show_bug.cgi?id=774641

7 years agoPass gint/guint pointers instead of enum pointers
Scott D Phillips [Thu, 17 Nov 2016 18:39:01 +0000 (10:39 -0800)]
Pass gint/guint pointers instead of enum pointers

The underlying integer type for enums are implementation defined and may
not be the same size as gint/guint. So implicitly casting from pointers-
to-enum-types to pointers-to-int-types is unsafe. MSVC warns on these.

https://bugzilla.gnome.org/show_bug.cgi?id=774641

7 years agoparse: Don't #include <unistd.h>
Scott D Phillips [Thu, 17 Nov 2016 18:35:50 +0000 (10:35 -0800)]
parse: Don't #include <unistd.h>

It isn't needed and isn't present in non-posix environments like windows
with MSVC or mingw.

https://bugzilla.gnome.org/show_bug.cgi?id=774641

7 years agoRevert "nlecomposition: Start task and initialize the stack after chaining up to...
Sebastian Dröge [Thu, 17 Nov 2016 07:40:38 +0000 (09:40 +0200)]
Revert "nlecomposition: Start task and initialize the stack after chaining up to parent's change state function"

This reverts commit 57d40bec1a3c5048baaad08403d7b7e641a9c55c.

Apparently it causes timeouts in the unit tests on Jenkins and
Thibault's machine, and in the gst-validate tests.

Caused by elements staying in PAUSED and waiting to be set to PLAYING.
Needs further investigation.

7 years agoRevert "nleobject: Start up in NULL->READY->PAUSED after the parent class did"
Sebastian Dröge [Thu, 17 Nov 2016 07:40:33 +0000 (09:40 +0200)]
Revert "nleobject: Start up in NULL->READY->PAUSED after the parent class did"

This reverts commit 5f7943c59d9def8c2dc9983936463462c1cdf63f.

7 years agonleobject: Start up in NULL->READY->PAUSED after the parent class did
Sebastian Dröge [Wed, 16 Nov 2016 16:11:00 +0000 (18:11 +0200)]
nleobject: Start up in NULL->READY->PAUSED after the parent class did

This keeps everything in a more consistent order and makes sure that the
base class is already set up completely before we start doing anything.
It also prevents from doing any setup if the base class fails, and
possibly not shutting things down again then.

https://bugzilla.gnome.org/show_bug.cgi?id=774480

7 years agonlecomposition: Start task and initialize the stack after chaining up to parent's...
Sebastian Dröge [Tue, 15 Nov 2016 15:56:00 +0000 (17:56 +0200)]
nlecomposition: Start task and initialize the stack after chaining up to parent's change state function

Otherwise we could set the state of the children to PAUSED already (i.e.
start dataflow) from the composition's task, while the composition
itself is currently chaining up to the parent class' change state
function and did not activate the pads yet. This causes buffers and
events to be discarded, and everything to stop with a not-negotiated
error.

https://bugzilla.gnome.org/show_bug.cgi?id=774480

7 years agotests: Fix running python unit tests
Thibault Saunier [Tue, 15 Nov 2016 21:34:44 +0000 (18:34 -0300)]
tests: Fix running python unit tests

Adding missing nose2-junit-xml.cfg.in file and minor fixes in
the way we call nose2 also making sure the .xunit files end
up in the right place.

7 years agoges: Check if GstDiscoverer could be created at init time
Thibault Saunier [Tue, 15 Nov 2016 18:09:10 +0000 (15:09 -0300)]
ges: Check if GstDiscoverer could be created at init time

And fail initialization if it is not the case, we make the assumption
it worked all around the codebase so we should really concider it fatal.

7 years agonlecomposition: Fix small remaining race in previous commit
Sebastian Dröge [Thu, 10 Nov 2016 13:17:50 +0000 (15:17 +0200)]
nlecomposition: Fix small remaining race in previous commit

The seek action might currently be handled (in which case it is not in
the actions list and the action lock is not locked), but not actually
handled completely yet (the seqnum is not stored yet).

To prevent this, we remember what the current action is that is being
handled, and also compare to that.

https://bugzilla.gnome.org/show_bug.cgi?id=774149

7 years agotests_: Use nose2 instead of nosetests
Alexandru Băluț [Wed, 19 Oct 2016 14:34:56 +0000 (16:34 +0200)]
tests_: Use nose2 instead of nosetests

Differential Revision: https://phabricator.freedesktop.org/D1394

7 years agoges: Fix documentation and debug comments
Alexandru Băluț [Wed, 19 Oct 2016 10:36:45 +0000 (12:36 +0200)]
ges: Fix documentation and debug comments

Reviewed-by: Thibault Saunier <thibault.saunier@collabora.com>
Differential Revision: https://phabricator.freedesktop.org/D1393

7 years agotimeline: reimplement snap_to_position a bit more appropriately.
Mathieu Duponchelle [Tue, 12 Jan 2016 14:51:55 +0000 (14:51 +0000)]
timeline: reimplement snap_to_position a bit more appropriately.

It could yet be made be simpler, but it would require
touching the rest of the timeline editing code.

Fixes https://phabricator.freedesktop.org/T7587

Reviewed-by: Thibault Saunier <thibault.saunier@collabora.com>
Differential Revision: https://phabricator.freedesktop.org/D657

7 years agonlecomposition: De-duplicate seek events based on their sequence number
Sebastian Dröge [Wed, 9 Nov 2016 15:14:19 +0000 (17:14 +0200)]
nlecomposition: De-duplicate seek events based on their sequence number

If there are e.g. multiple video sinks, we would get the same seek event
multiple times. But we only want to handle it once.

https://bugzilla.gnome.org/show_bug.cgi?id=774149

7 years agotest:validate: Port to python3
Thibault Saunier [Mon, 7 Nov 2016 21:01:51 +0000 (18:01 -0300)]
test:validate: Port to python3

7 years agoges-timeline: Fix typo in debug messages
Sebastian Dröge [Wed, 9 Nov 2016 09:48:09 +0000 (11:48 +0200)]
ges-timeline: Fix typo in debug messages

7 years agotimeline: Avoid creating extra transition when rippling clips
Thibault Saunier [Sat, 8 Oct 2016 08:43:07 +0000 (10:43 +0200)]
timeline: Avoid creating extra transition when rippling clips

In some cases when rippling clip we could get the algo lost because
a transition existed between two clips (for example at the end of c1
and at the begining of c2) but while rippling it would have required
a transition at the end of c2 and beginning of c1, and we were properly
not destroying the old one (as the two clips were in the moving context)
but we were still creating the other transition in the end...

Reviewed-by: Alex Băluț <alexandru.balut@gmail.com>
Differential Revision: https://phabricator.freedesktop.org/D1362

7 years agotimeline: Make sure transitions between rippled clips are never deleted
Thibault Saunier [Fri, 7 Oct 2016 13:31:40 +0000 (15:31 +0200)]
timeline: Make sure transitions between rippled clips are never deleted

Reviewed-by: Alex Băluț <alexandru.balut@gmail.com>
Differential Revision: https://phabricator.freedesktop.org/D1361

7 years agotimeline: Destroy transition if a neighbor is not being moved to a layer
Thibault Saunier [Thu, 6 Oct 2016 17:14:57 +0000 (19:14 +0200)]
timeline: Destroy transition if a neighbor is not being moved to a layer

And make sure that we move the transition to the right layer, not trying
to figure it out.

Differential Revision: https://phabricator.freedesktop.org/D1360

7 years agotests_: Check transition is gone when editing clip to another layer
Alexandru Băluț [Thu, 6 Oct 2016 12:00:23 +0000 (14:00 +0200)]
tests_: Check transition is gone when editing clip to another layer

Differential Revision: https://phabricator.freedesktop.org/D1359

7 years agomeson: Unset the plugin paths to generate the .gir files
Thibault Saunier [Fri, 4 Nov 2016 17:41:13 +0000 (14:41 -0300)]
meson: Unset the plugin paths to generate the .gir files

Avoiding problems when using subproject:

'Failed to load plugin something.so file too short'

7 years agomeson: update version
Tim-Philipp Müller [Tue, 1 Nov 2016 18:10:47 +0000 (18:10 +0000)]
meson: update version

7 years agoBack to development
Sebastian Dröge [Tue, 1 Nov 2016 16:53:15 +0000 (18:53 +0200)]
Back to development

7 years agoRelease 1.10.0
Sebastian Dröge [Tue, 1 Nov 2016 16:12:35 +0000 (18:12 +0200)]
Release 1.10.0

7 years agomeson: Don't depend on gstreamer-check-1.0 on windows
Scott D Phillips [Tue, 25 Oct 2016 15:54:11 +0000 (08:54 -0700)]
meson: Don't depend on gstreamer-check-1.0 on windows

https://bugzilla.gnome.org/show_bug.cgi?id=773114

7 years agoRevert "meson: move gstreamer-check-1.0 dependency to tests/check"
Nirbheek Chauhan [Tue, 25 Oct 2016 06:18:35 +0000 (11:48 +0530)]
Revert "meson: move gstreamer-check-1.0 dependency to tests/check"

This reverts commit 5665c2bfc9cae531c6dd9a75766d06a4af25ab9a.

Does not actually work. See:
https://bugzilla.gnome.org/show_bug.cgi?id=773114#c31

7 years agomeson: move gstreamer-check-1.0 dependency to tests/check
Thibault Saunier [Fri, 21 Oct 2016 08:49:18 +0000 (05:49 -0300)]
meson: move gstreamer-check-1.0 dependency to tests/check

7 years agomeson: mark gstreamer-check-1.0 as required: false
Scott D Phillips [Mon, 17 Oct 2016 16:34:27 +0000 (09:34 -0700)]
meson: mark gstreamer-check-1.0 as required: false

7 years agomeson: Don't set c_std to gnu99
Nirbheek Chauhan [Sat, 15 Oct 2016 16:51:24 +0000 (22:21 +0530)]
meson: Don't set c_std to gnu99

Use the default for each compiler on every platform instead. This
improves our compatibility with compilers that don't have gnu99 as
a c_std.

7 years agomeson: Use environment object to setup test environment variables
Thibault Saunier [Mon, 3 Oct 2016 20:44:04 +0000 (17:44 -0300)]
meson: Use environment object to setup test environment variables

Bump meson requirement to 0.35

7 years agotrack-element: Avoid dereferencing NULL pointer
Thibault Saunier [Mon, 10 Oct 2016 22:59:47 +0000 (00:59 +0200)]
track-element: Avoid dereferencing NULL pointer

We set TrackElement track type very early when creating effects
so it now uses that information to find TrackElement in clips
by track type.

Reviewed-by: Alex Băluț <alexandru.balut@gmail.com>
Differential Revision: https://phabricator.freedesktop.org/D1370

7 years agomeson: Add python tests
Thibault Saunier [Tue, 13 Sep 2016 15:31:54 +0000 (12:31 -0300)]
meson: Add python tests

7 years agomeson: Setup pre commit hook and fix getpluginsdir for standalone case
Thibault Saunier [Fri, 30 Sep 2016 14:35:42 +0000 (11:35 -0300)]
meson: Setup pre commit hook and fix getpluginsdir for standalone case

7 years agomeson: update version
Tim-Philipp Müller [Fri, 30 Sep 2016 13:56:48 +0000 (14:56 +0100)]
meson: update version

7 years agoRelease 1.9.90
Sebastian Dröge [Fri, 30 Sep 2016 10:04:39 +0000 (13:04 +0300)]
Release 1.9.90

7 years agomeson: Fix gtkdoc using new meson features
Thibault Saunier [Fri, 23 Sep 2016 23:41:04 +0000 (20:41 -0300)]
meson: Fix gtkdoc using new meson features

7 years agomeson: Fix installing configured files
Thibault Saunier [Wed, 21 Sep 2016 19:41:31 +0000 (16:41 -0300)]
meson: Fix installing configured files

7 years agouriclip: Remove some filesource leftovers
Thibault Saunier [Sat, 13 Aug 2016 23:54:22 +0000 (19:54 -0400)]
uriclip: Remove some filesource leftovers

Differential Revision: https://phabricator.freedesktop.org/D1329

7 years agotests_:python: Factor out common code
Thibault Saunier [Sun, 14 Aug 2016 01:09:53 +0000 (21:09 -0400)]
tests_:python: Factor out common code

Differential Revision: https://phabricator.freedesktop.org/D1328

7 years agonle: Drop tags getting out of the composition
Thibault Saunier [Sat, 17 Sep 2016 12:46:59 +0000 (09:46 -0300)]
nle: Drop tags getting out of the composition

Those tag are meaningless in for the new stream created by the composition

First step toward fixing T3070

Differential Revision: https://phabricator.freedesktop.org/D1327

7 years agotimeline: Properly compute the end of groups when checking snapping
Thibault Saunier [Thu, 11 Aug 2016 19:12:07 +0000 (15:12 -0400)]
timeline: Properly compute the end of groups when checking snapping

Computation was not taking into account the fact that the start of
the element being moved could be at the middle of a group and not
necessarily at the start!

Fixes T7544

Reviewed-by: Alex Băluț <alexandru.balut@gmail.com>
Differential Revision: https://phabricator.freedesktop.org/D1282

7 years agoges: Handle moving groups with effects inside
Thibault Saunier [Thu, 11 Aug 2016 17:19:44 +0000 (13:19 -0400)]
ges: Handle moving groups with effects inside

We were only concidering that we should let the group handle moving
transitions when changing transitions but in fact as soon as a
transition is happenning between two clips that are in a same group
the group properly handles moving the transition, so let the
group do its job.

Fixes T7543

Differential Revision: https://phabricator.freedesktop.org/D1281

7 years agoges: Deprecate GESTimelineElement::priority writability
Thibault Saunier [Thu, 11 Aug 2016 14:54:08 +0000 (10:54 -0400)]
ges: Deprecate GESTimelineElement::priority writability

GESLayer is now responsible for setting clips priorites. Also
GESClip top effects priorities are now set by the
ges_clip_set_top_effect_index method, the user should never call
ges_timeline_element_set_priority as it will anyway be overriden
by GES itself.

Differential Revision: https://phabricator.freedesktop.org/D1280

7 years agolayer: Handle operation priorities
Thibault Saunier [Thu, 11 Aug 2016 14:36:44 +0000 (10:36 -0400)]
layer: Handle operation priorities

All operations should have higher priorites and sources should be
on top of those. We now first set the operations priorities in
a first pass and then stack sources on top of those.

Differential Revision: https://phabricator.freedesktop.org/D1279

7 years agoges: transition: Make crossfade fade out at the same time as it fade in
Thibault Saunier [Thu, 11 Aug 2016 13:53:58 +0000 (09:53 -0400)]
ges: transition: Make crossfade fade out at the same time as it fade in

Until now fade out was just fading in the new clip, but this is not
correct and crossfade should at the same time fade out while fading
in.

Fixes https://phabricator.freedesktop.org/T3451

Differential Revision: https://phabricator.freedesktop.org/D1278

7 years agolayer: Make sure to resync priorities on commit
Thibault Saunier [Thu, 11 Aug 2016 13:42:32 +0000 (09:42 -0400)]
layer: Make sure to resync priorities on commit

In case effects have been added priorites might become wrong,
but until the timeline is not commited, it does not matter.

Make sure all priorities are correct before commiting compositions

Differential Revision: https://phabricator.freedesktop.org/D1277

7 years agoFinally move clip priority handling to GESLayer.
Thibault Saunier [Thu, 11 Aug 2016 13:14:42 +0000 (09:14 -0400)]
Finally move clip priority handling to GESLayer.

Fix all tests as we now have 1 priority inside the layer
dedicated to transitions (basically no source clip will
ever have a priority of 0 inside a layer).

Differential Revision: https://phabricator.freedesktop.org/D1276

7 years agoclip: Make top effect priority inside the clip priority range
Thibault Saunier [Thu, 11 Aug 2016 12:54:23 +0000 (08:54 -0400)]
clip: Make top effect priority inside the clip priority range

And simplify the way we start computing children priority
making min_priority already relative to the clip itself.

Differential Revision: https://phabricator.freedesktop.org/D1275

7 years agotests_: timelineedition: Fix test now that we ripple from start and not from end
Thibault Saunier [Thu, 11 Aug 2016 11:54:42 +0000 (07:54 -0400)]
tests_: timelineedition: Fix test now that we ripple from start and not from end

Differential Revision: https://phabricator.freedesktop.org/D1274

7 years agonlecomposition: Don't try to seek on an empty stack
Sebastian Dröge [Thu, 22 Sep 2016 15:28:21 +0000 (11:28 -0400)]
nlecomposition: Don't try to seek on an empty stack

We would seek on a NULL pad then, which gives ugly assertions.

https://bugzilla.gnome.org/show_bug.cgi?id=771843

7 years agonleurisource: Always provide a srcpad
Sebastian Dröge [Thu, 22 Sep 2016 15:25:18 +0000 (11:25 -0400)]
nleurisource: Always provide a srcpad

By putting uridecodebin into a bin with a ghostpad. Without this,
nlesource tries to get a srcpad too early (before uridecodebin added
one) and everything fails miserably.

This has to be fixed properly in nlesource at some point, by properly
handling dynamically added pads. Currently they can only work if they
are added in states <= READY, which is not the usual case.

https://bugzilla.gnome.org/show_bug.cgi?id=771843

7 years agonlesource: Fail prepare() if no valid source pad is found
Sebastian Dröge [Wed, 21 Sep 2016 22:23:56 +0000 (18:23 -0400)]
nlesource: Fail prepare() if no valid source pad is found

https://bugzilla.gnome.org/show_bug.cgi?id=771792

7 years agoges-launch: Be a bit more agressive sanitizing arguments
Thibault Saunier [Wed, 14 Sep 2016 17:32:19 +0000 (14:32 -0300)]
ges-launch: Be a bit more agressive sanitizing arguments

Otherwise GstStructure might fail parsing some fields
containing brackets

https://bugzilla.gnome.org/show_bug.cgi?id=771434

7 years agoconfigure: Depend on gstreamer 1.9.2.1
Sebastian Dröge [Wed, 14 Sep 2016 09:31:23 +0000 (11:31 +0200)]
configure: Depend on gstreamer 1.9.2.1

7 years agotimeline: Make get_groups public
Alexandru Băluț [Tue, 6 Sep 2016 12:27:38 +0000 (14:27 +0200)]
timeline: Make get_groups public

Had to separate timeline_emit_group_added from timeline_add_group
to avoid emitting group-added when the project is being loaded.

Reviewed-by: Thibault Saunier <thibault.saunier@collabora.com>
Differential Revision: https://phabricator.freedesktop.org/D1302

7 years agotests_: Make sure child-removed is emitted when ungrouping
Alexandru Băluț [Tue, 6 Sep 2016 13:49:49 +0000 (15:49 +0200)]
tests_: Make sure child-removed is emitted when ungrouping

Reviewed-by: Thibault Saunier <thibault.saunier@collabora.com>
Differential Revision: https://phabricator.freedesktop.org/D1301

7 years agotimeline: Fix documentation
Alexandru Băluț [Tue, 6 Sep 2016 11:03:11 +0000 (13:03 +0200)]
timeline: Fix documentation

Reviewed-by: Thibault Saunier <thibault.saunier@collabora.com>
Differential Revision: https://phabricator.freedesktop.org/D1300

7 years agoRemove obsolete dependency
Alexandru Băluț [Mon, 5 Sep 2016 10:23:30 +0000 (12:23 +0200)]
Remove obsolete dependency

Reviewed-by: Thibault Saunier <thibault.saunier@collabora.com>
Differential Revision: https://phabricator.freedesktop.org/D1299

7 years agoAutomatic update of common submodule
Jan Schmidt [Sat, 10 Sep 2016 10:52:38 +0000 (20:52 +1000)]
Automatic update of common submodule

From b18d820 to f980fd9

7 years agoAutomatic update of common submodule
Jan Schmidt [Fri, 9 Sep 2016 23:58:37 +0000 (09:58 +1000)]
Automatic update of common submodule

From f49c55e to b18d820

7 years agotests: Move -DGES_TEST_FILES_PATH to common_cflags
Thibault Saunier [Fri, 9 Sep 2016 20:14:43 +0000 (17:14 -0300)]
tests: Move -DGES_TEST_FILES_PATH to common_cflags

As it is needed to build the utils lib

7 years agotests: Fix the way we get tests assets
Thibault Saunier [Fri, 9 Sep 2016 19:42:13 +0000 (16:42 -0300)]
tests: Fix the way we get tests assets

Using __FILE__ won't work properly with meson.

7 years agotest:validate: Handle new expected_failures Test argument
Thibault Saunier [Fri, 9 Sep 2016 11:52:32 +0000 (08:52 -0300)]
test:validate: Handle new expected_failures Test argument

7 years agomeson: Do not build libges against libcheck
Thibault Saunier [Wed, 7 Sep 2016 19:53:06 +0000 (16:53 -0300)]
meson: Do not build libges against libcheck

7 years agomeson: Properly find where other GStreamer plugins are when using subprojects
Thibault Saunier [Mon, 5 Sep 2016 20:55:42 +0000 (17:55 -0300)]
meson: Properly find where other GStreamer plugins are when using subprojects

7 years agoRevert "validate: Blacklist racy tests"
Thibault Saunier [Mon, 5 Sep 2016 17:54:53 +0000 (14:54 -0300)]
Revert "validate: Blacklist racy tests"

This reverts commit ce35412ff260fbd6e07b374bc3ca677053c277e0.

https://bugzilla.gnome.org/show_bug.cgi?id=769894 has been fixed

7 years agomeson: Handle building the gir file when used as subproject
Thibault Saunier [Fri, 26 Aug 2016 22:55:33 +0000 (19:55 -0300)]
meson: Handle building the gir file when used as subproject

Add support for building GIR when used as subproject
Add an option to disable GIR generation
And bump version to 1.9.2

7 years agoBack to development
Sebastian Dröge [Thu, 1 Sep 2016 09:33:22 +0000 (12:33 +0300)]
Back to development

7 years agoRelease 1.9.2
Sebastian Dröge [Thu, 1 Sep 2016 09:33:13 +0000 (12:33 +0300)]
Release 1.9.2