platform/upstream/enlightenment.git
9 years agosend client/screen/pointer geometry to visibility effect edje
Mike Blumenkrantz [Mon, 15 Jun 2015 17:47:33 +0000 (13:47 -0400)]
send client/screen/pointer geometry to visibility effect edje

9 years agounify comp object visibility effect animation code
Mike Blumenkrantz [Mon, 15 Jun 2015 17:47:01 +0000 (13:47 -0400)]
unify comp object visibility effect animation code

9 years agoadd flag for detecting whether an effect is currently set on a comp object
Mike Blumenkrantz [Mon, 15 Jun 2015 17:44:47 +0000 (13:44 -0400)]
add flag for detecting whether an effect is currently set on a comp object

also improve refcounting for effect apis and add return values for use in
determining whether an effect was successfully set/started

9 years agoadd flag to E_Comp_Object for detecting show animation state
Mike Blumenkrantz [Mon, 15 Jun 2015 17:41:42 +0000 (13:41 -0400)]
add flag to E_Comp_Object for detecting show animation state

this allows improvements to the code which provides hide animations,
allowing clients to begin hiding during their show animations instead
of rendering a black rectangle

9 years agouse a safe iterator when deleting clients during comp shutdown
Mike Blumenkrantz [Mon, 15 Jun 2015 17:35:34 +0000 (13:35 -0400)]
use a safe iterator when deleting clients during comp shutdown

it's conceivable that, were there a bug in client refcounting,
this could become an infinite loop and prevent shutdown/restart.
since, at this point, we don't really care about deleting anything,
ensure that the loop will end

9 years agoe - e client, comp win etc. reffing fix ... part 2
Carsten Haitzler (Rasterman) [Mon, 15 Jun 2015 11:27:25 +0000 (20:27 +0900)]
e - e client, comp win etc. reffing fix ... part 2

this follows 56cabf59c672af27af87eda82079de4e2ee2d38e then
4e5521b4d82063a2fbf6632a117fda201cc1687d where i have been trying to
fix a crash with e client and comp win references etc. i have gone
over all referencing with a fine tooth comb and found all the nigglies
i can., no leaks now, no crashes, no valgrind complaints etc. so i
call this fixed now. as best i know this is new in e20, so not a
backport fix

9 years agoe - bring back fullscreen crash bug while fixing refcount correctness
Carsten Haitzler (Rasterman) [Sun, 14 Jun 2015 10:16:09 +0000 (19:16 +0900)]
e - bring back fullscreen crash bug while fixing refcount correctness

refs were inconsistent - thus this fixed the fullscreen quit bug by
never freeing a client. this brings the bug back by fixing this client
leak. i'll look again at this later.

9 years agoe - fix crash in desktop lookup+fixup where exe_inst is null
Carsten Haitzler (Rasterman) [Sun, 14 Jun 2015 08:50:14 +0000 (17:50 +0900)]
e - fix crash in desktop lookup+fixup where exe_inst is null

9 years agoe comp - fix refcounting for e_comp and comp clients
Carsten Haitzler (Rasterman) [Sun, 14 Jun 2015 07:06:37 +0000 (16:06 +0900)]
e comp - fix refcounting for e_comp and comp clients

the refcoutning for e_comp and e comp clients seemed to be a bit off -
i read over every ref and unref carefully and fix it. this leads to
the com-_data being null (properly now), so now check for that too.

9 years agoe - mixer - allow to 150% volume with some "stickynedd" from 100 to 120
Carsten Haitzler (Rasterman) [Sun, 14 Jun 2015 06:01:44 +0000 (15:01 +0900)]
e - mixer - allow to 150% volume with some "stickynedd" from 100 to 120

9 years agofixed window focus and keyboard input issues
Toan Pham [Fri, 12 Jun 2015 00:04:41 +0000 (20:04 -0400)]
fixed window focus and keyboard input issues

Summary: fixed window focus and keyboard input issues if the to-be-focused window is in an iconized state

Reviewers: zmike

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2644

9 years agoChanged search pattern to focus window by name to case-insensitive.
Toan Pham [Thu, 11 Jun 2015 18:38:10 +0000 (14:38 -0400)]
Changed search pattern to focus window by name to case-insensitive.

Summary: Sometimes a window has a name that's not like the application name, for example.  chromium-browser application will have a window name: Chromium-browser.  Most users will try to match the window name with "chromium-browser", but it wont work b/c the e's window name match is case sensitive.  Most users, would not know that the window name of chrome is "Chromium-browser", so it is pretty much impossible for them to setup a focus-to-window-name binding properly.

Reviewers: zmike

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2645

9 years agorename visibility effect struct member effect -> visibility_effect
Mike Blumenkrantz [Thu, 11 Jun 2015 16:47:04 +0000 (12:47 -0400)]
rename visibility effect struct member effect -> visibility_effect

feeble attempt at making future added effects more intelligible

9 years agoadd compositor visibility effects and matching configuration
Mike Blumenkrantz [Thu, 11 Jun 2015 15:35:44 +0000 (11:35 -0400)]
add compositor visibility effects and matching configuration

these are specific types of animation for use when toggling window visibility.
they combine with existing compositor window animations to provide nicer integration
for very specific types of windows

see https://www.youtube.com/watch?v=hIVdd0Z2K00 for a demo

9 years agoallow comp matching for UNKNOWN type windows
Mike Blumenkrantz [Thu, 11 Jun 2015 15:33:45 +0000 (11:33 -0400)]
allow comp matching for UNKNOWN type windows

sometimes this is a valid type and we want to match it

9 years agoUpdating italian translation
maxerba [Thu, 11 Jun 2015 12:31:24 +0000 (14:31 +0200)]
Updating italian translation

9 years agoe menu - an extra object ref causes menus to never be freed
Carsten Haitzler (Rasterman) [Wed, 10 Jun 2015 12:32:01 +0000 (21:32 +0900)]
e menu - an extra object ref causes menus to never be freed

this has been in e for ages - someone not noticed, but this fixes
visual artifacts of left over menus on the top-left. this extra ref
really makes no sense. it's not like this ref is then accomoanied by a
matching unref somewhere else (after much debugging).

@fix

9 years agowl-desktop-shell: Fix popup menu placement for xdg surfaces
Chris Michael [Tue, 9 Jun 2015 13:43:34 +0000 (09:43 -0400)]
wl-desktop-shell: Fix popup menu placement for xdg surfaces

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoe - mixer - clean up popup a bit with mute next to slider etc.
Carsten Haitzler (Rasterman) [Tue, 9 Jun 2015 08:22:40 +0000 (17:22 +0900)]
e - mixer - clean up popup a bit with mute next to slider etc.

9 years agoe mixer - fix segv on desklock if default sink not set yet
Carsten Haitzler (Rasterman) [Tue, 9 Jun 2015 07:18:24 +0000 (16:18 +0900)]
e mixer - fix segv on desklock if default sink not set yet

9 years agomixer - dont make list scrollable horizontally for nicer look
Carsten Haitzler (Rasterman) [Tue, 9 Jun 2015 07:17:07 +0000 (16:17 +0900)]
mixer - dont make list scrollable horizontally for nicer look

9 years agoRevert "Revert "fix mixer gcc class name""
Carsten Haitzler (Rasterman) [Tue, 9 Jun 2015 05:39:15 +0000 (14:39 +0900)]
Revert "Revert "fix mixer gcc class name""

This reverts commit 6bacf2070532d0bdf22afa4d303ac41a0ce52047.

oops - i was wrong. totally misread that patch i thinkg it changed
mixer -> emix

9 years agoe - shelf menu - handle deletion properly if cb is not for shelf menu
Carsten Haitzler (Rasterman) [Tue, 9 Jun 2015 05:29:25 +0000 (14:29 +0900)]
e - shelf menu - handle deletion properly if cb is not for shelf menu

it seems that some changes now make the shel menu post callback be
called for older menus not part of the shelf and thus shelf menu
stored != menu the cb is for - thus resulting in deletion of the wrong
menu

9 years agoe - mixer - don't do anything on desklock if no mixer context exists
Carsten Haitzler (Rasterman) [Tue, 9 Jun 2015 02:10:35 +0000 (11:10 +0900)]
e - mixer - don't do anything on desklock if no mixer context exists

9 years agoRevert "fix mixer gcc class name"
Carsten Haitzler (Rasterman) [Tue, 9 Jun 2015 02:09:12 +0000 (11:09 +0900)]
Revert "fix mixer gcc class name"

This reverts commit 718b3206cbc480b607bed771bbe679c2d312fe16.

no - this REPLACEs the mixer module. the same old mixer gadget that
was originally in e now will be replaced by this new epulse/emixer
gadget. thus the "mixer" gadcon name.

9 years agofix mixer gcc class name
Mike Blumenkrantz [Mon, 8 Jun 2015 19:02:24 +0000 (15:02 -0400)]
fix mixer gcc class name

9 years agoadd gadcon visibility handling, send hide signal to pager desks on hide
Mike Blumenkrantz [Mon, 8 Jun 2015 18:53:31 +0000 (14:53 -0400)]
add gadcon visibility handling, send hide signal to pager desks on hide

this (in combination with elm changes) should greatly reduce cpu usage

9 years agofix shelf visibility state signals
Mike Blumenkrantz [Mon, 8 Jun 2015 18:52:40 +0000 (14:52 -0400)]
fix shelf visibility state signals

previous behavior resulted in hide being sent when animations began
instead of at end

9 years agocompletely freeze all deskmirror object events/animations during hide
Mike Blumenkrantz [Mon, 8 Jun 2015 18:51:58 +0000 (14:51 -0400)]
completely freeze all deskmirror object events/animations during hide

9 years agodisable animations on clients during hide
Mike Blumenkrantz [Mon, 8 Jun 2015 18:47:42 +0000 (14:47 -0400)]
disable animations on clients during hide

9 years agoonly freeze/thaw edje objects on non-input clients
Mike Blumenkrantz [Mon, 8 Jun 2015 18:46:33 +0000 (14:46 -0400)]
only freeze/thaw edje objects on non-input clients

9 years agofix eo errors for ilist icon select state
Mike Blumenkrantz [Mon, 8 Jun 2015 16:35:16 +0000 (12:35 -0400)]
fix eo errors for ilist icon select state

9 years agowl-desktop-shell: Fix T2357: Update xdg_shell for version 5
Chris Michael [Mon, 8 Jun 2015 16:23:40 +0000 (12:23 -0400)]
wl-desktop-shell: Fix T2357: Update xdg_shell for version 5

Summary: This fixes the Enlightenment wayland desktop shell by
updating it to support xdg_shell version 5. Fixes T2357 where gtk apps
would not work. Confirmed working now.

@fix

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoupdate po files
Carsten Haitzler (Rasterman) [Mon, 8 Jun 2015 12:14:46 +0000 (21:14 +0900)]
update po files

9 years agoe mixer - replace with epulse/emixer
Carsten Haitzler (Rasterman) [Mon, 8 Jun 2015 12:00:23 +0000 (21:00 +0900)]
e mixer - replace with epulse/emixer

this is emixer (epulse) from

http://git.enlightenment.org/devs/ceolin/epulse.git

the emixer binar is rewritten though and the emix lib is compiled-in
into the module and into the binary as oppopsed to a shared lib with
loadable modules. this supports alsa and pulse. a much more solid mixer.

9 years agofix leak in randr2 when matching clone modes
Carsten Haitzler (Rasterman) [Mon, 8 Jun 2015 11:53:13 +0000 (20:53 +0900)]
fix leak in randr2 when matching clone modes

9 years agoe desklock - listen to randr screen changes and re-do locks
Carsten Haitzler (Rasterman) [Mon, 1 Jun 2015 09:38:31 +0000 (18:38 +0900)]
e desklock - listen to randr screen changes and re-do locks

fix due to new randr2 stuff

9 years agolower minimum backlight clamp to 5%
Mike Blumenkrantz [Fri, 5 Jun 2015 13:13:23 +0000 (09:13 -0400)]
lower minimum backlight clamp to 5%

sometimes 10% is still way too bright

9 years agoenlightenment deskenv support - fix dialog labelling and desktop name
Carsten Haitzler (Rasterman) [Thu, 28 May 2015 02:41:28 +0000 (11:41 +0900)]
enlightenment deskenv support - fix dialog labelling and desktop name

1. Enlightenment, not X-Enlihgtenment for the desktop. we're a proper DE name
2. Have a fixed desktop env list and efreet gives very... odd results
(eg X-Geequie because some rogue desktop file decides it's good to
have this in it :)).

9 years agoe start - path manipulation - don't add a : if there already is one
Carsten Haitzler (Rasterman) [Tue, 26 May 2015 23:43:13 +0000 (08:43 +0900)]
e start - path manipulation - don't add a : if there already is one

avoids adding more : chars to make a broken path even worse

9 years agoResize the canvas at the end of canvas init
Derek Foreman [Fri, 22 May 2015 16:24:07 +0000 (12:24 -0400)]
Resize the canvas at the end of canvas init

Summary:
wl_drm sets up a resize callback, so if the resize occurs before zone setup
wl_drm's call to e_comp_canvas_update will setup extra zones (breaking the
wizard).

Reviewers: zmike, devilhorns

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2567

 #butts

9 years agofix focusing client on mouse events which trigger actions (with exceptions)
Mike Blumenkrantz [Thu, 21 May 2015 16:31:00 +0000 (12:31 -0400)]
fix focusing client on mouse events which trigger actions (with exceptions)

block focus events only if focus has changed to another client or the current
client's layer-actual has been lowered

regression from 719b2baeba8133a021663e8281e6fd53662e8b16

9 years agoe - fix profiles to have correct adjust value for bl adjust in bindings
Carsten Haitzler (Rasterman) [Thu, 21 May 2015 04:43:00 +0000 (13:43 +0900)]
e - fix profiles to have correct adjust value for bl adjust in bindings

@fix

9 years agoe randr2 - add some logic for knowing when to ask to configure a screen
Carsten Haitzler (Rasterman) [Sat, 16 May 2015 14:18:32 +0000 (23:18 +0900)]
e randr2 - add some logic for knowing when to ask to configure a screen

start of a "what do you want to do with this screen (clone, extend etc.)"
feature

9 years agoe - randr2 - filter to the "best" common mode match on clone config
Carsten Haitzler (Rasterman) [Sat, 16 May 2015 13:56:01 +0000 (22:56 +0900)]
e - randr2 - filter to the "best" common mode match on clone config

this should fix T2100

9 years agouse layout for theme import dialog
Mike Blumenkrantz [Wed, 20 May 2015 12:18:48 +0000 (08:18 -0400)]
use layout for theme import dialog

fixes min sizing

9 years agoremove legacy entry widget pointer setting code
Mike Blumenkrantz [Wed, 20 May 2015 12:11:59 +0000 (08:11 -0400)]
remove legacy entry widget pointer setting code

fix T2424

9 years agoProperly fix T2182 (add a --version option to enlightenment)
Chris Michael [Mon, 18 May 2015 18:41:45 +0000 (14:41 -0400)]
Properly fix T2182 (add a --version option to enlightenment)

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoallow starting enlightenment with the valgrind option again
Chris Michael [Mon, 18 May 2015 18:22:26 +0000 (14:22 -0400)]
allow starting enlightenment with the valgrind option again

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoallow new fullscreening clients to replace current nocomp where applicable
Mike Blumenkrantz [Mon, 18 May 2015 17:19:56 +0000 (13:19 -0400)]
allow new fullscreening clients to replace current nocomp where applicable

9 years agoforce xwindow stacking to obey nocomp policy and stack below current nocomp
Mike Blumenkrantz [Mon, 18 May 2015 17:19:33 +0000 (13:19 -0400)]
force xwindow stacking to obey nocomp policy and stack below current nocomp

fix T2277

9 years agoreject client fullscreening based on config and desk visibility
Mike Blumenkrantz [Mon, 18 May 2015 17:00:23 +0000 (13:00 -0400)]
reject client fullscreening based on config and desk visibility

9 years agofix nocomp popping on comp input ungrabbing
Mike Blumenkrantz [Mon, 18 May 2015 16:11:36 +0000 (12:11 -0400)]
fix nocomp popping on comp input ungrabbing

regression from e24666b9296d1972f320d89218ce9990e9c44420

9 years agoe_comp_wl: fix not sending a mouse enter event to client.
Boram Park [Mon, 18 May 2015 15:53:05 +0000 (11:53 -0400)]
e_comp_wl: fix not sending a mouse enter event to client.

Summary:
First of all, currently cdata->ptr.x,y contains e_client related pos.
But, cdata is owned by e_comp, not per e_client. So cdata->ptr.x,y should contain
absolute pos.
And, when a pointer resource is created, enter event should be sent to client.

Change-Id: I21cb031e293fe281e35ba89f3a96116a28a48856
Signed-off-by: Boram Park <boram1288.park@samsung.com>
Test Plan:
1. run e as wayland server
2. move cursor to around 50,50
3. run elementary_test
4. click the mouse left button on slide toggle widget.
   (Don't move a cursor out of elementary_test. If you do, you can't find bug)

Reviewers: gwanglim, raster, devilhorns, ManMower, zmike

Reviewed By: zmike

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2330

9 years agofix call to e_comp_wl_input_keymap_set for building with
Chris Michael [Mon, 18 May 2015 15:48:31 +0000 (11:48 -0400)]
fix call to e_comp_wl_input_keymap_set for building with
wayland-client support

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoremove cpp warning
Chris Michael [Mon, 18 May 2015 15:42:51 +0000 (11:42 -0400)]
remove cpp warning

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoELM_WIN_DIALOG_BASIC is broken in wl right now, so for wl create
Chris Michael [Mon, 18 May 2015 15:32:09 +0000 (11:32 -0400)]
ELM_WIN_DIALOG_BASIC is broken in wl right now, so for wl create
normal elm windows for now.

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoremove unused variable
Chris Michael [Mon, 18 May 2015 13:24:01 +0000 (09:24 -0400)]
remove unused variable

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoadd missing EINA_UNUSED for function paramater
Chris Michael [Mon, 18 May 2015 13:22:39 +0000 (09:22 -0400)]
add missing EINA_UNUSED for function paramater

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoFix call to e_comp_wl_input_keymap_set
Chris Michael [Mon, 18 May 2015 13:21:03 +0000 (09:21 -0400)]
Fix call to e_comp_wl_input_keymap_set

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoadd missing EINA_UNUSED for function paramater
Chris Michael [Mon, 18 May 2015 13:20:16 +0000 (09:20 -0400)]
add missing EINA_UNUSED for function paramater

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agowayland: Remove a lot of E_Comp_Data pointers
Derek Foreman [Mon, 18 May 2015 13:08:52 +0000 (09:08 -0400)]
wayland: Remove a lot of E_Comp_Data pointers

Summary: We should just use e_comp->wl_comp_data everywhere we can.

Reviewers: devilhorns, zmike

Reviewed By: zmike

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2507

9 years agoe_menu: remove an unnecessary comparison
MinJeong Kim [Mon, 18 May 2015 12:36:09 +0000 (08:36 -0400)]
e_menu: remove an unnecessary comparison

Summary:
Remove an unnecessary comparison.
It fixes coverity CID 1291841 Same on both sides.
@fix

Reviewers: raster, devilhorns, zmike

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2523

9 years agowayland-compositor: Set server mode
Derek Foreman [Fri, 15 May 2015 18:44:21 +0000 (14:44 -0400)]
wayland-compositor: Set server mode

Summary:
If we don't set server mode we can deadlock when trying to use functions
that way for server mode to become set.

This can currently cause a startup freeze in e_scale_init().

Reviewers: zmike, devilhorns

Reviewed By: devilhorns

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2520

9 years agoadded a macro and a fucntion to switch to desktop D on screen S as specified in https...
Nicolas Marcotte [Fri, 15 May 2015 07:00:31 +0000 (16:00 +0900)]
added a macro and a fucntion to switch to desktop D on screen S as specified in https://phab.enlightenment.org/T2218

Summary: Merge branch 'master' of https://git.enlightenment.org/core/enlightenment initnal implementation of https://phab.enlightenment.org/T2218

Reviewers: raster

Subscribers: raster, cedric

Differential Revision: https://phab.enlightenment.org/D2231

Fixed:
Modified patch to use existing macros for walking all screens.

9 years agowl-drm: Don't enable/disable an output until after we have set the current mode
Chris Michael [Thu, 14 May 2015 12:38:13 +0000 (08:38 -0400)]
wl-drm: Don't enable/disable an output until after we have set the current mode

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agodrm randr: More work on getting RandR functional in E-Wl. This adds a
Chris Michael [Wed, 13 May 2015 18:44:51 +0000 (14:44 -0400)]
drm randr: More work on getting RandR functional in E-Wl. This adds a
function to remove an output from e_comp_wl also.

NB: Not entirely complete yet

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoconf_randr: Reduce debug output noise
Chris Michael [Wed, 13 May 2015 18:43:26 +0000 (14:43 -0400)]
conf_randr: Reduce debug output noise

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agowl-desktop-shell: Quiet debug output
Chris Michael [Wed, 13 May 2015 16:56:18 +0000 (12:56 -0400)]
wl-desktop-shell: Quiet debug output

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoremove unused 'serial' variable in mulit-move callback
Chris Michael [Wed, 13 May 2015 17:20:33 +0000 (13:20 -0400)]
remove unused 'serial' variable in mulit-move callback

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agofix numerous compile errors from D2482
Chris Michael [Wed, 13 May 2015 16:26:07 +0000 (12:26 -0400)]
fix numerous compile errors from D2482

NB: Mental Note: COMPILE BEFORE PUSH !!

 # butts

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoSupport wl_touch and send wl_touch events to client
jhyuni.kang [Wed, 13 May 2015 16:05:57 +0000 (12:05 -0400)]
Support wl_touch and send wl_touch events to client

Summary:
Currently enlightenment-wayland support pointer and keyboard events.(in input)
         So I added EVAS_CALLBACK_MULTI_* event handler and hanling functions for wl_touch interface.
         This code send wl_touch events to client like a pointer event.

Test Plan:
In wayland + enlightenment environment, generate multi touch events.
           The enlightenment will be send wl_touch events to client.

Reviewers: raster, devilhorns, zmike

Reviewed By: devilhorns, zmike

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2482

9 years agosend screen paramaters in proper order to e_comp_wl_output_init
Chris Michael [Wed, 13 May 2015 16:05:17 +0000 (12:05 -0400)]
send screen paramaters in proper order to e_comp_wl_output_init

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoe_comp_wl: fix segfault in case parent's surface is destroyed before subsurface is...
Boram Park [Tue, 12 May 2015 18:43:00 +0000 (14:43 -0400)]
e_comp_wl: fix segfault in case parent's surface is destroyed before subsurface is destroyed

Summary:
A parent surface can be destroyed. In this case, ec->comp_data->sub.data->parent has
a wrong address which has already destroyed. Then, it occurs segmentation fault.
To avoid segfault, ec->comp_data->sub.data->parent should be NULL when a parent surface
is destroyed.

Signed-off-by: Boram Park <boram1288.park@samsung.com>
Test Plan: Run attached test program(subsurface)

Reviewers: gwanglim, cedric, devilhorns, zmike

Reviewed By: devilhorns, zmike

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2484

9 years agoremove str(n)dupa usages
Daniel Kolesa [Mon, 11 May 2015 15:33:42 +0000 (16:33 +0100)]
remove str(n)dupa usages

str(n)dupa are GNU extensions that duplicate a string, using an alloca'd
buffer. This patch removes their definitions from e.h (which should only
contain E's own API, without fallback definitions for libc functions)
which were wrong anyway (they failed in cases where str(n)dupa was an
actual function, not a macro).

Instead, we replace them depending on context with alloca+memcpy+strlen
or a static buffer (used in contexts where we are sure that the buffer
will contain the string entirely)

@fix

9 years agoFix misspelling in e_client.h
Yomi [Thu, 7 May 2015 23:01:04 +0000 (19:01 -0400)]
Fix misspelling in e_client.h

Summary: Fix misspelling in e_client.h

Reviewers: zmike

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2477

9 years agoFix misspelling in e_randr2.c
Yomi [Thu, 7 May 2015 22:41:10 +0000 (18:41 -0400)]
Fix misspelling in e_randr2.c

Summary: Fix misspelling in e_randr2.c

Reviewers: zmike

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2476

9 years agowl-drm: Start on making the RandR config dialog usable in drm
Chris Michael [Thu, 7 May 2015 19:44:50 +0000 (15:44 -0400)]
wl-drm: Start on making the RandR config dialog usable in drm

NB: 'Apply' does not function yet, needs another API or 2 in ecore_drm

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agomake e_acpi_lid_is_closed function callable from wl_drm module
Chris Michael [Thu, 7 May 2015 19:44:22 +0000 (15:44 -0400)]
make e_acpi_lid_is_closed function callable from wl_drm module

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoenlightenment: Make E build again with EFL from git
Chris Michael [Thu, 7 May 2015 19:13:10 +0000 (15:13 -0400)]
enlightenment: Make E build again with EFL from git

sed -i 's/EAPI/E_API/g'

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoFix T2385 : internal windows not showing resize cursor
Chris Michael [Mon, 4 May 2015 16:40:06 +0000 (12:40 -0400)]
Fix T2385 : internal windows not showing resize cursor

If we are going to set the internal elm windows to borderless and
handle frames inside E, then we need to also unset any existing
pointer objects which may be on the wayland Ecore_Evas.

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoCorrect misspelling.
Yomi [Sun, 3 May 2015 21:45:17 +0000 (17:45 -0400)]
Correct misspelling.

Summary:
Correct misspelling.
Changes: Fix misspelling in e_desk.c

     src/bin/e_desk.c

Reviewers: zmike, raster

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2466

9 years agoe_ipc: fix creation of IPC server when $DISPLAY is a path
Jean Guyomarc'h [Sat, 2 May 2015 14:04:28 +0000 (10:04 -0400)]
e_ipc: fix creation of IPC server when $DISPLAY is a path

Summary:
Some implementations of the X server (e.g. Xquartz) define their $DISPLAY
as a path. Since ecore_ipc_server_add() does not create non-existant
directories, and since it may not worth to hide the socket in a complex
path, this patch aims at reducing the $DISPLAY by only keeping its basename.

Reviewers: zmike

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D2465

9 years agofix infinite loop when changing ibar sources
Mike Blumenkrantz [Thu, 30 Apr 2015 21:04:47 +0000 (17:04 -0400)]
fix infinite loop when changing ibar sources

fix T1951

9 years agoonly apply fwin focus to main efm object if no icon is editing
Mike Blumenkrantz [Thu, 30 Apr 2015 20:06:09 +0000 (16:06 -0400)]
only apply fwin focus to main efm object if no icon is editing

9 years agoadd efm function for getting an editing icon
Mike Blumenkrantz [Thu, 30 Apr 2015 20:05:55 +0000 (16:05 -0400)]
add efm function for getting an editing icon

9 years agodon't destroy efm icon entry on client focus-in if icon's client is being focused
Mike Blumenkrantz [Thu, 30 Apr 2015 20:05:21 +0000 (16:05 -0400)]
don't destroy efm icon entry on client focus-in if icon's client is being focused

9 years agofix comp input ungrabbing to read more clearly and not pop nocomp wrongly
Mike Blumenkrantz [Thu, 30 Apr 2015 18:50:20 +0000 (14:50 -0400)]
fix comp input ungrabbing to read more clearly and not pop nocomp wrongly

9 years agoprune module config list when loading a module
Mike Blumenkrantz [Thu, 30 Apr 2015 18:39:28 +0000 (14:39 -0400)]
prune module config list when loading a module

it's somehow possible to get multiple instances of a module in the config.
since I don't know whether this is a current or past problem, prune the list
on every module load to remove subsequent instances of the module

fix T2181

9 years agofix notification text to conform to spec and libnotify implementation
Mike Blumenkrantz [Thu, 30 Apr 2015 18:11:06 +0000 (14:11 -0400)]
fix notification text to conform to spec and libnotify implementation

fix T2053
fix T2187

9 years agotry hicolor icon theme as fallback for fdo icon setting if user's theme fails
Mike Blumenkrantz [Thu, 30 Apr 2015 16:32:31 +0000 (12:32 -0400)]
try hicolor icon theme as fallback for fdo icon setting if user's theme fails

fix T2142

9 years agochange default icon theme to hicolor
Mike Blumenkrantz [Thu, 30 Apr 2015 16:31:37 +0000 (12:31 -0400)]
change default icon theme to hicolor

ref T2142

9 years agoforce visibility on clients before attempting to hide them
Mike Blumenkrantz [Wed, 29 Apr 2015 21:21:14 +0000 (17:21 -0400)]
force visibility on clients before attempting to hide them

this ensures that edje visibility states get reset in the event of toggling
visibility extremely quickly

probably fixes T2260

9 years agoonly apply mouse down focus events if a binding did not run on the same event
Mike Blumenkrantz [Wed, 29 Apr 2015 21:15:43 +0000 (17:15 -0400)]
only apply mouse down focus events if a binding did not run on the same event

fixes the case of enabling "raise on click" and also having a mouse binding
for "lower on click" active

fix T2364

9 years agofreeze evas events on clients during keyboard move/resize actions
Mike Blumenkrantz [Wed, 29 Apr 2015 21:02:36 +0000 (17:02 -0400)]
freeze evas events on clients during keyboard move/resize actions

fix T2112

9 years agorename client menu move/resize items to indicate real function
Mike Blumenkrantz [Wed, 29 Apr 2015 21:02:16 +0000 (17:02 -0400)]
rename client menu move/resize items to indicate real function

ref T2112

9 years agofreeze comp object edjes while not hidden
Mike Blumenkrantz [Wed, 29 Apr 2015 16:55:02 +0000 (12:55 -0400)]
freeze comp object edjes while not hidden

requested by cedric

9 years agoe_place - dont need extrra zone check here anymore
Carsten Haitzler (Rasterman) [Mon, 27 Apr 2015 22:51:11 +0000 (07:51 +0900)]
e_place - dont need extrra zone check here anymore

9 years agoe_shelf - check visiblity with zone and desk FIRST not later
Carsten Haitzler (Rasterman) [Mon, 27 Apr 2015 22:48:16 +0000 (07:48 +0900)]
e_shelf - check visiblity with zone and desk FIRST not later

es->cfg->desk_show_mode if false (default) implies a shelf is always
visible - true, as it's sticky (visible on all desktops). but this
didn't check zone. check zone first THEN this.

9 years agoadd focus callback to override elm_win being stupid
Mike Blumenkrantz [Mon, 27 Apr 2015 20:03:42 +0000 (16:03 -0400)]
add focus callback to override elm_win being stupid

plz elm_win, you can't focus yourself. you aren't a real widget.