platform/upstream/enlightenment.git
9 years agomake comp cfd unresizable
Mike Blumenkrantz [Thu, 15 Jan 2015 00:55:21 +0000 (19:55 -0500)]
make comp cfd unresizable

the size of the tabs is smaller than the size of the content, which makes this look stupid when it covers the screen

9 years agodon't override entry min sizing in filepreview
Mike Blumenkrantz [Thu, 15 Jan 2015 00:41:21 +0000 (19:41 -0500)]
don't override entry min sizing in filepreview

9 years agoremove unnecessary size hinting
Mike Blumenkrantz [Thu, 15 Jan 2015 00:41:05 +0000 (19:41 -0500)]
remove unnecessary size hinting

9 years agoe_label -> elm_label conversion
Mike Blumenkrantz [Wed, 14 Jan 2015 19:43:29 +0000 (14:43 -0500)]
e_label -> elm_label conversion

text is text is text is text text is text is text is text

 #teamborker

9 years agoadd workaround for ELM_DIALOG_WIN_BASIC not working in wayland.
Chris Michael [Wed, 14 Jan 2015 21:33:19 +0000 (16:33 -0500)]
add workaround for ELM_DIALOG_WIN_BASIC not working in wayland.
add #warning to remove when resolved.

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agodon't crash if the client comp_data has no surface yet
Chris Michael [Wed, 14 Jan 2015 21:11:53 +0000 (16:11 -0500)]
don't crash if the client comp_data has no surface yet

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoupdate entry widget to not use deprecated elm scrollbar policy function
Chris Michael [Wed, 14 Jan 2015 20:07:51 +0000 (15:07 -0500)]
update entry widget to not use deprecated elm scrollbar policy function

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoforce internal window recentering after resize when no move/resize is active
Mike Blumenkrantz [Wed, 14 Jan 2015 19:47:54 +0000 (14:47 -0500)]
force internal window recentering after resize when no move/resize is active

9 years agoremove unused vars
Mike Blumenkrantz [Wed, 14 Jan 2015 03:46:14 +0000 (22:46 -0500)]
remove unused vars

9 years agode-e_comp_get()ify e_fwin
Mike Blumenkrantz [Wed, 14 Jan 2015 03:44:23 +0000 (22:44 -0500)]
de-e_comp_get()ify e_fwin

9 years agoautoupdate widget min sizes
Mike Blumenkrantz [Wed, 14 Jan 2015 02:15:07 +0000 (21:15 -0500)]
autoupdate widget min sizes

9 years agoredo fileman tooltips to size correctly
Mike Blumenkrantz [Wed, 14 Jan 2015 01:04:19 +0000 (20:04 -0500)]
redo fileman tooltips to size correctly

exhibits edje min size calc bug/workaround

9 years agoRevert "fix centering of elm wayland windows"
Mike Blumenkrantz [Tue, 13 Jan 2015 22:24:33 +0000 (17:24 -0500)]
Revert "fix centering of elm wayland windows"

This reverts commit 1ec2a4e91c1ab1402b2a7f0c758684bdc37f3ef5.

this may have fixed wayland window centering, but only because it effectively turns the function into a no-op instead of actually centering the window

9 years agoset evas size hint min in widget min size
Mike Blumenkrantz [Tue, 13 Jan 2015 00:45:41 +0000 (19:45 -0500)]
set evas size hint min in widget min size

9 years agoe_table -> elm_table conversion
Mike Blumenkrantz [Tue, 13 Jan 2015 00:24:52 +0000 (19:24 -0500)]
e_table -> elm_table conversion

if rebases were monopoly money, this commit would be able to buy mediterranean ave.

 #teamborker

9 years agoe - force a config flush before sys actions that could kill e
Carsten Haitzler (Rasterman) [Wed, 14 Jan 2015 02:48:05 +0000 (11:48 +0900)]
e - force a config flush before sys actions that could kill e

@fix

this should fix T1988. after some looking here is what i think is
happening. he is asking the system (sysvinit or systemd) to
halt/reboot, but these end up killing the e process before it can save
its config that is lurking around still in ram as opposed to on disk,
so puttting in a config flush before beginning such a system action
will ensure everything is on disk.

9 years agofix centering of elm wayland windows
Chris Michael [Mon, 12 Jan 2015 18:36:58 +0000 (13:36 -0500)]
fix centering of elm wayland windows

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agocomp-wl: Set argb and no_shape_cut for new clients. This fixes the
Chris Michael [Mon, 12 Jan 2015 18:03:52 +0000 (13:03 -0500)]
comp-wl: Set argb and no_shape_cut for new clients. This fixes the
black shadow issue

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoFix Coverity CID1261286. Identical code for different branches
Chris Michael [Mon, 12 Jan 2015 17:12:52 +0000 (12:12 -0500)]
Fix Coverity CID1261286. Identical code for different branches

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoFix Coverity CID1261288. Idential code for different branches
Chris Michael [Mon, 12 Jan 2015 17:11:26 +0000 (12:11 -0500)]
Fix Coverity CID1261288. Idential code for different branches

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agopager_plain: Fix Coverity CID1261290 reporting logically dead code,
Chris Michael [Mon, 12 Jan 2015 17:09:43 +0000 (12:09 -0500)]
pager_plain: Fix Coverity CID1261290 reporting logically dead code,
but in fact it looks like a copy/paste issue

@fix

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoE comp: Adjust according to elm function rename.
Tom Hacohen [Mon, 12 Jan 2015 10:18:11 +0000 (10:18 +0000)]
E comp: Adjust according to elm function rename.

9 years agode-e_comp_get()-ify e_shelf
Mike Blumenkrantz [Sun, 11 Jan 2015 20:57:07 +0000 (15:57 -0500)]
de-e_comp_get()-ify e_shelf

9 years agofix build from tarball when configured mount backend is different from origin
Mike Blumenkrantz [Sun, 11 Jan 2015 17:56:47 +0000 (12:56 -0500)]
fix build from tarball when configured mount backend is different from origin

previously the generated e_fm_shared_types.h file was wrongly distributed with the rest of the source tree when it should not have been

9 years agoe_entry -> elm_entry conversion
Mike Blumenkrantz [Sun, 11 Jan 2015 17:20:38 +0000 (12:20 -0500)]
e_entry -> elm_entry conversion

another dice roll in the board game of enlightenment development. next, I will attempt to roll doubles to get out of jail.

 #teamborker

fix T1906

9 years agosimplify comp usage in e_fm
Mike Blumenkrantz [Sun, 11 Jan 2015 17:20:35 +0000 (12:20 -0500)]
simplify comp usage in e_fm

9 years agouse elm_win_fake() to create a fake elm_win for widget parenting/focus
Mike Blumenkrantz [Sun, 11 Jan 2015 17:20:32 +0000 (12:20 -0500)]
use elm_win_fake() to create a fake elm_win for widget parenting/focus

ugh

9 years agoset comp X window hint to normal
Mike Blumenkrantz [Sun, 11 Jan 2015 17:20:21 +0000 (12:20 -0500)]
set comp X window hint to normal

ee is broken in that it will never update the withdrawn property for a canvas unless you 1) explicitly set this 2) have an override window

9 years agounwrap INLINE_IMAGE window creation
Mike Blumenkrantz [Tue, 6 Jan 2015 21:05:39 +0000 (16:05 -0500)]
unwrap INLINE_IMAGE window creation

9 years ago+e_win_evas_win_get()
Mike Blumenkrantz [Tue, 6 Jan 2015 21:00:04 +0000 (16:00 -0500)]
+e_win_evas_win_get()

9 years agocomp-wl: Remove unused variable
Chris Michael [Fri, 9 Jan 2015 18:44:46 +0000 (13:44 -0500)]
comp-wl: Remove unused variable

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agocomp-wl: Add missing function prototype
Chris Michael [Fri, 9 Jan 2015 18:43:40 +0000 (13:43 -0500)]
comp-wl: Add missing function prototype

Since we call this function before it's actually defined in the
source, we'll just add a small function prototype at the top

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agowl-desktop-shell: Fix issue where internal elm windows were not
Chris Michael [Fri, 9 Jan 2015 17:47:08 +0000 (12:47 -0500)]
wl-desktop-shell: Fix issue where internal elm windows were not
borderless

This fixes the issue of elementary windows having 2 borders, however
it does expose a new issue where those same internal elm windows now
have a black shadow border around them.

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agocomp_wl: Fix issue with multiple clients being created
Chris Michael [Fri, 9 Jan 2015 17:45:03 +0000 (12:45 -0500)]
comp_wl: Fix issue with multiple clients being created

This fixes an issue where internal elementary dialogs would end up
creating 2 pixmaps and 2 clients for each window. This happened due to
the pixmap id not matching the elm window id.

This patch also handles reparenting internal elm windows if needed.

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agofix crash when activating menu that is not in range
Duna Oh [Fri, 9 Jan 2015 05:51:40 +0000 (14:51 +0900)]
fix crash when activating menu that is not in range

Summary: add the condition check code if the pressed key is not in range of menu items

Test Plan:
(1) Left-click on the desktop to bring up the menu.
(2) Press down the number key to activate the menu item.
(3) When you press down the number key that is not in range, you will see the segmentation error.

below is the crash log which I encoutered.
-------------------------------------------------------------------
Program received signal SIGSEGV, Segmentation fault.
_e_menu_item_activate_nth (n=n@entry=5) at src/bin/e_menu.c:2318
2318       _e_menu_item_ensure_onscreen(mi);

Reviewers: zmike, gwanglim, raster

Reviewed By: raster

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D1845

9 years agoe zone - add stow/unstow apis
Carsten Haitzler (Rasterman) [Fri, 9 Jan 2015 02:50:35 +0000 (11:50 +0900)]
e zone - add stow/unstow apis

9 years agodon't change shade.x/y during unshade or after shade
Mike Blumenkrantz [Thu, 8 Jan 2015 22:51:47 +0000 (17:51 -0500)]
don't change shade.x/y during unshade or after shade

fix T1930

9 years agocheck focus after X nocomp end
Mike Blumenkrantz [Thu, 8 Jan 2015 22:24:50 +0000 (17:24 -0500)]
check focus after X nocomp end

fix T1768

9 years agofix issue where wayland modules were always creating a new compositor
Chris Michael [Tue, 6 Jan 2015 18:04:04 +0000 (13:04 -0500)]
fix issue where wayland modules were always creating a new compositor

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agowl-desktop-shell: Remove use of deprecated e_comp_get function
Chris Michael [Tue, 6 Jan 2015 14:30:30 +0000 (09:30 -0500)]
wl-desktop-shell: Remove use of deprecated e_comp_get function

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoremove use of deprecated e_comp_get function
Chris Michael [Tue, 6 Jan 2015 14:28:58 +0000 (09:28 -0500)]
remove use of deprecated e_comp_get function

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoadd E_VERSION_MAJOR define for external module usage
Mike Blumenkrantz [Mon, 5 Jan 2015 20:49:57 +0000 (15:49 -0500)]
add E_VERSION_MAJOR define for external module usage

this should simplify writing modules with conditional code for different versions of e

9 years agounused param--
Mike Blumenkrantz [Mon, 5 Jan 2015 20:38:51 +0000 (15:38 -0500)]
unused param--

9 years agoremove some e_comp_get() usage in core comp/client files
Mike Blumenkrantz [Mon, 5 Jan 2015 20:35:11 +0000 (15:35 -0500)]
remove some e_comp_get() usage in core comp/client files

9 years agodisable client stack debugging
Mike Blumenkrantz [Mon, 5 Jan 2015 20:29:41 +0000 (15:29 -0500)]
disable client stack debugging

this should have been turned off for e19 release but I forgot

9 years agoremove e_comp_list(), deprecate all related functions for pending removal
Mike Blumenkrantz [Mon, 5 Jan 2015 20:27:41 +0000 (15:27 -0500)]
remove e_comp_list(), deprecate all related functions for pending removal

there is only one E_Comp which can now be accessed by the e_comp global.

if you're editing a file with some uses of these deprecated functions, replace their usages with appropriate references to this variable

pass -Wno-deprecated-declarations to ignore these warnings during build

9 years agoenforce no multihead in X compositor
Mike Blumenkrantz [Mon, 5 Jan 2015 20:26:47 +0000 (15:26 -0500)]
enforce no multihead in X compositor

this hasn't worked since E17, may was well stop trying

9 years agowl_drm: Add event handler for session activation signal.
Seunghun Lee [Mon, 5 Jan 2015 14:09:59 +0000 (09:09 -0500)]
wl_drm: Add event handler for session activation signal.

Summary:
Compositor need to change the compositing mode to offscreen or render state depend on session activation state.
this patch is for it.

Test Plan:
(1) Run enlightenment on X.
(2) Change another VT.
(3) Run enlightenment on wayland and drm.
(4) Change the VT to enlightenment which is ran (1).
enlightenment on wayland will not render during session deactivation state.

Reviewers: zmike, devilhorns

Reviewed By: devilhorns

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D1816

9 years agoTiling: don't try to change the border of borderless windows.
Tom Hacohen [Sun, 4 Jan 2015 13:45:16 +0000 (13:45 +0000)]
Tiling: don't try to change the border of borderless windows.

@fix

9 years agoadd support for new netwm csd X atoms
Mike Blumenkrantz [Sat, 3 Jan 2015 05:42:42 +0000 (00:42 -0500)]
add support for new netwm csd X atoms

nothing uses these yet so there's no way to test them.

fix T1963

9 years agocompletely block border changes on MWM borderless windows
Mike Blumenkrantz [Sat, 3 Jan 2015 05:07:48 +0000 (00:07 -0500)]
completely block border changes on MWM borderless windows

ref T1963

seriously.

9 years agodon't set gravity on client windows during X move/resize
Mike Blumenkrantz [Sat, 3 Jan 2015 04:56:11 +0000 (23:56 -0500)]
don't set gravity on client windows during X move/resize

fix T1940

9 years agoremove e_box from POTFILES.in
Mike Blumenkrantz [Thu, 1 Jan 2015 22:57:28 +0000 (17:57 -0500)]
remove e_box from POTFILES.in

seriously why is a file with no translatable text in here

9 years agoupdate copyright year
Yomi [Thu, 1 Jan 2015 22:45:08 +0000 (17:45 -0500)]
update copyright year

Summary: update copyright year

Reviewers: zmike, bu5hm4n

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D1836

9 years agoadd compile warning note for elm hack
Mike Blumenkrantz [Thu, 1 Jan 2015 22:31:25 +0000 (17:31 -0500)]
add compile warning note for elm hack

9 years agoset elm log level to CRI
Mike Blumenkrantz [Thu, 1 Jan 2015 22:31:21 +0000 (17:31 -0500)]
set elm log level to CRI

current elm usage triggers infinite (harmless) widget-parenting errors. to avoid filling hard disks with log files, shut this up for now

9 years agodon't recalc winlist on client add event if client is not added to winlist
Mike Blumenkrantz [Thu, 1 Jan 2015 22:31:16 +0000 (17:31 -0500)]
don't recalc winlist on client add event if client is not added to winlist

9 years agoe_box -> elm_box conversion
Mike Blumenkrantz [Thu, 1 Jan 2015 22:29:25 +0000 (17:29 -0500)]
e_box -> elm_box conversion

the next installment in the exciting adventure game Elementary Wars!

happy new year from #teamborker

9 years agocomp pending resizes should not require both w+h to change
Mike Blumenkrantz [Tue, 30 Dec 2014 01:42:27 +0000 (20:42 -0500)]
comp pending resizes should not require both w+h to change

9 years agodon't show mixer popup during nocomp
Mike Blumenkrantz [Mon, 29 Dec 2014 22:31:08 +0000 (17:31 -0500)]
don't show mixer popup during nocomp

fix T1270

9 years agouse frame geometry in win resize trap to get accurate sizes
Mike Blumenkrantz [Mon, 29 Dec 2014 22:25:43 +0000 (17:25 -0500)]
use frame geometry in win resize trap to get accurate sizes

ref T1905

9 years agoadd back notification merging
Mike Blumenkrantz [Sun, 28 Dec 2014 22:25:59 +0000 (17:25 -0500)]
add back notification merging

this was removed without warning or explanation in r82118:

    e: Change notification module to use internal notification API
    Patch by: José Roberto de Souza  <zehortigoza@profusion.mobi>

currently broken by T1954 for lines longer than 36 characters

9 years agoconfig: Fix segfault
Marcel Hollerbach [Wed, 24 Dec 2014 12:16:01 +0000 (13:16 +0100)]
config: Fix segfault

We can pass here NULL, win does not have a specific saved e_comp.
So we pass NULL to e_comp_get and get the current active one.

This fixes T1937.

9 years agoRevert "all dialogs are now resizable"
Carsten Haitzler (Rasterman) [Mon, 22 Dec 2014 14:02:05 +0000 (23:02 +0900)]
Revert "all dialogs are now resizable"

This reverts commit 9157ca8ca739e3b62442fef9d73367d381ae7306.

this also makes all dialogs that have content in a toolbar that goes
off window at min size resizable - this is actually what you wanted ..
bgut this previous "all dialogs are resizable" thing was uber-lazy.

Conflicts:
src/modules/bluez4/e_mod_main.c
src/modules/conf_bindings/e_int_config_signalbindings.c
src/modules/conf_theme/e_int_config_theme.c
src/modules/mixer/conf_gadget.c
src/modules/shot/e_mod_main.c

9 years agocall e_powersave_init Before we call e_screensaver_init because the
Chris Michael [Thu, 18 Dec 2014 15:13:27 +0000 (10:13 -0500)]
call e_powersave_init Before we call e_screensaver_init because the
init of e_screensaver sets up an event handler for the
powersave_update event (which would be 0 unless powersave is init
first).

Thanks to dt9 for the report ! ;)

Signed-off-by: Chris Michael <cp.michael@samsung.com>
9 years agoe_actions: Fix to parse syntax correctly in key binding settings
Jee-Yong Um [Thu, 18 Dec 2014 07:08:34 +0000 (16:08 +0900)]
e_actions: Fix to parse syntax correctly in key binding settings

Summary:
Correct the way to parse syntax in key bindings for launching application according to syntax guide
and pop an error dialog when a space is detected while activating the action.

Example:
Guide is given like "exe:xterm" with "No whitespace" between param's name and contents.
However, existing way to parse syntax should require "ONE whitespace" between name and contents.
This modification will parse syntax correctly.

@fix

Reviewers: seoz, zmike, Hermet

Subscribers: raster, cippp, cedric

Differential Revision: https://phab.enlightenment.org/D1699

9 years agoe - randr include - move to master includes (only if building for x)
Carsten Haitzler (Rasterman) [Thu, 18 Dec 2014 05:19:07 +0000 (14:19 +0900)]
e - randr include - move to master includes (only if building for x)

9 years agorandr: improve code readability in lid update
Sebastian Dransfeld [Wed, 17 Dec 2014 10:39:19 +0000 (11:39 +0100)]
randr: improve code readability in lid update

Be sure we have active non-lids before deactivating lids.

9 years agorandr: fix bad copy paste error
Sebastian Dransfeld [Wed, 17 Dec 2014 10:35:23 +0000 (11:35 +0100)]
randr: fix bad copy paste error

9 years agosystray: revert change which shouldn't go in
Sebastian Dransfeld [Wed, 17 Dec 2014 10:20:22 +0000 (11:20 +0100)]
systray: revert change which shouldn't go in

9 years agorandr: add error print
Sebastian Dransfeld [Wed, 17 Dec 2014 10:19:25 +0000 (11:19 +0100)]
randr: add error print

9 years agoRevert "randr: always call _e_randr_output_active_set"
Sebastian Dransfeld [Wed, 17 Dec 2014 10:16:43 +0000 (11:16 +0100)]
Revert "randr: always call _e_randr_output_active_set"

This reverts commit 7f14d81e6a4afe7217a7ed4d6dac06e4fbf3a983.

Broke LID open/close for me, note to self, more testing before commit.

9 years agorandr: TODO++ and error print
Sebastian Dransfeld [Wed, 17 Dec 2014 09:35:54 +0000 (10:35 +0100)]
randr: TODO++ and error print

9 years agorandr: don't try to enable crtc with no outputs
Sebastian Dransfeld [Wed, 17 Dec 2014 09:35:12 +0000 (10:35 +0100)]
randr: don't try to enable crtc with no outputs

9 years agorandr: always call _e_randr_output_active_set
Sebastian Dransfeld [Wed, 17 Dec 2014 09:29:27 +0000 (10:29 +0100)]
randr: always call _e_randr_output_active_set

_e_randr_output_active_set should handle if we set active on an active
output

9 years agorandr: handle (x, y) > 0
Sebastian Dransfeld [Wed, 17 Dec 2014 09:22:48 +0000 (10:22 +0100)]
randr: handle (x, y) > 0

x and y can be larger than 0, so we must set a high start value.

9 years agorandr: remove weird code
Sebastian Dransfeld [Wed, 17 Dec 2014 07:56:24 +0000 (08:56 +0100)]
randr: remove weird code

If we do E_Randr_Output *out2 = _e_randr_output_find(output->xid), then
output should be equal to out2, else we have several output variables per
output xid which should be fixed correctly.

9 years agorandr: cleanup try 2
Sebastian Dransfeld [Wed, 17 Dec 2014 07:49:40 +0000 (08:49 +0100)]
randr: cleanup try 2

This does not change any behaviour, just moves code:

1. We only try to locate a crtc for an output in _e_randr_output_active_set

And we only do this when we set active.

2. We only search for a crtc in _e_randr_output_crtc_find

This includes looping known crtcs without querying X if it matches.

Please don't revert, but help cleaning. As the current code does not
work for me...

9 years agoRevert "Randr: Fix null dereference."
Carsten Haitzler (Rasterman) [Wed, 17 Dec 2014 00:33:45 +0000 (09:33 +0900)]
Revert "Randr: Fix null dereference."
This reverts commit 4aef218d28ed66e80d882f052c704d92d11c3c81.

Revert "randr: remember crtcid for fast lookup on reconnect"
This reverts commit 55b4ad41d19f1e4573203bf4296356d8f991e809.

Revert "randr: further cleanup"
This reverts commit 3e02824663a2d6b54a2a9ba6d5d9ab760c42aff2.

these make nvidia randr resolution changing work again. sorry seb.
even though cleanups might be nice... this code is FIDDLY and every
driver seems to work differently, so take it easy and be very careful
- make no assumptions. i know it works on my intel desktop and nvidia
deskop atm. once i get a vga cable i'll fix up intel laptop with
external display that has problems. i can try radeon too later. but
right now - this stuff needs to work. being clean is far less
important than working. :)

9 years agoRevert "randr: Don't track active outputs"
Carsten Haitzler (Rasterman) [Wed, 17 Dec 2014 00:12:32 +0000 (09:12 +0900)]
Revert "randr: Don't track active outputs"

This reverts commit 01f43e3618d07917a15345e2063b9efd866e4b7d.

this broke nvidia randr - i cant tunr my screens on or off anymore.

9 years agopo: Remove file from now disabled access module to pass distcheck
Stefan Schmidt [Tue, 16 Dec 2014 15:47:38 +0000 (16:47 +0100)]
po: Remove file from now disabled access module to pass distcheck

9 years agoRandr: Fix null dereference.
Tom Hacohen [Tue, 16 Dec 2014 14:56:10 +0000 (14:56 +0000)]
Randr: Fix null dereference.

CID1181926

9 years agorandr: Don't track active outputs
Sebastian Dransfeld [Tue, 16 Dec 2014 09:13:55 +0000 (10:13 +0100)]
randr: Don't track active outputs

Remove tracking of active outputs to reduce code complexity.

9 years agorandr: TODO++
Sebastian Dransfeld [Tue, 16 Dec 2014 08:56:02 +0000 (09:56 +0100)]
randr: TODO++

9 years agorandr: remember crtcid for fast lookup on reconnect
Sebastian Dransfeld [Tue, 16 Dec 2014 08:21:02 +0000 (09:21 +0100)]
randr: remember crtcid for fast lookup on reconnect

When an output is set inactive, it is disconnected from a crtc. Now we
remember the id of the last crtc it was connected to, so we get a faster
lookup on reconnect.

9 years agoe randr - restore edp checks
Carsten Haitzler (Rasterman) [Tue, 16 Dec 2014 02:31:00 +0000 (11:31 +0900)]
e randr - restore edp checks

the dp checks were looking for edp stuff. but due to naming they appeard as
eDP or maybe edp or EDP - thus were simplified to dp. put more explicit
checks there.

9 years agorandr: Don't match DP as LID
Sebastian Dransfeld [Mon, 15 Dec 2014 11:51:16 +0000 (12:51 +0100)]
randr: Don't match DP as LID

On my intel laptop DP are my DisplayPort connections, not LID.

9 years agorandr: further cleanup
Sebastian Dransfeld [Mon, 15 Dec 2014 11:25:22 +0000 (12:25 +0100)]
randr: further cleanup

Primarly two things removed:
1. crtc->outputs should only contain outputs which should be enabled
So looping them twice in e_randr_apply is definitly wrong.

2. _e_randr_crtc_find should find a crtc for an output
So looping our crtcs again in _active_set is definitly wrong. If
_e_randr_crtc_find does not find us a crtc, it should be fixed there.

9 years agoe randr - make it work on nvidia - lots of "wrong" things fixed
Carsten Haitzler (Rasterman) [Sun, 14 Dec 2014 03:39:36 +0000 (12:39 +0900)]
e randr - make it work on nvidia - lots of "wrong" things fixed

i've spent a while testing randr on my nvidia box and drviers. it now
works - i can turn screens on and off. i couldn't before. e_randr was
trying to be too smart for its own good. it wasnt matching up outputs
to crtcs, was thinking outputs active when they were not, etc. etc.
etc. - reality is e_randr could do witha  lot of simplification, but
at this stage i'm just going to hammer it into shape to work across as
many drivers and setups as i can test against.

9 years agoonly use elm win title for internal windows if it has been set
Mike Blumenkrantz [Sun, 14 Dec 2014 20:40:25 +0000 (15:40 -0500)]
only use elm win title for internal windows if it has been set

9 years agoe_win: Fix Dialog Titles
Marcel Hollerbach [Sun, 14 Dec 2014 11:21:16 +0000 (12:21 +0100)]
e_win: Fix Dialog Titles

We should not set the title to "E" in the trap function of show.
If we do this we will override the icccm title and the netwm name, this
mean e_comp_x will set E as title of the window if the call to
elm_win_title_set was BEFORE the window was shown.

Now the name which was set before will be inserted.

9 years agoRegister the new window into out evas.
Marcel Hollerbach [Fri, 12 Dec 2014 08:39:51 +0000 (09:39 +0100)]
Register the new window into out evas.

Summary:
Otherwise the mouse up event is emitted to the evas but the evas does
not work with it, because it does not know the window it is from. This
leads to a evas_event_down_count_get bigger than 1 which leads to a
missing mouse_out event. With this missing mouse_out event _post_up_handle
in evas_event.c is not called anymore, this means all the click events
are getting emitted to the object which was last clicked.

This fixes T1894.

Test Plan: Drag and Drop a icon somewhere and click somewhere after that, the evas will not get stuck anymore

Reviewers: raster

Subscribers: cedric, abyomi0, zmike, raster

Maniphest Tasks: T1894

Differential Revision: https://phab.enlightenment.org/D1771

9 years agostore hide_logo state for desklock backgrounds
Mike Blumenkrantz [Thu, 11 Dec 2014 17:24:16 +0000 (12:24 -0500)]
store hide_logo state for desklock backgrounds

fix T1714

9 years agoapply client size based on correct remember value, only apply if value != 0
Mike Blumenkrantz [Mon, 8 Dec 2014 22:06:13 +0000 (17:06 -0500)]
apply client size based on correct remember value, only apply if value != 0

ref T999

9 years agoremembering size on a window should also remember maximize state
Mike Blumenkrantz [Mon, 8 Dec 2014 22:05:58 +0000 (17:05 -0500)]
remembering size on a window should also remember maximize state

9 years agogreatly reduce FAILURE_MAX
Mike Blumenkrantz [Mon, 8 Dec 2014 22:03:23 +0000 (17:03 -0500)]
greatly reduce FAILURE_MAX

ref T1820

9 years agoremove access module from build
Mike Blumenkrantz [Mon, 8 Dec 2014 20:38:02 +0000 (15:38 -0500)]
remove access module from build

this hasn't been updated in over 1.5 years and has caused me to waste time repeatedly fixing build-related issues. in addition, it uses symbols which are no longer even available in efl.

if nobody updates/maintains this, it will be removed for e20 release

src/modules/access/e_mod_main.c: In function '_cover_new':
src/modules/access/e_mod_main.c:765:4: warning: implicit declaration of function 'ecore_x_input_multi_select' [-Wimplicit-function-declaration]

9 years agoreduce shape recalc for clients which cannot currently receive input
Mike Blumenkrantz [Mon, 8 Dec 2014 01:16:42 +0000 (20:16 -0500)]
reduce shape recalc for clients which cannot currently receive input

9 years agoredo shaped client checks to use shape flags instead of rect pointer, fix shapeless...
Mike Blumenkrantz [Mon, 8 Dec 2014 01:14:34 +0000 (20:14 -0500)]
redo shaped client checks to use shape flags instead of rect pointer, fix shapeless clients

according to the shape extension protocol, the number of rectangles returned should be checked to determine a client's shape. if 0 is returned, the client has no shape, meaning that it either should not be drawn or should have no input region. this improves behavior with various client window types such as tooltips

ref T1820

9 years agoadd failure count for client rendering, drop clients after FAILURE_MAX tries
Mike Blumenkrantz [Mon, 8 Dec 2014 01:11:46 +0000 (20:11 -0500)]
add failure count for client rendering, drop clients after FAILURE_MAX tries

this traps clients which are showing and hiding themselves extremely quickly and prevents them from crashing us when we try to draw them

fix T1820