platform/upstream/enlightenment.git
9 years agoe exec tracker - fix tracking to update desktop files on efreet change
Carsten Haitzler (Rasterman) [Mon, 2 Nov 2015 03:28:27 +0000 (12:28 +0900)]
e exec tracker - fix tracking to update desktop files on efreet change

efreet will send change events meaning out desktop file icons are
invalid now. we have refs but content is junk. update these to new
desktop ptr handles by using the exec key to look up new ones on the
change event. this fixes ibar gainign a bunch of blank unknown icons
when any desktop files are touched on the system or user dirs

@fix

9 years agosimplify client SMART/EXPAND maximize
Mike Blumenkrantz [Fri, 30 Oct 2015 19:03:29 +0000 (15:03 -0400)]
simplify client SMART/EXPAND maximize

calling e_client_resize_limit() here was misleading since the values
were never used

9 years agoadjust non-x11 client's saved frame geometry when changing csd
Mike Blumenkrantz [Thu, 29 Oct 2015 18:33:15 +0000 (14:33 -0400)]
adjust non-x11 client's saved frame geometry when changing csd

this seems to be a more comprehensive solution for retaining previous
window sizes after toggling various csd-affecting window states in wayland

9 years agoRevert "when maximizing a wayland client with csd, use window geometry"
Mike Blumenkrantz [Thu, 29 Oct 2015 18:21:42 +0000 (14:21 -0400)]
Revert "when maximizing a wayland client with csd, use window geometry"

This reverts commit 6e5e51c2b8647a41f156e2854be409780492a679.

not a complete solution, makes client code somewhat unreadable

9 years agoadd E_COMP_ENGINE env var to determine sw/gl
Mike Blumenkrantz [Wed, 28 Oct 2015 16:55:09 +0000 (12:55 -0400)]
add E_COMP_ENGINE env var to determine sw/gl

accepted values are 1, 2, sw, gl

toggling config options is a hassle when debugging

9 years agoclean up some systray leaks
Mike Blumenkrantz [Mon, 26 Oct 2015 17:11:32 +0000 (13:11 -0400)]
clean up some systray leaks

9 years agoalways unalias x11 client parent window <-> pixmap relation during delete
Mike Blumenkrantz [Mon, 26 Oct 2015 17:02:34 +0000 (13:02 -0400)]
always unalias x11 client parent window <-> pixmap relation during delete

deleted state is not relevant, this should always be removed during deletion
to prevent further referencing of a deleted client

fix T2803

9 years agomodules:mixer: Avoid pa backend from crash
Flavio Ceolin [Sat, 24 Oct 2015 02:22:48 +0000 (00:22 -0200)]
modules:mixer: Avoid pa backend from crash

Check the pulseaudio's context error and if info is not null before
get server's information.

9 years agoe - ibar - fix clash when 2 icons of the same path get added somehow
Carsten Haitzler (Rasterman) [Sat, 24 Oct 2015 02:01:56 +0000 (11:01 +0900)]
e - ibar - fix clash when 2 icons of the same path get added somehow

i encontered a situation where the icon_hash contained a garbage entry
- had been freed already. the only way i can see this happening is if
the desktop file changed path during runtime thus the icon was never
removed from hash on free as string didnt match. store string used
when adding to hash so removal is guarannteed to work and also for
good measure protect against double-adding (and generate a new string
for storage using timestamp which should be unique).

so this fixes a crash i was just looping on.

@fix

9 years agounset release mode
Mike Blumenkrantz [Fri, 23 Oct 2015 18:04:42 +0000 (14:04 -0400)]
unset release mode

9 years ago20.0-beta NEWS updates
Mike Blumenkrantz [Fri, 23 Oct 2015 17:32:29 +0000 (13:32 -0400)]
20.0-beta NEWS updates

9 years ago20.0-beta release
Mike Blumenkrantz [Thu, 1 Oct 2015 17:42:45 +0000 (13:42 -0400)]
20.0-beta release

9 years agoadd NEWS from 19.12 release
Mike Blumenkrantz [Fri, 23 Oct 2015 17:28:15 +0000 (13:28 -0400)]
add NEWS from 19.12 release

9 years agowhen maximizing a wayland client with csd, use window geometry
Mike Blumenkrantz [Fri, 23 Oct 2015 16:56:56 +0000 (12:56 -0400)]
when maximizing a wayland client with csd, use window geometry

wayland clients which have csd must be resized according to window geometry,
not client (surface) geometry. this is somewhat tricky to handle because x11
clients which have csd work the exact opposite way and must continue to be
managed using client geometry

this is not my ideal solution for this issue, but I can't think of a
better one at this time which fully fixes wayland client maximization

9 years agoreject frame theme changes for clients which have CSD
Mike Blumenkrantz [Thu, 22 Oct 2015 23:27:01 +0000 (19:27 -0400)]
reject frame theme changes for clients which have CSD

9 years agoonly apply frame geometry deltas for CSD if the CSD region exists
Mike Blumenkrantz [Thu, 22 Oct 2015 23:25:07 +0000 (19:25 -0400)]
only apply frame geometry deltas for CSD if the CSD region exists

ref e931e86161d7f419e44652b533fea4dc9b16241b

9 years agoset client input rects more accurately, accounting for SSD
Mike Blumenkrantz [Thu, 22 Oct 2015 23:24:11 +0000 (19:24 -0400)]
set client input rects more accurately, accounting for SSD

9 years agoenforce cw->obj pass_events state more accurately
Mike Blumenkrantz [Thu, 22 Oct 2015 23:22:48 +0000 (19:22 -0400)]
enforce cw->obj pass_events state more accurately

changing the edje file of cw->shobj can change the pass_events state
of its child object, cw->obj

9 years agoclip client input rect to client geometry, not window geometry
Mike Blumenkrantz [Thu, 22 Oct 2015 22:44:51 +0000 (18:44 -0400)]
clip client input rect to client geometry, not window geometry

when using csd, the window geometry will be smaller than the surface,
resulting in an unusually small input region

9 years agodefer all module error dialogs
Mike Blumenkrantz [Thu, 22 Oct 2015 18:14:20 +0000 (14:14 -0400)]
defer all module error dialogs

due to the (current) synchronous creation of wayland surfaces in efl,
creating these dialogs during startup will result in a crash. the only
way to prevent this and still display the dialogs is to wait a bit and
show the dialogs after init has completed

9 years agocompletely remove access module
Mike Blumenkrantz [Thu, 22 Oct 2015 17:50:18 +0000 (13:50 -0400)]
completely remove access module

as promised almost a year ago, this module has been unmaintained to
the point where it cannot function in any way, and attempting to keep
it in the tree is a futile effort

this module can now be found at devs/discomfitor/e_module-access.git

ref 77717523f3e6c327ddd797ad064a97c467e6f90b

9 years agoremove dead modules from whitelist
Mike Blumenkrantz [Thu, 22 Oct 2015 17:18:14 +0000 (13:18 -0400)]
remove dead modules from whitelist

9 years agounconditionally show opengl option in compositor settings
Mike Blumenkrantz [Thu, 22 Oct 2015 16:49:51 +0000 (12:49 -0400)]
unconditionally show opengl option in compositor settings

enlightenment is not always running in x11, so making this option
dependent on xlib will break other environments

9 years agoadd hacks to work around pulseaudio+xwayland integration deadlocks
Mike Blumenkrantz [Wed, 21 Oct 2015 21:17:49 +0000 (17:17 -0400)]
add hacks to work around pulseaudio+xwayland integration deadlocks

enlightenment is (I think) the first wayland compositor to run with
in-process pulseaudio integration for audio playback and not just mixer
support. hooray.

this results in a fun issue: if DISPLAY is set, as it must be for x11
clients to function, pulseaudio will unconditionally attempt to use a
blocking socket connection to create a connection to the running xserver.
the only exception here is if x11 support has been compiled out of pulseaudio,
but probably no distro will do that ever.

so, what happens when the compositor thread tries to create a socket connection
to the xserver that the compositor thread has not yet started? absolutely nothing.
forever.

the easiest solution which continues to provide the key press sounds that everyone
loves is to ensure that the pulseaudio connection is created before DISPLAY is ever
set, namely in the xwayland module init.

this will now occur automatically now in the case when the mixer module detects
pulseaudio support.

TL;DR: don't disable mixer module if you use xwayland

9 years agorequire efl >= 1.16 for wayland support
Mike Blumenkrantz [Wed, 21 Oct 2015 19:58:07 +0000 (15:58 -0400)]
require efl >= 1.16 for wayland support

the fact is that nobody developing this has tested with 1.15.x recently,
and almost none of the changes have been backported to that branch. better
to require a newer version than to present a failing user experience

9 years agomake comp config unresizable...again
Mike Blumenkrantz [Wed, 21 Oct 2015 19:17:25 +0000 (15:17 -0400)]
make comp config unresizable...again

ref 5178197f602da6cff94e03f20c522d5b1a631ae7

9 years agodo not use ecore-x functions in scale config while running in wayland
Mike Blumenkrantz [Wed, 21 Oct 2015 18:20:24 +0000 (14:20 -0400)]
do not use ecore-x functions in scale config while running in wayland

9 years agoappend new keyboard resources for wayland focused-client to the focused list
Mike Blumenkrantz [Wed, 21 Oct 2015 18:05:02 +0000 (14:05 -0400)]
append new keyboard resources for wayland focused-client to the focused list

fixes all cases where applications would fail to receive focus initially
and then be permanently broken

9 years agofix typo in wayland keyboard resource unbind list management
Mike Blumenkrantz [Wed, 21 Oct 2015 17:49:13 +0000 (13:49 -0400)]
fix typo in wayland keyboard resource unbind list management

9 years agodo not use x11 input windows for grab dialogs when running in wayland
Mike Blumenkrantz [Wed, 21 Oct 2015 16:23:44 +0000 (12:23 -0400)]
do not use x11 input windows for grab dialogs when running in wayland

9 years agoFix e_menu trying to swallow an object from a different canvas
Chris Michael [Wed, 21 Oct 2015 14:26:18 +0000 (10:26 -0400)]
Fix e_menu trying to swallow an object from a different canvas

_e_menu_realize was trying to swallow the menu->container_object into
the menu->bg_object, but the menu->bg_object was being created on the
compositor canvas, and the container object was being created on the
e_comp->elm_win.

With recent EFL changes, this causes an abort inside Evas.

To fix this, set the menu->evas to be the Evas from the e_comp->elm_win.

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoupdate csd using deltas of previous values
Mike Blumenkrantz [Mon, 19 Oct 2015 21:19:32 +0000 (17:19 -0400)]
update csd using deltas of previous values

when applying new csd to a window which already has csd, the previous
csd must be removed in order to apply any new csd offsets in order to
avoid unwanted moving/resizing

9 years agoe-comp-wl-input: don't leak the initial keymap.
Seunghun Lee [Mon, 19 Oct 2015 15:43:07 +0000 (11:43 -0400)]
e-comp-wl-input: don't leak the initial keymap.

Summary: remove xkb_map_ref() for keymap in _e_comp_wl_input_keymap_update().

Test Plan: N/A

Reviewers: devilhorns, zmike

Subscribers: ManMower, cedric

Differential Revision: https://phab.enlightenment.org/D3173

9 years agomouse_config: fix accel threshold and numerator limits.
Thierry [Fri, 16 Oct 2015 20:29:19 +0000 (13:29 -0700)]
mouse_config: fix accel threshold and numerator limits.

man xset =>  If the `threshold' parameter is provided and 0, the
`acceleration' parameter will be used in the exponent of a more
natural and continuous formula, giving precise control for slow motion
but big reach for fast motion, and a progressive transition for motions
in between.  Recommended `acceleration' value in this case is 3/2 to 3,
but not limited to that range.

9 years agoqueue pending buffers for wayland shm clients
Mike Blumenkrantz [Fri, 16 Oct 2015 19:45:23 +0000 (15:45 -0400)]
queue pending buffers for wayland shm clients

due to how deferred rendering works, it's possible for a client to
send a second buffer before enlightenment has rendered the first one.
in this situation, it seems that the best solution is to queue successive
buffers (frames) and pop the queue after each render

ref T2784

9 years agoforce a software render in all cases when finalizing x11 client iconify
Mike Blumenkrantz [Thu, 15 Oct 2015 19:33:28 +0000 (15:33 -0400)]
force a software render in all cases when finalizing x11 client iconify

since ICCCM requires that clients be unmapped while iconified, it's necessary
for the compositor to perform one last render prior to the unmap in order to
ensure that mirror objects will still appear as expected. this render must use
the pixmap buffer data in order to avoid timing issues due to async/deferred
rendering, and it is only necessary for the case of clients rendering with a
native surface

fix T2788

9 years agoclear x11 client pixmap when returning from iconic
Mike Blumenkrantz [Thu, 15 Oct 2015 19:32:38 +0000 (15:32 -0400)]
clear x11 client pixmap when returning from iconic

ensure pixmap remains usable during the entire hide process

9 years agocheck systray dbus name existence before attempting to recover session
Mike Blumenkrantz [Thu, 15 Oct 2015 17:10:58 +0000 (13:10 -0400)]
check systray dbus name existence before attempting to recover session

in the case that a system error occurs the dbus method for returning the
session name can fail, leading to a null return

CID 1327399

9 years agouse dbus unique name for tracking systray sessions
Mike Blumenkrantz [Thu, 15 Oct 2015 16:58:32 +0000 (12:58 -0400)]
use dbus unique name for tracking systray sessions

the env var DBUS_SESSION_BUS_ADDRESS may not be set for all platforms
where enlightenment+dbus are usable, so it's necessary to get the dbus
id to ensure that this value exists and is properly tracked

9 years agoreject deleted clients from x11 sync alarm event callback
Mike Blumenkrantz [Thu, 15 Oct 2015 16:40:00 +0000 (12:40 -0400)]
reject deleted clients from x11 sync alarm event callback

E_Client->comp_data is null after a client has been deleted, so
attempting to handle events which require the dereferencing of this
pointer after a client has been deleted will result in a crash

these events should be rejected after delete regardless, since at this
time the compositor has stopped handling events for the client

ref f42c6aa1871cca6c6ffb39b65e2bfa3815bff35c

9 years agoe - comps sync - ec client was null - i think, so protect against crash
Carsten Haitzler (Rasterman) [Thu, 15 Oct 2015 10:26:00 +0000 (19:26 +0900)]
e - comps sync - ec client was null - i think, so protect against crash

i got a crash here and the bt was broken and i couldnt check if
_e_comp_x_client_data_get() returned null, but it's the only thing
that would make sense, so protect against this to avoid a crash. as
this was a one-off, i can't find out more,

@fix

9 years agosystray: search for the best available resolution of a icon
Marcel Hollerbach [Wed, 14 Oct 2015 10:18:41 +0000 (12:18 +0200)]
systray: search for the best available resolution of a icon

For example in telegram the first icon which was transmitted was in a
pretty bad resolution, the new code searches the best res.

9 years agocache systray items for each dbus session
Mike Blumenkrantz [Wed, 14 Oct 2015 20:30:41 +0000 (16:30 -0400)]
cache systray items for each dbus session

the current spec does not directly require any behavior from clients
when a systray host, leading to an issue where clients do not re-register
their items when a new host appears

when using an in-process systray watcher, as the current implementation does,
the best choice for maintaining consistency for systray items across restarts
is to cache them according to the current dbus session. the process of setting
up the item will validate it on subsequent restarts, and changes to the session
will clear the cache

fix T2786

9 years ago e_flowlayout: typecast return value to int
Amitesh Singh [Wed, 14 Oct 2015 18:02:07 +0000 (14:02 -0400)]
 e_flowlayout: typecast return value to int

Summary:
return value is small enough, so there is no risk in casting from unsigned int to signed int.

CID: 1267211

Reviewers: devilhorns, raster, cedric, zmike

Subscribers: seoz, sachin.dev

Differential Revision: https://phab.enlightenment.org/D3179

9 years agotrap shaped x11 clients and prevent compositor blocking with high rect count
Mike Blumenkrantz [Wed, 14 Oct 2015 16:51:45 +0000 (12:51 -0400)]
trap shaped x11 clients and prevent compositor blocking with high rect count

in the case where a shaped window with many rects exists, there is a high
probability of the damage rect count being huge, leading to massive blocking for
each frame as the compositor attempts to fetch all of these rects from the xserver.

instead, the compositor can shortcut this by forcing a full-window damage any time
the rect count is sufficiently high, trading a blocking socket operation for some
amount of (potential) overdraw.

testing in affected scenarios has shown huge improvements: where previously the entire
compositor would lock up, things work as expected now

see https://bugzilla.mozilla.org/show_bug.cgi?id=1214746 for a sample case

9 years agodo not return when x11 damage event returns no rects
Mike Blumenkrantz [Wed, 14 Oct 2015 16:50:17 +0000 (12:50 -0400)]
do not return when x11 damage event returns no rects

ensure first_damage flag is set in all cases

9 years agopulseaudio: pass big data as pass by reference
Amitesh Singh [Wed, 14 Oct 2015 16:21:00 +0000 (21:51 +0530)]
pulseaudio: pass big data as pass by reference

CID: 1308394

9 years agoenlightenment: Make sure we have a zone before calling
Chris Michael [Wed, 14 Oct 2015 14:59:31 +0000 (10:59 -0400)]
enlightenment: Make sure we have a zone before calling
_bgpreview_viewport_update

@fix CID1324753

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoenlightenment: Use 'fabs' to compute absolute value of floating point
Chris Michael [Wed, 14 Oct 2015 14:38:14 +0000 (10:38 -0400)]
enlightenment: Use 'fabs' to compute absolute value of floating point
types

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoenlightenment: #if 0 out an unused variable
Chris Michael [Wed, 14 Oct 2015 14:30:33 +0000 (10:30 -0400)]
enlightenment: #if 0 out an unused variable

NB: The function which actually uses this variable is also #if 0'd out

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoenlightenment: Use 'fabs' to get absolute value of floating point types
Chris Michael [Wed, 14 Oct 2015 14:28:38 +0000 (10:28 -0400)]
enlightenment: Use 'fabs' to get absolute value of floating point types

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoenlightenment: Fix E_CONFIG_LIMIT to use the proper min & max values
Chris Michael [Wed, 14 Oct 2015 14:22:19 +0000 (10:22 -0400)]
enlightenment: Fix E_CONFIG_LIMIT to use the proper min & max values

e_config->backlight.idle_dim is actually an unsigned char in the
structure, so use a 0 & 1 for min & max

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoenlightenment: Add missing field initializers for Eina_Rectangle
Chris Michael [Wed, 14 Oct 2015 14:19:43 +0000 (10:19 -0400)]
enlightenment: Add missing field initializers for Eina_Rectangle

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoenlightenment: Add missing field initializers for Evas_Point
Chris Michael [Wed, 14 Oct 2015 14:18:04 +0000 (10:18 -0400)]
enlightenment: Add missing field initializers for Evas_Point

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoenlightenment: Add missing EINA_UNUSED for unused function paramaters
Chris Michael [Wed, 14 Oct 2015 14:17:22 +0000 (10:17 -0400)]
enlightenment: Add missing EINA_UNUSED for unused function paramaters

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoenlightenment: Add missing EINA_UNUSED for unused function parameters
Chris Michael [Wed, 14 Oct 2015 14:16:25 +0000 (10:16 -0400)]
enlightenment: Add missing EINA_UNUSED for unused function parameters

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoenlightenment: Add missing EINA_UNUSED for unused function paramaters
Chris Michael [Wed, 14 Oct 2015 14:14:16 +0000 (10:14 -0400)]
enlightenment: Add missing EINA_UNUSED for unused function paramaters

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agosubtract x11 client damage region even when forcing full override damage
Mike Blumenkrantz [Wed, 14 Oct 2015 02:38:50 +0000 (22:38 -0400)]
subtract x11 client damage region even when forcing full override damage

fix T2785

9 years agoforce full damage for first frame of x11 override clients
Mike Blumenkrantz [Tue, 13 Oct 2015 21:07:30 +0000 (17:07 -0400)]
force full damage for first frame of x11 override clients

it seems that damages for popup windows in some applications,
namely blink-based browsers, are reported incorrectly, resulting
in sometimes having a blank window

9 years agoFix make distcheck
Marcel Hollerbach [Tue, 13 Oct 2015 19:58:21 +0000 (21:58 +0200)]
Fix make distcheck

9 years agomixer: start app from bin directory
Marcel Hollerbach [Tue, 13 Oct 2015 18:59:45 +0000 (20:59 +0200)]
mixer: start app from bin directory

before the app was installed into the modules directory, now its
installed to the bin directory

9 years agomixer: install icon and .desktop file
Marcel Hollerbach [Tue, 13 Oct 2015 18:55:50 +0000 (20:55 +0200)]
mixer: install icon and .desktop file

9 years agomixer: install app to bin directory
Marcel Hollerbach [Tue, 13 Oct 2015 18:55:15 +0000 (20:55 +0200)]
mixer: install app to bin directory

9 years agouse correct param ordering for setting wayland client frame geometry
Mike Blumenkrantz [Tue, 13 Oct 2015 17:31:47 +0000 (13:31 -0400)]
use correct param ordering for setting wayland client frame geometry

function uses l,r,t,b and not x,y,xx,yy

fix T2624

9 years agoremove duplicate surface size update for wayland clients during commit
Mike Blumenkrantz [Tue, 13 Oct 2015 17:25:15 +0000 (13:25 -0400)]
remove duplicate surface size update for wayland clients during commit

this has already happened in e_comp_wl_surface_attach()

9 years agodon't forget wl_desktop_shell's header in tarball
Tomáš Čech [Mon, 12 Oct 2015 15:56:45 +0000 (11:56 -0400)]
don't forget wl_desktop_shell's header in tarball

Summary: Merge branch 'master' of git://git.enlightenment.org/core/enlightenment

Reviewers: zmike

Subscribers: cedric

Maniphest Tasks: T2776

Differential Revision: https://phab.enlightenment.org/D3170

9 years agoe mixer - be more robust if pulse exits/crashes and def sink is null
Carsten Haitzler (Rasterman) [Sat, 10 Oct 2015 02:07:24 +0000 (11:07 +0900)]
e mixer - be more robust if pulse exits/crashes and def sink is null

ensure there are protections everywhere for when default sink is null
- mixer did crash when pulse crashed and a popup was to be shown...

9 years agoexplicitly disable screensaver for wayland urgency wakeups
Mike Blumenkrantz [Fri, 9 Oct 2015 19:23:15 +0000 (15:23 -0400)]
explicitly disable screensaver for wayland urgency wakeups

9 years agomake wake-on-urgent apply for all cases of window urgency
Mike Blumenkrantz [Fri, 9 Oct 2015 19:21:41 +0000 (15:21 -0400)]
make wake-on-urgent apply for all cases of window urgency

in the case that an urgent window is set as urgent again, wakeup should
still occur

9 years agoapply _GTK_FRAME_EXTENTS for xwayland clients
Mike Blumenkrantz [Fri, 9 Oct 2015 18:20:25 +0000 (14:20 -0400)]
apply _GTK_FRAME_EXTENTS for xwayland clients

not entirely sure this is worthwhile but may as well

9 years agoIntl: fix potential stack buffer overflow.
Tom Hacohen [Fri, 9 Oct 2015 08:26:45 +0000 (09:26 +0100)]
Intl: fix potential stack buffer overflow.

CID 1298051

@fix

9 years agoClient: fix potential out of bounds read.
Tom Hacohen [Fri, 9 Oct 2015 08:19:51 +0000 (09:19 +0100)]
Client: fix potential out of bounds read.

This is an unsigned int underflow that could lead to an out
of bounds read.

CID 1291837

@fix

9 years agoenlightenment: Update wayland readme
Chris Michael [Thu, 8 Oct 2015 19:28:55 +0000 (15:28 -0400)]
enlightenment: Update wayland readme

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agodon't crash desktop gadget systrays when adjusting image sizing
Mike Blumenkrantz [Thu, 8 Oct 2015 16:16:37 +0000 (12:16 -0400)]
don't crash desktop gadget systrays when adjusting image sizing

desktop gadgets (gadman) do not have a shelf, so derefing that pointer
without checking will guarantee a crash

9 years agoe comp data - actually ... NO visibility effect by default is right
Carsten Haitzler (Rasterman) [Wed, 7 Oct 2015 23:27:38 +0000 (08:27 +0900)]
e comp data - actually ... NO visibility effect by default is right

so a bit of a config wipe had my comp config reset (some dev
shenanigans) and i got to see what current e "out of box" config is -
and it was horrid for menus. visibilitiy effect was broken. vertial.
default was none - always. forever. like empty. so go back to that. in
fact why do this visibility effect? that was the point of having a
different shadow in the theme - each has its own look + effect. this
makes things far more complex...

9 years agowhen removing VERTICAL maximize state, also remove LEFT and RIGHT states
Mike Blumenkrantz [Wed, 7 Oct 2015 19:43:29 +0000 (15:43 -0400)]
when removing VERTICAL maximize state, also remove LEFT and RIGHT states

LEFT and RIGHT are supersets of VERTICAL which serve no purpose on their own,
so remove them when removing VERTICAL

fix T2770

9 years agoe comp - default config was rather broken - fix
Carsten Haitzler (Rasterman) [Wed, 7 Oct 2015 10:54:21 +0000 (19:54 +0900)]
e comp - default config was rather broken - fix

default config didnt select menu style shadow - was selecting
vertical/still which meant animation was rather horrible. fix.

9 years agoonly use fast-path pixmap size updating after a resize for x11 clients
Mike Blumenkrantz [Mon, 5 Oct 2015 21:28:26 +0000 (17:28 -0400)]
only use fast-path pixmap size updating after a resize for x11 clients

updating these values after other operations will result in the pixmap size
being set inaccurately, leading to rendering inconsistencies

9 years agoupdate wayland readme
Mike Blumenkrantz [Mon, 5 Oct 2015 20:36:39 +0000 (16:36 -0400)]
update wayland readme

9 years agorepopulate app list dialogs when efreet deskop cache updates
Mike Blumenkrantz [Mon, 5 Oct 2015 20:01:10 +0000 (16:01 -0400)]
repopulate app list dialogs when efreet deskop cache updates

should improve usage when installing/uninstalling apps

9 years agoattempt to load default e_bindings if initial load fails
Mike Blumenkrantz [Mon, 5 Oct 2015 19:56:28 +0000 (15:56 -0400)]
attempt to load default e_bindings if initial load fails

9 years agodo not set frame geometry on frame-possessing wayland clients
Mike Blumenkrantz [Mon, 5 Oct 2015 19:56:04 +0000 (15:56 -0400)]
do not set frame geometry on frame-possessing wayland clients

9 years agothrow an ACK when trying to set comp frame geometry if frame exists
Mike Blumenkrantz [Mon, 5 Oct 2015 19:55:44 +0000 (15:55 -0400)]
throw an ACK when trying to set comp frame geometry if frame exists

9 years agoclear menu cache when saving a menu file
Mike Blumenkrantz [Mon, 5 Oct 2015 17:37:45 +0000 (13:37 -0400)]
clear menu cache when saving a menu file

force regeneration of menus after editing them

ref T2271

9 years agomake icon selection dialog resizable in .desktop editor
Mike Blumenkrantz [Mon, 5 Oct 2015 17:33:13 +0000 (13:33 -0400)]
make icon selection dialog resizable in .desktop editor

ref T2271

9 years agomove configure init after intl init has completed
Mike Blumenkrantz [Mon, 5 Oct 2015 17:09:38 +0000 (13:09 -0400)]
move configure init after intl init has completed

gettext calls will not return the correct language string until after
setup has completed, leading to untranslatable strings

fix T2760

9 years agoclear the tiler for input region after setting to comp object.
Seunghun Lee [Mon, 5 Oct 2015 16:47:29 +0000 (12:47 -0400)]
clear the tiler for input region after setting to comp object.

Summary:
there is no need to set repeatedly input region even if it's already applied.
and this patch remove the code to clear tiler from client's unmapped case.
this fixes that tiler for input region is removed before applying it to comp object in case client is unmmaped yet.

Reviewers: devilhorns, zmike

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D3076

9 years agoAdd support for the ISO 8601 date format to the clock module
Peter Kjellerstedt [Sun, 4 Oct 2015 16:56:46 +0000 (12:56 -0400)]
Add support for the ISO 8601 date format to the clock module

Reviewers: zmike

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D3129

9 years ago e fm: Use correct string length when copying path.
Chidambar Zinnoury [Sun, 4 Oct 2015 16:44:38 +0000 (18:44 +0200)]
 e fm: Use correct string length when copying path.

 Sometimes, trying to go to the parent directory wouldn’t work and end up showing some garbage. Unfortunately, the code ended up not NULL-ending the path, which made searching for the path separator buggy.

9 years agoAdd "(drag only)" to title for drag only edge bindings
Peter Kjellerstedt [Sun, 4 Oct 2015 15:53:31 +0000 (11:53 -0400)]
Add "(drag only)" to title for drag only edge bindings

Reviewers: zmike

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D3128

9 years ago e menus: Fix invalid comparisons.
Chidambar Zinnoury [Sun, 4 Oct 2015 14:35:43 +0000 (16:35 +0200)]
 e menus: Fix invalid comparisons.

 Fixes CID 1267215 and 1267216.

9 years agofix systray to work with spec-breaking apps, eg. steam
Mike Blumenkrantz [Sun, 4 Oct 2015 12:41:07 +0000 (08:41 -0400)]
fix systray to work with spec-breaking apps, eg. steam

according to the StatusNotifierItem specification, applications
register "service org.freedesktop.StatusNotifierItem-PID-ID" on the
session bus, and then "must register the unique instance name
to the StatusNotifierWatcher".

some applications, such as steam, instead register the path that they
will run on (/org/ayatana/NotificationItem/steam) and then expect the
watcher to register the method call's send id bus: this is totally bogus.

to catch this, when registering the new item the enlightenment watcher must
first determine if the item is spec-conforming. if yes, proceed as normal.
if no, pretend the application knows what it's doing and try to make things
work as expected anyway

for more details, read the full spec here
http://www.freedesktop.org/wiki/Specifications/StatusNotifierItem

fix T2763

9 years agomove client post-update adding into render function
Mike Blumenkrantz [Fri, 2 Oct 2015 17:49:57 +0000 (13:49 -0400)]
move client post-update adding into render function

in some cases, it's possible for a client which expects to render on
the next frame to actually render on the frame after. in these cases,
the compositor must not clear the pixmap image until after the render
has occurred in order to avoid inaccuracies. for this reason, the best
place to flag a client for post-render work is at the time of the client's
render

ref T2762
ref D3120

9 years agounify comp render returns for argb and non-argb clients
Mike Blumenkrantz [Fri, 2 Oct 2015 17:44:37 +0000 (13:44 -0400)]
unify comp render returns for argb and non-argb clients

9 years agofix comp render return value for failed draws of non-argb clients
Mike Blumenkrantz [Fri, 2 Oct 2015 17:42:33 +0000 (13:42 -0400)]
fix comp render return value for failed draws of non-argb clients

9 years agofix return value of comp render for wayland clients
Mike Blumenkrantz [Fri, 2 Oct 2015 17:41:17 +0000 (13:41 -0400)]
fix return value of comp render for wayland clients

9 years agoadd RENDER_DEBUG for forced renders when setting dirty flag
Mike Blumenkrantz [Fri, 2 Oct 2015 17:40:55 +0000 (13:40 -0400)]
add RENDER_DEBUG for forced renders when setting dirty flag

9 years agoshould remove a frame resource from pending.frames list also
Boram Park [Fri, 2 Oct 2015 16:38:30 +0000 (12:38 -0400)]
should remove a frame resource from pending.frames list also

Summary: Signed-off-by: Boram Park <boram1288.park@samsung.com>

Reviewers: stefan_schmidt, gwanglim, raster, zmike, devilhorns

Reviewed By: zmike, devilhorns

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D3079

9 years agomd5: fix wrong sizeof argument (SIZEOF_MISMATCH)
Amitesh Singh [Thu, 1 Oct 2015 20:20:23 +0000 (16:20 -0400)]
md5: fix wrong sizeof argument (SIZEOF_MISMATCH)

Summary: fixes CID: 222382

Reviewers: raster, zmike, cedric, devilhorns

Reviewed By: devilhorns

Subscribers: cedric, seoz, sachin.dev

Differential Revision: https://phab.enlightenment.org/D3125

9 years agounset release mode
Mike Blumenkrantz [Thu, 1 Oct 2015 17:52:35 +0000 (13:52 -0400)]
unset release mode