Commit Graph

2977 Commits

Author SHA1 Message Date
Emmanuele Bassi
f25cdf066c [actor] Allow unsetting width and height separately
It is possible to unset the size of an actor specified with set_width()
and set_height() by using:

  clutter_actor_set_size (actor, -1, -1);

Which works by unsetting the :min-*-set and the :natural-*-set properties.

Calling set_width(-1) and set_height(-1) separately, though, doesn't work
thus implicitly breaking the assumption that set_size() is nothing more
than set_width()+set_height(). This was obviously due to the face that
pre-1.0 set_width() and set_height() took an unsigned integer as an
argument.
2009-08-13 16:13:07 +01:00
Colin Walters
d3e06bc1e5 Add (allow-none) for clutter_stage_set_key_focus()
The @actor argument can be NULL to unset the key focus.

Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-08-13 10:49:39 +01:00
Emmanuele Bassi
9766fb1300 [script] Fix Actor's ad-hoc parser
ClutterActor parses positional and dimensional properties with a
custom deserializer. We need to:

  - handle G_TYPE_INT64, the default integer type for JSON-GLib
  - use G_TYPE_FLOAT for properties, since Actor switched to it
    for the pixel-based ones

This makes ClutterScript work again.
2009-08-12 16:02:23 +01:00
Emmanuele Bassi
cfbbacb807 [script] Fix the default deserialization
Properties parsed from ClutterScript should now use G_TYPE_INT64
and automatically convert between G_TYPE_INT and G_TYPE_INT64.
2009-08-12 16:01:25 +01:00
Emmanuele Bassi
4d52245637 [json] Update the internal copy of JSON-GLib
JSON-GLib switch to 64 bit integers upstream, to avoid the undefined
integer size of the JSON RFC. All integers are automatically promoted
to gint64.
2009-08-12 16:00:13 +01:00
Emmanuele Bassi
2b70b0d4ed Include clutter-json.h, not json-types.h
The json-types.h header is found by the mere fact of it being
in the project; if we are compiling against the system JSON-GLib
this could be horribly out of date.

We need to use clutter-json.h, which will include the right
header for us.
2009-08-12 15:57:25 +01:00
Emmanuele Bassi
cc6aa6e4d5 Clutter GIR depends on ClutterJson GIR conditionally
If we're building Clutter with a system JSON-GLib we need to skip
the typelib compilation of ClutterJson since we depend on Json-1.0.gir
directly.
2009-08-12 15:32:18 +01:00
Emmanuele Bassi
6fcf5c1820 [json] Include only json-glib.h
JSON-GLib moved to a single include scheme, so we should only include
json-glib.h. If we use the internal copy it doesn't matter, since the
header does the right thing.
2009-08-12 15:31:16 +01:00
Emmanuele Bassi
8c49e22753 Add copyright and license notice
The clutter-script-parser.c does not have a copyright and license
notices; even though the LGPL is a per-project license and not a
per-file license, having those notices in every source file is a
good idea.
2009-08-12 10:30:11 +01:00
zhou jiangwei
13e055a351 [eglx] Update the EGLX backend
The EGLX backend needs to be updated to the new ClutterBackendX11
code.

Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-08-11 14:19:24 +01:00
zhou jiangwei
2ab9bef587 [cogl] Fix the GLES 2.0 backend
Update the GLES 2.0 backend to the new COGL API after the breakage
introduced by the journal.

Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-08-11 14:19:24 +01:00
Robert Bragg
dbfd3f0f44 [cogl_clear] Also pass the alpha component of the CoglColor to glClearColor
Previously we were passing a constant of 0.0 for the alpha component which
wasn't as flexible.
2009-08-11 14:15:44 +01:00
Emmanuele Bassi
f0cb337d92 [build] Fix LDADD -> LIBADD usage 2009-08-11 13:27:25 +01:00
Emmanuele Bassi
dfc32b60c4 [osx] Clean up Makefile.am
The OS X backend Makefile.am was missing a line concatenation, and
so the -xobjective-c directive was always ignored.

Instead of dumping everything into INCLUDES and LDADD we should follow
what the rest of the backends do, and use per-target CFLAGS and LDADD,
and reserve the INCLUDES to -D and -I directives.

Thanks to: Christian Hergert <chris@dronelabs.com>
2009-08-11 13:11:34 +01:00
Emmanuele Bassi
d0748b3b9b Merge branch 'system-json-glib'
* system-json-glib:
  [script] Simplify the parser code
  Update JSON-GLib dependency
2009-08-10 17:21:20 +01:00
Emmanuele Bassi
d263e7a78b [build] Install JSON headers inside the right directory
Don't install inside the clutter-MAJOR_MINOR/ directory, but use
the API_VERSION (1.0).

Otherwise we'd have the Clutter headers for 1.x inside:

        $includedir/clutter-1.0/clutter

And the JSON-related headers inside:

        $includedir/clutter-1.<minor>/clutter
2009-08-10 17:19:04 +01:00
Emmanuele Bassi
19a53054fd Bump to 1.1.1 after branching 1.0 2009-08-07 15:27:25 +01:00
Emmanuele Bassi
420ed353b5 [x11] Do not ask to destroy an empty Window
The fix for bug 1750 inside commit b190448e made Clutter-GTK spew
BadWindow errors. The reason for that is that we call XDestroyWindow()
without checking if the old Window is None; this happens if we call
clutter_x11_set_stage_foreign() on a new ClutterStage before it has
been realized.

Since Clutter-GTK does not need to realize the Stage it is going to
embed anymore (the only reason for that was to obtain a proper Visual
but now there's ClutterBackendX11 API for that), the set_stage_foreign()
call is effectively setting the StageX11 Window for the first time.
2009-08-07 08:26:18 +01:00
Emmanuele Bassi
335fb07ab0 [actor] Add missing return_if_fail()
The queue_relayout() method was missing the argument check.

Thanks to: Neil Patel <njpatel@gmail.com>
2009-08-05 17:50:31 +01:00
Neil Roberts
1434d72134 Small doc fix to clutter_text_get_color
Bug reported by Rob Bradford.
2009-08-05 15:51:49 +01:00
Emmanuele Bassi
b190448ed4 [x11] Destroy the Window we own when setting a foreign one
When we replace the stage Window using a foreign one we also need to
destroy the Window we created, if needed, to avoid leaking resources
all around.

Fixes bug:

  http://bugzilla.openedhand.com/show_bug.cgi?id=1750
2009-08-04 16:59:15 +01:00
Emmanuele Bassi
7963254c23 [docs] Close a tag 2009-08-04 11:39:09 +01:00
Emmanuele Bassi
79288b46ed [docs] Fix a missing parameter 2009-08-04 11:38:48 +01:00
Emmanuele Bassi
9cda8d130f [actor] Be more explicit when warning about invariants
The "catch all" warning for a the mapped invariant violation is too
generic: it doesn't tell you why the invariant was broken in case
we are trying to map an unparented actor - e.g. through a Clone.
2009-08-04 11:15:05 +01:00
Emmanuele Bassi
f8a964b422 [cogl] Expose cogl_is_vertex_buffer()
The function is public and available in the shared object, but
it's not present in the header, so this is not an API/ABI change.
2009-08-04 11:15:05 +01:00
Robert Bragg
c9f57452e6 [build] remove reference to light1.png in tests/interactive/Makefile.am
This file doesn't exist in the repo and isn't referenced by any unit tests.
2009-08-03 15:16:42 +01:00
Robert Bragg
71c4fa0f73 [build] dist tests/interactive/wrapper.sh
in tests/interactive/Makefile.am add wrapper.sh to EXTRA_DIST otherwise
interactive unit tests wont be runnable when building from distributed
tarballs.
2009-08-03 15:08:13 +01:00
Damien Lespiau
c9c87543e8 [debug] CLUTTER_{NOTE,MARK,DBG,GLERR,TIMESTAMP} should always be statements
In case someone tries to do:
	if (foo)
		CLUTTER_NOTE(DOMAIN, "%s", blah);

CLUTTER_NOTE () should always be a statement
2009-08-03 14:25:09 +01:00
Damien Lespiau
cd35f3b384 [debug] Fix __GNUC__ typo
The right gcc define is __GNUC__ not __GNUC_. This typo had the side
effect that we were using the non gcc specific debug macros leading to
a less optmised CLUTTER_NOTE () than one could have dreamed of.
2009-08-03 14:25:09 +01:00
Damien Lespiau
94b35a4a5e [build] GCC_FLAGS is no more, MAINTAINER_CFLAGS is the One.
CC_FLAGS was a left over from the 0.8.x age. Killed it and use the
newer MAINTAINER_CFLAGS variable instead.
2009-08-03 14:25:08 +01:00
Damien Lespiau
4ee49eded0 [gitignore] Ignore new test-color-hls-roundtrip test 2009-08-03 14:25:08 +01:00
Emmanuele Bassi
b2ab2b9535 [actor] Use the right vertex
The vertex that should be used by the apply_relative_transform
is the one passed in as const, and the result should be placed
inside the non-const ClutterVertext. Currently, we are using
the latter, and thus the function is completely useless.
2009-08-02 18:34:17 +01:00
Emmanuele Bassi
ae2a861a0b [x11] Force a redraw before mapping the stage
Does as it says on the tin: run a paint cycle before calling
XMapWindow().
2009-07-31 10:29:33 +01:00
Joshua Lock
abdcf4debb [docs] Update building instructions for OS X
Recommend OS X users install the dependencies using MacPorts and build
Clutter themselves.

Signed-off-by: Emmanuele Bassi <ebassi@linux.intel.com>
2009-07-31 10:28:50 +01:00
Neil Roberts
b3842a6b62 [mingw] Update the README
The README has been updated to say that you can use the script under
MSYS and also to contain a pointer to OAH. It also contains the
updated contents of a blog post¹ about building with MSYS.

1. http://www.busydoingnothing.co.uk/blog/2008/07/13#ClutterWin32
2009-07-29 19:06:36 +01:00
Neil Roberts
8ad446f3e9 [mingw-cross-compile.sh] Fix to use git rather than svn
The script includes a part to automatically checkout the Clutter
source and build it but it was using the old svn repo instead of git.
2009-07-29 18:13:19 +01:00
Neil Roberts
d796d3c88c [mingw-cross-compile.sh] Add a note about MSYS in the comment
The comment previously implied that you could only use the script to
cross compile but it works just as well under MSYS.
2009-07-29 18:07:40 +01:00
Neil Roberts
1e14d3041d [mingw-cross-compile.sh] Update version numbers of all the dependencies
This updates the URLs for the binary dependencies to use the latest
versions.
2009-07-29 18:07:21 +01:00
Emmanuele Bassi
a3c323c52c [docs] Remove mentions of Clutter-Cairo
The mingw README file still mentions Clutter-Cairo as a known issue.
2009-07-29 17:22:18 +01:00
Emmanuele Bassi
0ec6793738 [mingw/win32] Remove clutter-cairo
The Clutter-Cairo library has been removed, so there's no point
in checking it out -- from Subversion, even.
2009-07-29 17:21:13 +01:00
Emmanuele Bassi
fba1045bcb Post-release bump to 1.0.1 2009-07-28 18:39:24 +01:00
Emmanuele Bassi
f245f3049d [release] 1.0.0 2009-07-28 13:57:46 +01:00
Emmanuele Bassi
93ecd6d0c4 Update the NEWS 2009-07-28 13:57:46 +01:00
Emmanuele Bassi
77fdd7494e Fix compiler warnings when COGL debug level=minimum 2009-07-28 13:57:46 +01:00
Emmanuele Bassi
85b69237a9 [docs] Point the TODO to Bugzilla 2009-07-28 13:46:11 +01:00
Emmanuele Bassi
c683079167 [docs] Update requirements and release notes 2009-07-28 13:46:11 +01:00
Emmanuele Bassi
a5d6b23308 [units] Add binding-friendly initializers
We should follow the convention for boxed types initializers of:

  <type_name>_from_<another_type> (boxed, value)

For ClutterUnits as well; so:

  clutter_units_pixels -> clutter_units_from_pixels
  clutter_units_em     -> clutter_units_from_em
  ...

We should still keep the short-hand version as a macro, though.
2009-07-28 13:46:11 +01:00
Emmanuele Bassi
57baa2f0bc [animation] Fix variable use
Instead of using the proper typed variables we were overwriting
the gpointer for the Interval's final value.
2009-07-28 13:21:57 +01:00
Neil Roberts
f4ad3761c4 [win32] Remove the call to clutter_redraw in clutter_stage_win32_show
Since commit 7b811f8b this is done in clutter_stage_show so there is
no need for it in the backends.
2009-07-28 12:13:44 +01:00
Neil Roberts
df586eb507 [win32] Move the show/hide to be implementations of the interface
This makes clutter_stage_win32_show/hide be implementations of
ClutterStageWindowIface rather than overriding the methods in
ClutterActor. This reflects the changes in e4ff24bc for the X11
backend.
2009-07-28 12:13:44 +01:00