gtk4.git
7 years agogl renderer: Don't add clip ops if the clip didn't change
Timm Bäder [Thu, 17 Jan 2019 07:22:00 +0000 (08:22 +0100)]
gl renderer: Don't add clip ops if the clip didn't change

7 years agoMerge branch 'get-timings-transfer' into 'master'
Emmanuele Bassi [Wed, 16 Jan 2019 23:43:14 +0000 (23:43 +0000)]
Merge branch 'get-timings-transfer' into 'master'

g-i: Add transfer none annotations for frame clock timings getters

See merge request GNOME/gtk!511

7 years agog-i: Add transfer none annotations for frame clock timings getters
Tomasz Miąsko [Wed, 16 Jan 2019 00:00:00 +0000 (00:00 +0000)]
g-i: Add transfer none annotations for frame clock timings getters

7 years agocolorscale: Redraw correct widget after setting rgba
Timm Bäder [Wed, 16 Jan 2019 18:20:53 +0000 (19:20 +0100)]
colorscale: Redraw correct widget after setting rgba

The entire color scale hack is still done in GtkRange, which draws the
color scale in the range gizmo. So, to correctly redraw the color scale
when setting a new color, we need to redraw the proper widget and that's
the trough widget.

Fixes #1453

7 years agoentry: Remove _gtk_entry_grab_focus
Timm Bäder [Wed, 16 Jan 2019 18:16:47 +0000 (19:16 +0100)]
entry: Remove _gtk_entry_grab_focus

We have API for both the select and !select behavior these days, so use
that one.

Fixes #1457

7 years agowindow: Directly find the cursor in the widget hierarchy
Timm Bäder [Wed, 16 Jan 2019 16:37:42 +0000 (17:37 +0100)]
window: Directly find the cursor in the widget hierarchy

Instead of recording the way up from the target widget to the grab
widget (or toplevel) and then walking that path upwards, just walk the
parent chain and look at the cursor.

7 years agoscrolledwindow: Remove unused drag_device member
Timm Bäder [Wed, 16 Jan 2019 09:55:15 +0000 (10:55 +0100)]
scrolledwindow: Remove unused drag_device member

7 years agosnapshot: Add gtk_snapshot_from_parent
Timm Bäder [Wed, 16 Jan 2019 07:30:42 +0000 (08:30 +0100)]
snapshot: Add gtk_snapshot_from_parent

Most of the time, the GtkSnapshot objects we create while snapshotting
widgets don't end up containing all that many nodes or states in their
respective node or state stack. This undermines the amortized allocation
behavior of the G(Ptr)Array we use for the stacks. So instead, use the
(until now unused) parent_snapshot GtkSnapshot* passed to
gtk_widget_create_render_node and reuse its node and state stack.

We do not avoid allocating a new GtkSnapshot object, but we do avoid
allocating a ton of G(Ptr)Array objects and we also avoid realloc'ing
their storage.

7 years agowindow: Don't manually queue a draw on the new focus widget
Timm Bäder [Tue, 15 Jan 2019 05:11:03 +0000 (06:11 +0100)]
window: Don't manually queue a draw on the new focus widget

This is either unneeded and OK to remove, or it is needed and there's a
bug somewhere else.

7 years agotreeview: Don't get separator_height more than once
Timm Bäder [Tue, 15 Jan 2019 15:26:13 +0000 (16:26 +0100)]
treeview: Don't get separator_height more than once

Closes !209

7 years agotreeview: cache expander size
Timm Bäder [Tue, 15 Jan 2019 15:23:18 +0000 (16:23 +0100)]
treeview: cache expander size

We need to know it all the time and since we get it from CSS, doing that
is fairly expensive.

7 years agoUpdated Spanish translation
Daniel Mustieles [Wed, 16 Jan 2019 08:54:54 +0000 (09:54 +0100)]
Updated Spanish translation

7 years agoUpdate Esperanto translation
Carmen Bianca BAKKER [Mon, 14 Jan 2019 21:16:22 +0000 (21:16 +0000)]
Update Esperanto translation

7 years agorbtree: Fix removing last element
Benjamin Otte [Mon, 14 Jan 2019 17:42:18 +0000 (18:42 +0100)]
rbtree: Fix removing last element

The refactoring of automatically updating tree->root when setting a
node's parent works very well - unless all nodes get removed and no
node's parent got updated.

7 years agorbtree: Make node getters not require tree argument
Benjamin Otte [Mon, 14 Jan 2019 01:12:02 +0000 (02:12 +0100)]
rbtree: Make node getters not require tree argument

The tree is not needed to walk around the nodes.

It is however still needed for anything that requires modifying the
tree.

There is no immediate benefit in changing this API, but there might be
situations in the future where we can avoid looking up the tree when we
just want to check some details about the node.

7 years agorbtree: Add gtk_rb_tree_node_get_tree()
Benjamin Otte [Mon, 14 Jan 2019 00:55:23 +0000 (01:55 +0100)]
rbtree: Add gtk_rb_tree_node_get_tree()

Store a link to the tree in the root node. This allows looking up the
tree in O(log N) from the node without any extra memory usage.

This is useful because code can just store a pointer to the node and
doesn't need to keep the tree pointer around. And that can (for large
trees) save quite a bit of memory.

7 years agorbtree: Access node->parent only via accessors
Benjamin Otte [Mon, 14 Jan 2019 00:44:07 +0000 (01:44 +0100)]
rbtree: Access node->parent only via accessors

This also adds a set_parent() function that automatically takes care of
updating tree->root for root nodes.

7 years agorbtree: Remove gtk_rb_tree_find()
Benjamin Otte [Sat, 12 Jan 2019 14:09:37 +0000 (15:09 +0100)]
rbtree: Remove gtk_rb_tree_find()

Searching through the tree is too specific to use a general function.
All the existing code just copies and slightly adapts the same 20 lines
instead, so there's no reason to keep the complicated API.

7 years agoRename GtkCssRBTree to GtkRBTree
Matthias Clasen [Mon, 7 Jan 2019 04:37:32 +0000 (23:37 -0500)]
Rename GtkCssRBTree to GtkRBTree

This is a generally usable red-black tree api,
and it is not used in css at all.

7 years agoRename GtkRBTree to GtkTreeRBTree
Matthias Clasen [Mon, 7 Jan 2019 03:28:09 +0000 (22:28 -0500)]
Rename GtkRBTree to GtkTreeRBTree

This frees up the generic name for a more
generic rbtree implementation.

7 years agogl renderer: Force blur shadow node children offscreen
Timm Bäder [Sun, 13 Jan 2019 07:21:58 +0000 (08:21 +0100)]
gl renderer: Force blur shadow node children offscreen

This broke the overlay blur demoe when resizing the window to a size
that would completely move the image below a button, causing the
GtkSnapshot code to remove the clip node below the blur node.

7 years agooverlay: Directly create cairo_region
Timm Bäder [Sun, 13 Jan 2019 06:47:53 +0000 (07:47 +0100)]
overlay: Directly create cairo_region

Instead of doing a cairo_region_create + union, just pass the rect
directly to cairo_region_create_rectangle

7 years agorendernodepaintable: Don't try to snapshot 0×0 nodes
Timm Bäder [Sun, 13 Jan 2019 06:23:10 +0000 (07:23 +0100)]
rendernodepaintable: Don't try to snapshot 0×0 nodes

Considering the operations that some of the rendernode constructors
do, nodes with width or height 0 (or both of course) are very well
possible. This would break in the rendernodepaintable when adding a
transform, which divides by width/height of the rendernode.

7 years agooverlay: Blur the clipped node, not clip the blurred node
Timm Bäder [Sun, 13 Jan 2019 05:35:44 +0000 (06:35 +0100)]
overlay: Blur the clipped node, not clip the blurred node

Blurring is a lot more expensive than clipping, so try to blur less by
clipping first.

7 years agooverlay: Trivial indentation fix
Timm Bäder [Sun, 13 Jan 2019 04:49:26 +0000 (05:49 +0100)]
overlay: Trivial indentation fix

7 years agodemo: Fix tranparency overlay demo
Timm Bäder [Sun, 13 Jan 2019 04:46:06 +0000 (05:46 +0100)]
demo: Fix tranparency overlay demo

We'd like the rose picture to be bigger than 16×16. Also remove the
scrolledwindow since the GtkPicture now automatically scales down the
rose image. This also fixes the picture always being allocated at y=0.

7 years agooverlay: Don't offset the main window snapshot
Timm Bäder [Sun, 13 Jan 2019 04:41:43 +0000 (05:41 +0100)]
overlay: Don't offset the main window snapshot

We need the node without an offset applied.

Fixes the rose picture being offset in tests/testoverlay

7 years agotestoverlay: Re-enable blurring
Timm Bäder [Sun, 13 Jan 2019 04:28:44 +0000 (05:28 +0100)]
testoverlay: Re-enable blurring

Accidentally disabled in 8b14c8d0a820e6617bd72a39a53d2ae6fa7b242b

7 years agolabel: Add missing nullable annotation
Timm Bäder [Sat, 12 Jan 2019 07:03:47 +0000 (08:03 +0100)]
label: Add missing nullable annotation

7 years agoUpdate Polish translation
Piotr Drąg [Sat, 12 Jan 2019 12:19:35 +0000 (13:19 +0100)]
Update Polish translation

7 years agospinbutton: Use increment/decrement icon names
Timm Bäder [Fri, 11 Jan 2019 17:17:08 +0000 (18:17 +0100)]
spinbutton: Use increment/decrement icon names

And add list-{remove,add}-symbolic as fallback.

Fixes #1569

7 years agolabel: Skip updating link state if we have no layout
Timm Bäder [Fri, 11 Jan 2019 16:46:12 +0000 (17:46 +0100)]
label: Skip updating link state if we have no layout

This can happen whenever the ::activate-link handler sets different
markup on the label, causing all links to be recreated. In this case,
the GtkLabelLink* passed to emit_activate_link is garbage after the
g_signal_emit call and we shouldn't try to do anything with it.

Fixes #1498

7 years agosnapshot: Get new GArray item directly
Timm Bäder [Fri, 11 Jan 2019 09:05:44 +0000 (10:05 +0100)]
snapshot: Get new GArray item directly

Instead of building it up ourselves, then coping it into the GArray and
then getting the last element of that array.

7 years agoUse the IEC power symbols for GtkSwitch labels
Emmanuele Bassi [Fri, 11 Jan 2019 15:21:09 +0000 (15:21 +0000)]
Use the IEC power symbols for GtkSwitch labels

Unicode 9.0 introduced glyps for the "on" and "off" power states, in the
form of:

 - U+23FD POWER ON SYMBOL, or ⏽
 - U+2B58 HEAVY CIRCLE, or ⭘

With `HEAVY CIRCLE` as "power off symbol" selected to avoid adding yet
another circle to the standard.

Since we moved GtkSwitch to always show glyphs instead of (translatable)
strings, asking the localisation teams to either come up with a suitable
short string to replace the English "ON" and "OFF", or to fall back to
Unicode glyphs, we should ensure we're using the appropriate symbols to
begin with.

See also: gtk!503 for the corresponding gtk-3-24 change.

7 years agoMerge branch 'wip/muktupavels/listbox-master' into 'master'
Timm Bäder [Fri, 11 Jan 2019 11:38:27 +0000 (11:38 +0000)]
Merge branch 'wip/muktupavels/listbox-master' into 'master'

listbox: set selected_row before emitting signal

See merge request GNOME/gtk!502

7 years agolistbox: set selected_row before emitting signal
Alberts Muktupāvels [Tue, 1 Jan 2019 23:14:17 +0000 (01:14 +0200)]
listbox: set selected_row before emitting signal

Signal emittion was added in 6f857f87dcd9 commit and it seems that
this is only place where selected_row is set after emitting signal.

Because of this gtk_list_box_get_selected_row currently returns NULL
as selected row if selection mode is set to GTK_SELECTION_BROWSE.

7 years agoMerge branch 'wip/dboles/EventControllerKey-docs-4' into 'master'
Timm Bäder [Fri, 11 Jan 2019 09:39:25 +0000 (09:39 +0000)]
Merge branch 'wip/dboles/EventControllerKey-docs-4' into 'master'

EventControllerKey: Add missing documentation bits

See merge request GNOME/gtk!454

7 years agoswitch: Always use I/O
Matthias Clasen [Thu, 10 Jan 2019 22:20:45 +0000 (17:20 -0500)]
switch: Always use I/O

This allows for smaller switches, and goes better with
modern themes that have round switches.

7 years agogl renderer: Dont't try to draw glyphs at scale 0
Timm Bäder [Thu, 10 Jan 2019 16:14:48 +0000 (17:14 +0100)]
gl renderer: Dont't try to draw glyphs at scale 0

Fixes #1475

7 years agogl glyphcache: Only support one dirty glyph per atlas
Timm Bäder [Sun, 6 Jan 2019 14:35:54 +0000 (15:35 +0100)]
gl glyphcache: Only support one dirty glyph per atlas

7 years agotestoverlay: Fix fancy effect
Timm Bäder [Sun, 6 Jan 2019 14:05:25 +0000 (15:05 +0100)]
testoverlay: Fix fancy effect

7 years agoiconcache: Remove unused defines
Timm Bäder [Sun, 6 Jan 2019 10:08:09 +0000 (11:08 +0100)]
iconcache: Remove unused defines

7 years agogl renderer: Remove debug code
Timm Bäder [Sun, 6 Jan 2019 09:59:21 +0000 (10:59 +0100)]
gl renderer: Remove debug code

7 years agogl renderer: Fix hidpi transform nodes
Timm Bäder [Sun, 6 Jan 2019 09:59:01 +0000 (10:59 +0100)]
gl renderer: Fix hidpi transform nodes

7 years agoshadowvalue: Remove unused API
Timm Bäder [Sat, 5 Jan 2019 08:09:58 +0000 (09:09 +0100)]
shadowvalue: Remove unused API

7 years agogl renderer: Only compile the vertex shader once
Timm Bäder [Sun, 6 Jan 2019 05:55:31 +0000 (06:55 +0100)]
gl renderer: Only compile the vertex shader once

All our programs use the same vertex shader, so don't compile it over
and over again. This improves startup times by at least 0.001%, I swear.

7 years agogl renderer: Don't forget the offset when rendering blur nodes
Timm Bäder [Sat, 5 Jan 2019 08:18:27 +0000 (09:18 +0100)]
gl renderer: Don't forget the offset when rendering blur nodes

7 years agoUpdate Spanish translation
Daniel Mustieles [Thu, 10 Jan 2019 08:50:55 +0000 (08:50 +0000)]
Update Spanish translation

7 years agoMerge branch 'patch-1' into 'master'
Timm Bäder [Thu, 10 Jan 2019 06:23:09 +0000 (06:23 +0000)]
Merge branch 'patch-1' into 'master'

Fix a typo in “popup-context-menu” documentation.

See merge request GNOME/gtk!500

7 years agoFix a typo in “popup-context-menu” documentation.
bodqhrohro [Wed, 9 Jan 2019 14:24:29 +0000 (14:24 +0000)]
Fix a typo in “popup-context-menu” documentation.

7 years agoMerge branch 'wip/sadiq/750-gtk4' into 'master'
Timm Bäder [Wed, 9 Jan 2019 11:53:45 +0000 (11:53 +0000)]
Merge branch 'wip/sadiq/750-gtk4' into 'master'

gtkpopover: Fix popover when parent widget is partially visible

Closes #750

See merge request GNOME/gtk!496

7 years agogtkpopover: Fix popover when parent widget is partially visible
Mohammed Sadiq [Wed, 9 Jan 2019 11:06:06 +0000 (16:36 +0530)]
gtkpopover: Fix popover when parent widget is partially visible

fixes https://gitlab.gnome.org/GNOME/gtk/issues/750

7 years agoUpdated Spanish translation
Daniel Mustieles [Wed, 9 Jan 2019 08:52:53 +0000 (09:52 +0100)]
Updated Spanish translation

7 years agoMerge branch 'meson-warning' into 'master'
Emmanuele Bassi [Tue, 8 Jan 2019 14:59:21 +0000 (14:59 +0000)]
Merge branch 'meson-warning' into 'master'

build: Avoid a Meson deprecation warning

See merge request GNOME/gtk!493

7 years agobuild: Avoid a Meson deprecation warning
Emmanuele Bassi [Tue, 8 Jan 2019 14:46:46 +0000 (14:46 +0000)]
build: Avoid a Meson deprecation warning

Using an empty `configuration_data` object to copy a configuration file
is deprecated since Meson 0.47 (released July 2018); the equivalent
behaviour is available by using `copy: true`.

7 years agorevealer: Handle child-visible properly
Benjamin Otte [Tue, 8 Jan 2019 00:29:41 +0000 (01:29 +0100)]
revealer: Handle child-visible properly

The target position is irrelevant for determining if the child should be
visible. When the current position is 0, it needs to be hidden, period.

Fixes #1355

7 years agosettings: Don't special case IM modules
Benjamin Otte [Mon, 7 Jan 2019 23:39:15 +0000 (00:39 +0100)]
settings: Don't special case IM modules

The previous fixes made it unnecessary to hardcode IM modules for
different display types. The code now automatically skips system IM
modules for other displays.

7 years agoimmodule: Check if module matches display
Benjamin Otte [Mon, 7 Jan 2019 23:36:58 +0000 (00:36 +0100)]
immodule: Check if module matches display

When loading IM modules from environment variables or GtkSettings, we
would accept "xim" on Wayland. That is a bad idea.

7 years agoimmodule: Don't load unsupported IM module
Benjamin Otte [Mon, 7 Jan 2019 23:25:08 +0000 (00:25 +0100)]
immodule: Don't load unsupported IM module

The code would technically allow loading the xim module when X11 support
was not compiled in.

This is probably an artificial concern, because it's pretty hard to
compile XIM support without X11 support, but it makes the code clearer,
so there we go.

7 years agoimmodule: Pass a GdkDisplay for default context ID
Benjamin Otte [Mon, 7 Jan 2019 23:22:14 +0000 (00:22 +0100)]
immodule: Pass a GdkDisplay for default context ID

After all, the context depends on the display - we want to use the
wayland context for Wayland,  xim for X11 and so on.

7 years agoimmulticontext: Don't have a global_context_id
Benjamin Otte [Mon, 7 Jan 2019 23:16:52 +0000 (00:16 +0100)]
immulticontext: Don't have a global_context_id

Context IDs are dependant on the display - both because displays can use
different backends, but also because changing the GtkSetting is a
per-display operation.

So just remove the cache.
If it turns out we need a per-display cache, we can add one to
GtkSettings.

7 years agoa11y: Check display in `*grab_cell_focus()`
Olivier Fourdan [Mon, 7 Jan 2019 21:06:57 +0000 (21:06 +0000)]
a11y: Check display in `*grab_cell_focus()`

Calling the accessibility function `grab_focus()` on a `GtkCell` under
Wayland will cause the client to crash.

This is another case of `gdk_x11_get_server_time()` being called
regardless of the actual windowing backend used.

Closes: https://gitlab.gnome.org/GNOME/gtk/issues/1507
7 years agoflowbox: Accept multipress gesture late
Carlos Garnacho [Wed, 2 Jan 2019 18:10:49 +0000 (19:10 +0100)]
flowbox: Accept multipress gesture late

So it's able to operate properly with the DnD gesture set by
gtk_drag_source_set(). We usually just react on button release,
that's the right time to claim the gesture.

Closes: https://gitlab.gnome.org/GNOME/gtk/issues/1557
7 years agosearchenginequartz: Fix naming of Beagle @ comment
Daniel Boles [Mon, 7 Jan 2019 20:34:32 +0000 (20:34 +0000)]
searchenginequartz: Fix naming of Beagle @ comment

As funny as that is, it's not very instructive. :P

https://gitlab.gnome.org/GNOME/gtk/merge_requests/490#note_400993

7 years ago[GtkSearchEngineQuartz] Limit the returned hits.
John Ralls [Mon, 7 Jan 2019 00:19:49 +0000 (16:19 -0800)]
[GtkSearchEngineQuartz] Limit the returned hits.

When the result set is larger than the limit.

7 years ago[GtkSearchEngineQuartz] Resolve the path to a GtkFile* for the hit list.
Daniel Boles [Sun, 6 Jan 2019 23:56:57 +0000 (15:56 -0800)]
[GtkSearchEngineQuartz] Resolve the path to a GtkFile* for the hit list.

Instead of a char* path, which the search_hit parser doesn't expect,
causing a crash.

Closes: $https://gitlab.gnome.org/GNOME/gtk/issues/815
7 years agoMerge branch 'fix-corner-resize' into 'master'
Carlos Garnacho [Mon, 7 Jan 2019 18:14:33 +0000 (18:14 +0000)]
Merge branch 'fix-corner-resize' into 'master'

Fix resize corners

See merge request GNOME/gtk!481

7 years agoMerge branch 'commit-docs' into 'master'
Matthias Clasen [Mon, 7 Jan 2019 16:44:41 +0000 (16:44 +0000)]
Merge branch 'commit-docs' into 'master'

docs: Move commit style docs to the contribution guide

See merge request GNOME/gtk!358

7 years agoUpdate Japanese translation
Ryuta Fujii [Sat, 5 Jan 2019 01:16:56 +0000 (01:16 +0000)]
Update Japanese translation

7 years agoUpdate Japanese translation
Ryuta Fujii [Fri, 4 Jan 2019 22:21:18 +0000 (22:21 +0000)]
Update Japanese translation

(cherry picked from commit 69ddfcf1ed165eb9b092f4da8a4b7396290487b0)

7 years agoUpdated Lithuanian translation
Aurimas Černius [Fri, 4 Jan 2019 21:23:00 +0000 (23:23 +0200)]
Updated Lithuanian translation

7 years agoimcontextwayland: (Hopefully) fix if expression
Timm Bäder [Fri, 4 Jan 2019 04:55:26 +0000 (05:55 +0100)]
imcontextwayland: (Hopefully) fix if expression

7 years agoUpdate POTFILES.in
Piotr Drąg [Thu, 3 Jan 2019 18:21:09 +0000 (19:21 +0100)]
Update POTFILES.in

7 years agobuttonbox: Remove redundant if statement
Timm Bäder [Wed, 26 Dec 2018 05:59:51 +0000 (06:59 +0100)]
buttonbox: Remove redundant if statement

This used to do something in the past.

7 years agocolorchooser: Remove extra custom color
Timm Bäder [Sat, 29 Dec 2018 06:38:55 +0000 (07:38 +0100)]
colorchooser: Remove extra custom color

7 years agopopover demo: Fix size-allocate callback signature
Timm Bäder [Sat, 29 Dec 2018 16:09:27 +0000 (17:09 +0100)]
popover demo: Fix size-allocate callback signature

7 years agowindow: Fix state_flags_changed impl name
Timm Bäder [Sun, 30 Dec 2018 05:08:55 +0000 (06:08 +0100)]
window: Fix state_flags_changed impl name

This is unrelated to GdkSurface and was changed by accident.

7 years agostack: Don't queue an unnecessary redraw
Timm Bäder [Sun, 23 Dec 2018 07:23:03 +0000 (08:23 +0100)]
stack: Don't queue an unnecessary redraw

Both queue_allocate and queue_resize already queue a draw.

7 years agoentry: Remove unused struct
Timm Bäder [Tue, 1 Jan 2019 11:21:41 +0000 (12:21 +0100)]
entry: Remove unused struct

7 years agoCheck for the session bus before using it
Matthias Clasen [Thu, 3 Jan 2019 01:35:19 +0000 (20:35 -0500)]
Check for the session bus before using it

Patch by Chris Allen, #475

7 years agoMerge branch 'wip/nbenitezl/move-bookmark-placeholder-index-4' into 'master'
Matthias Clasen [Thu, 3 Jan 2019 01:27:49 +0000 (01:27 +0000)]
Merge branch 'wip/nbenitezl/move-bookmark-placeholder-index-4' into 'master'

gtkplacessidebar.c: move bookmark at the placeholder index

See merge request GNOME/gtk!459

7 years agoMerge branch 'patch-1' into 'master'
Matthias Clasen [Thu, 3 Jan 2019 01:02:07 +0000 (01:02 +0000)]
Merge branch 'patch-1' into 'master'

Zero-fill new GtkTextIter

See merge request GNOME/gtk!455

7 years agoMerge branch 'wip/carlosg/imwayland-for-master' into 'master'
Matthias Clasen [Thu, 3 Jan 2019 00:36:35 +0000 (00:36 +0000)]
Merge branch 'wip/carlosg/imwayland-for-master' into 'master'

Use zwp_text_input_v3 on gtk4

Closes #1317 and #1525

See merge request GNOME/gtk!462

7 years agoFix XChangeProperty calls
Matthias Clasen [Thu, 3 Jan 2019 00:26:56 +0000 (19:26 -0500)]
Fix XChangeProperty calls

They must be long...

Closes: #1556
7 years agoFix resize corners
Matthias Clasen [Thu, 3 Jan 2019 00:08:02 +0000 (19:08 -0500)]
Fix resize corners

I was stuck in an X session and noticed that my resize corners
all got east or north cursors. It turns out that gnome-shell
does not properly advertise support for edge constraints under X11,
and the absence of that makes the code for determining the edge
under the cursor misbehave.

This change should fix that.

7 years agoMerge branch 'wip/carlosg/revealer-fix-interrupted-animations-master' into 'master'
Timm Bäder [Wed, 2 Jan 2019 08:50:25 +0000 (08:50 +0000)]
Merge branch 'wip/carlosg/revealer-fix-interrupted-animations-master' into 'master'

revealer: Fully set the target state if unmapped during animation

See merge request GNOME/gtk!466

7 years agoimage: Avoid some unnecessary signal connections
Timm Bäder [Sun, 30 Dec 2018 19:37:48 +0000 (20:37 +0100)]
image: Avoid some unnecessary signal connections

Equivalent if fd9879e5ffba2c58b897155b39ff79584b3618e0 but for GtkImage.

7 years agoimage: Indentation fix
Timm Bäder [Sun, 30 Dec 2018 19:25:21 +0000 (20:25 +0100)]
image: Indentation fix

7 years agogl renderer: Add a clip stack
Timm Bäder [Sun, 30 Dec 2018 14:25:01 +0000 (15:25 +0100)]
gl renderer: Add a clip stack

So we can check that the currently set clip is the first one and now
intersect with it. This first clip is always the entire viewport or the
entire render_area and we don't want to end up drawing things to a
texture because of it.

7 years agopicture: Avoid some unnecessary signal connections
Timm Bäder [Fri, 21 Dec 2018 08:26:38 +0000 (09:26 +0100)]
picture: Avoid some unnecessary signal connections

We don't need to ever invalidate the picture size if the paintable tells
us its size is static. Same for the contents.

7 years agoimage: Remove private struct from public header
Timm Bäder [Fri, 21 Dec 2018 08:14:20 +0000 (09:14 +0100)]
image: Remove private struct from public header

7 years agoCellAccessible: Fix return type of row|column_span
Daniel Boles [Sun, 30 Dec 2018 18:15:11 +0000 (18:15 +0000)]
CellAccessible: Fix return type of row|column_span

See:
 * c876c74eb73fe6daa279b42f9e0a9566ba1ced87
 * 63e0eb5de5d5a14afadebab31fe679beb6706cb2

7 years agoUpdate Catalan translation
Jordi Mas [Sat, 29 Dec 2018 08:31:36 +0000 (09:31 +0100)]
Update Catalan translation

7 years agodocs: Clarify the section on dependencies
Emmanuele Bassi [Fri, 28 Dec 2018 18:43:45 +0000 (18:43 +0000)]
docs: Clarify the section on dependencies

7 years agodocs: Clarify the section on commit access
Emmanuele Bassi [Mon, 15 Oct 2018 09:48:05 +0000 (10:48 +0100)]
docs: Clarify the section on commit access

The section is about rules for direct commit access to the upstream
repository, so let's be more clear about it.

7 years agodocs: Clarify autoclosing issues syntax
Emmanuele Bassi [Mon, 15 Oct 2018 09:46:35 +0000 (10:46 +0100)]
docs: Clarify autoclosing issues syntax

Link to the GitLab documentation, and clarify that if no single commit
in a merge requests closes an issue, you should add a reference to the
issue in the commit message anyway.

7 years agodocs: Make the contribution guide slightly more friendly
Emmanuele Bassi [Wed, 10 Oct 2018 15:42:15 +0000 (16:42 +0100)]
docs: Make the contribution guide slightly more friendly

This is an important document for newcomers, so we should err on the
side of being more detailed on what kind of contributions we expect,
and how we expect them.

The text is heavily modelled on the contributing-template by Nadia
Eghbal available here:

  https://github.com/nayafia/contributing-template

7 years agodocs: Move commit style docs to the contribution guide
Emmanuele Bassi [Wed, 26 Sep 2018 13:53:32 +0000 (14:53 +0100)]
docs: Move commit style docs to the contribution guide

There's no point in having a separate file detailing how commits ought
to work, considering we already have a contribution guide.

7 years agoUpdate Catalan translation
Jordi Mas [Fri, 28 Dec 2018 08:55:51 +0000 (09:55 +0100)]
Update Catalan translation

7 years agoUpdate Catalan translation
Jordi Mas [Tue, 25 Dec 2018 08:57:57 +0000 (09:57 +0100)]
Update Catalan translation