SVN revision: 63877
Reverted due to the following build failure:
eina_main.c:117:16: error: conflicting types for ‘_eina_tracking_lock’
../../src/include/eina_inline_lock_posix.x:83:29: note: previous declaration of ‘_eina_tracking_lock’ was here
make[3]: *** [libeina_la-eina_main.lo] Error 1
_eina_tracking_lock is used for implementation of Eina_Locks, so cannot be an Eina_Lock
Signed-off-by: Mike McCormack <mj.mccormack@samsung.com>
SVN revision: 63904
EWS is meant to run in FB or other systems without actual windows, but
as we test on X11 let's forbid to change the resolution.
One is still able to do it using ecore_evas_ews_setup().
SVN revision: 63860
The old version works because in every function in which
this macro is used ``l'' is the length and ``d'' is the
destination. This patch prevents future headaches when
those constraints no longer hold.
Patch by: Jim Kukunas <james.t.kukunas@linux.intel.com>
SVN revision: 63856
EWS is a new Ecor_Evas engine that builds on top of other engines. It
will create a backing store Ecore_Evas and ecore_evas_ews_new()
windows are created in it as images, but transparent to the outside
users (similar to buffer's ecore_evas_object_image_new()).
It provides a basic windowing system, with a known background object
that can be changed to your pleasure, and issue Ecore_Events to notify
of new windows and changes like movement, etc. Then you can write a
simple window manager based on it. (See example, Elementary will
contain one as well).
Backing store is determined by your best engine (as in
ecore_evas_new()) or specified with ecore_evas_ews_engine_set() or
environment variable $ECORE_EVAS_EWS (format:
engine-name❌y:w:h:options). The size can be set with
ecore_evas_ews_setup().
SVN revision: 63848
2011/9/30 Jaehwan Kim <jae.hwan.kim@samsung.com>:
>
> I found some bugs in edje_cc_handlers.c.
>
> One is about data.item. It is saved in hash list.
> When it is saved, I freed the memory of the key.
> But the memory is used. so I deleted the free(key).
>
> Second is it is possible in
st_collections_group_parts_part_description_state.
> So I changed the location of the conditional statement.
>
> Please apply this patch :)
>
> Jaehwan Kim.
SVN revision: 63822
NOTE: we need the help of the window manager to make this
really work. So for the moment, it half work. As soon as I
hack E17, the Xv fast path for Emotion will only work when
used with E17 and it will fallback to canvas inlined rendering
in other case.
SVN revision: 63802
Subject: [E-devel] [Patch] Evas touch event patch.
Nice to meet you.
I'm Eunmi Lee, developing mobile web browser and working on WebKit EFL port.
I need new type of event for touch, so I've made patch to add
EVAS_CALLBACK_TOUCH event to the evas.
I will explain history of this patch.
Currently, many web applications and sites use TouchEvent and they can
do everything(scrolling, zooming and so on) like native application
using TouchEvent.
So, I'm also want to provide TouchEvent for web in the WebKit EFL port,
but I got a problem during making TouchEvent because EFL's touch
event's structure (Mouse, Multi Event) is different from Web
TouchEvent's one.
Let me explain about Web TouchEvent firstly.
Web TouchEvent is consist of type and touch points list simply.
There are 3 kinds of type.
TouchStart: Happens every time a finger is placed on the screen.
TouchEnd: Happens every time a finger is removed from the screen.
TouchMove: Happens as a finger already placed on the screen is moved
across the screen.
for example, we can make (1 finger starts to touch), (2 fingers are
moving), (1 finger is released duirng 3 fingers are moving) and so on.
You can see the detailed information in the following url:
http://www.sitepen.com/blog/2008/07/10/touching-and-gesturing-on-the-iphone
However, EFL's touch event is consist of six kinds of type :
MOUSE_DOWN, MOUSE_UP, MOUSE_MOVE, MULTI_DOWN, MULTI_UP, MULTI_MOVE.
So, I have to make a converter to make web touch event from EFL's
touch event.
You can reference attatched image file : evas_touch_event.png.
To tell the truth, converting code is not a big one.
But, I want to reduce this additional job and make code simple.
In the WebKit QT port, they don't have to make converting code for
TouchEvent,
because they have QTouchEvent, it has type and touchPoints list and
they can be mapped to Web TouchEvent one by one.
I think iPhone and Android also have such kind of event.
That's all why I want to add new touch event type to the evas.
about my patch:
- EVAS_CALLBACK_TOUCH event is added
- touch_points Eina_List is added to the Evas structure to maintain
current touch lists.
- process MOUSE/MULTI UP, DOWN, MOVE to make TOUCH event.
It is my first time to modify eves codes and actually I don't know too
much about evas.
So, I will be grateful if you send any feedback and comments.
SVN revision: 63796
It seems that depending on the system, vlc can't use a NULL pointer to
the pixels where it should write its data.
So a small amount of memory should be allocated and passed to its
rendering callbacks (specifically, the lock callback) when the file is
being opened and decoded for the first time. Then this memory can be
freed, since the real rendering will happen on the shared memory area.
SVN revision: 63777
Info added includes: cursor position of the action, content related to the
action, and the action that happened (added/removed).
Also fixed tab insertion to delete the current selection.
SVN revision: 63773
arch does sse3. the wasy sse3 is done makes this stupidly ugly. so too
bad for 32bit builds - no sse3 for u unless u specifically enable at
build time and then u have a binary that never runs on pre-sse3 chips.
SVN revision: 63745
Subject: [E-devel] [Patch] support Animation gif's disposal mode
I make patch support animation gif disposal mode.
Before, gif loader only decode & render based on previous frame.
This patch can support "do not dispose mode" & "restore background
mode".
So It solve after image problem of restore background mode.
SVN revision: 63716
several problems with it... but SOMEONE... (lucas) committed it
without even so much as replying to the list saying he was going to...
:)
SVN revision: 63705
mul_256_sse3
sub4_alpha_sse3
interp4_256_sse3
mul_sym_sse3
mul4_sym_sse3
mul3_sym_sse3
LOOP_ALIGNED_U1_A48_SSE3
__attribute__((always_inline)) is needed to coax GCC (< 4.6.0)
into inlining the common blend ops. Not inlining these functions
causes a steep performance penalty.
Patch by: Jim Kukunas <james.t.kukunas@linux.intel.com>
SVN revision: 63698
getnameinfo is a blocking call would would resolve a sockaddr into
its hostname, not into an IP address, so the ip_get should use the
inet_ntop call instead.
SVN revision: 63682
Subject: [E-devel] [PATCH] update cursor location when the position of
entry is changed
The cursor location information should be transferred when the position of
entry is changed.
For example, an entry can be moved by comformant, rotate event.
The transferred cursor location is used for letting candidate word window
move the appropriate position.
SVN revision: 63660
Yunhan <spbear@gmail.com>
On Wed, Sep 21, 2011 at 10:51 PM, Kim Yunhan <spbear@gmail.com> wrote:
> Hello!
>
> elm_map uses Ecore Con with CURL.
> I tested elm_map many times on my device.
> But sometimes UI interaction is held when data connection is poor.
> So I tried to debug and I found that this code lead to hold an Ecore
main
> loop.
>
> ----------------------------------------------
> In ecore_con_url.c
>
> while (curl_multi_perform(_curlm, &still_running) ==
> CURLM_CALL_MULTI_PERFORM) ;
> ----------------------------------------------
>
> curl_multi_perform() is CURL's asynchronous API.
> But above code hold an Ecore main loop.
> When it takes long time in libcurl, UI interaction is delayed.
>
> For example, If you have poor data connection.
> libcurl is trying to resolve DNS in this step.
> But it have to wait until timeout.
> At that time it looks like being locked.
>
> So I write a patch for fixing it.
SVN revision: 63636
In C++ we can't assign a void pointer to another type pointer without casts. We
now rely on typeof() operator *when using C++*.
We may provide another version later for those compilers without typeof()
support.
SVN revision: 63568
allocate memory for list variables & rectangles unless we need to (ie:
If there are no evas_render_updates to do, then there is no need to
allocate extra list variables & rectangles).
SVN revision: 63528
Reported by WooHyun. Thanks a lot, great catch, also told me where and what
the issue is exactly.
Also added a test to verify this works.
SVN revision: 63493
This possibly doesn't scale as good but it's good enough for everything I've
tried. It's a lot easier to maintain comparing to the rbtree, and takes a
lot less memory. Next step is probably changing the array size according
to the actual content of the textblock.
SVN revision: 63474
Otherwise ecore-evas will think that the event is for multi touch
and will ignore the actual values of the structure and will only use
the uninitialized data from the multi struct
SVN revision: 63431
The new function emotion_object_keep_aspect_set() allow to set a policy
that will automatically change the emotion object border property, based
on the video aspect ratio and object current aspect ratio.
Possible values are:
* EMOTION_ASPECT_KEEP_NONE
* EMOTION_ASPECT_KEEP_WIDTH
* EMOTION_ASPECT_KEEP_HEIGHT
* EMOTION_ASPECT_KEEP_BOTH
* EMOTION_ASPECT_CROP
* EMOTION_ASPECT_CUSTOM
SVN revision: 63416
This is a simple API that implements colors for the borders specified by
emotion_object_border_set(), using a background rectangle behind the
emotion object.
SVN revision: 63415
The emotion_object_border_set/get APIs were added, allowing one to
specify a positive (border) or negative (crop) value for each of the
edjes of the video being played. This allows easier setup on an Emotion
object to keep the aspect ratio, either by adding black areas on it, or
by cropping the exceding area of the video.
NOTE: The specified border/crop is based on the original size of the
video. If it is scaled (emotion object size is different from the video
size), the borders will be scaled as well.
Patch based on an original patch sent by
Hugo Camboulive <hugo.camboulive@gmail.com>.
SVN revision: 63414