from their git repo. This changes the way events are dispatched, and
adds the new wl_registry listeners in place of
wl_display_add_global_listener. Also fixes calls to non-existant
wl_display_iterate. Fixes binding the various Wayland interfaces
(compositor, shell, etc). Fixes binding of output and input.
Signed-off-by: Christopher Michael <cp.michael@samsung.com>
SVN revision: 78050
"frame object". The resulting smart object is set to the frame object.
Add the Smart_Data Frame Text to be part of the Smart Object.
Signed-off-by: Christopher Michael <cp.michael@samsung.com>
SVN revision: 78048
ecore_poll.c: In function ‘_poller_constructor’:
ecore_poll.c:273:10: warning: ‘return’ with a value, in function returning void
Signed-off-by: Mike McCormack <mikem@atratus.org>
SVN revision: 77997
Subject: [E-devel] [ecore] Recursive rm
I looked at the ecore_file_recursive_rm() function
(in ecore/src/lib/ecore_file/), and I wonder why this
is so ugly/complicated.
We are doing a readlink() and two stat(). Why not simply
do a lstat() ?
It takes less memory and it's simplier. Here is a patch.
(And readlink() is ssize_t)
SVN revision: 77979
From: Vikram Narayanan <vikram186@gmail.com>
Date: Sun, 14 Oct 2012 22:57:05 +0530
Subject: [PATCH] ecore/ecore_directfb: Fix a pedantic bug
Though someone can argue it is correct, (as the two structures have the
same size and numbers) it gives a room for a little confusion. Fix this.
Signed-off-by: Vikram Narayanan <vikram186@gmail.com>
SVN revision: 77971
Subject: [E-devel] [Patch] [Ecore] Allow to send SelectionNotify event
if there is no data from XGetWindowProperty.
The issue I want to fix is that elementary clipboard does not return
paste callback if selection does not contain any data.
Currently when ecore_x_window_prop_property_get function is called
asking for AnyProperty it returns 0 if there is no data (number of
items stored equals 0). This causes ecore_x_event_handle_selection
_notify to exit before SelectionNotify event is added and eventually
elementary paste callback is not fired.
My patch removes this condition and changes behaviour of this function
by allowing to add SelectionNotify event even if data returned is empty.
SVN revision: 77732
This is the method necessary to synchronize Wayland and Evas rendering, similar
to vblank sync that is used in X11. This solves the Wayland SHM render issues
(tearing effect) without the need to add double buffering to the engine.
In a near future, the SHM implementation will feature double (or triple)
buffering and this syncrhonization mechanism will be changed to be optional.
Patch based on the work started by Robert Bradford on ticket #1280http://trac.enlightenment.org/e/ticket/1280
Signed-off-by: Eduardo Lima (Etrunko) <eduardo.lima@intel.com>
SVN revision: 77580
I did not do changelog, news or backport here as it has zero effect on the
functionality. Just makes the code easier to udnerstand and read. If you think
it needs to be done neverless howl and let me know what you actually want in
it. :)
SVN revision: 77441
when the screensaver cycle interval passes. That is not what we are
interested in here. We only need to know when the screensaver
activates/deactivates.
NB: This cuts cpu usage down incredibly when screensavers are set with
a short timeout.
SVN revision: 77310
Removed unecessary ensure_pool_size function move the checks to _pool_create.
Signed-off-by: Eduardo Lima (Etrunko) <eduardo.lima@intel.com>
SVN revision: 77097
pool
_ecore_evas_wl_buffer_free() and _ecore_evas_wl_shm_pool_free() were added,
avoid repeating the same calls everywhere.
Signed-off-by: Eduardo Lima (Etrunko) <eduardo.lima@intel.com>
SVN revision: 77096
Subject: Re: [E-devel] [Patch] [Ecore] Patch with NULL checks to some
functions in ecore module
I had prepared patch for some Ecore functions which, in my opinion,
critically need NULL checks. I had tried to keep each file convention.
SVN revision: 76852
Subject: [E-devel] [Patch] [Ecore] Protection for giving below zero
values for ecore_poll and ecore_timer
Please also check attached patch for ecore_timer_interval_set and
ecore_poller_poll_interval_set.
It checks if given time is not zero or below zero.
SVN revision: 76805
Subject: [E-devel] [Patch] [Ecore] ecore_file_escape_name fix for \n
and \t
Please review and consider the attached patch for ecore_file_escape_name
function, now it correctly escapes tab and new line characters.
SVN revision: 76803
Timeout messages are given a 408 HTTP status code (request timeout), instead
of a CRIT() message; this gives chance to the completed callback do whatever
is necessary when a timeout happens.
SVN revision: 76419
ecore_x_randr to not work for pretty much everyone. Short version,
don't memcpy something potentially Larger into something Certainly
smaller. (read on for the details).
NB: This 'should' fix all the randr problems in ecore_x (tho I have
not tested Everything).
NB: Ok, here goes:
XRRGetScreenResources returns a struct. Inside that struct is a
list of RROutputs.
RROutput is defined as (from randrproto.h):
#define RROutput CARD32
CARD32 is defined as (from X11/Xmd.h):
# ifdef LONG64
typedef unsigned long CARD64;
typedef unsigned int CARD32;
# else
typedef unsigned long CARD32;
# endif
so CARD32 can change based on the system (32/64 bit).
Ecore_X_Randr_Output is defined as (Ecore_X.h):
typedef Ecore_X_ID Ecore_X_Randr_Output;
(for reference: typedef unsigned int Ecore_X_ID)
Double bonus points if you have already spotted the problem !! ;)
SVN revision: 76306
NB: Based on initial work from Alex <zhiwen.wu@intel.com> (thanks for
the patch), but modified to clean up compiler warnings, fixed function
declarations, fixed function namespacing.
SVN revision: 75809
Christopher Michael
Carsten Haitzler
Mike McCormack
rustyBSD
Vikram Narayanan
Jihoon Kim
Cedric BAIL
Michal Pakula vel Rutka
Yakov Goldberg
Sebastian Dransfeld
Eduardo Lima (Etrunko)
Vincent Torri
Stefan Schmidt
Christopher Michael
Patryk Kaczmarek
Mike Blumenkrantz
Leandro Pereira
Jiyoun Park
Nicolas Aguirre