gtk issueshttps://gitlab.gnome.org/GNOME/gtk/-/issues2021-04-03T11:09:28Zhttps://gitlab.gnome.org/GNOME/gtk/-/issues/3398textview selection highlight may be shown when no selection2021-04-03T11:09:28ZMohammed Sadiqtextview selection highlight may be shown when no selectionSometime, with large text, the selection is still shown (and broken) when there should be no selection
How to reproduce:
1. Open `gtk4-demo --run=markup`
1. enable Source toggle
1. Select all text there and copy paste 15 times or so s...Sometime, with large text, the selection is still shown (and broken) when there should be no selection
How to reproduce:
1. Open `gtk4-demo --run=markup`
1. enable Source toggle
1. Select all text there and copy paste 15 times or so so that the buffer is large
1. Select pretty large portion of text using mouse click drag (up to the end of buffer)
1. Click somewhere in the selection
Result: The highlight of the line that was clicked last is gone, and the rest stays
![selection-textview](/uploads/66e409a784300925019c4017229cd87b/selection-textview.webm)https://gitlab.gnome.org/GNOME/gtk/-/issues/3396changing textview text properties breaks existing tags in widget-factory2021-04-03T11:09:28ZMohammed Sadiqchanging textview text properties breaks existing tags in widget-factoryChanging textview text properties breaks the existing text tags when the selection changes.
This was tested with `gtk4-widget-factory` textview in page 1. See attached video for details
![text-view-factory](/uploads/190175c3ab1e59ca3f2...Changing textview text properties breaks the existing text tags when the selection changes.
This was tested with `gtk4-widget-factory` textview in page 1. See attached video for details
![text-view-factory](/uploads/190175c3ab1e59ca3f250bbf99641f1a/text-view-factory.webm)https://gitlab.gnome.org/GNOME/gtk/-/issues/3391Change check- and radio-menuitems behavior2020-12-04T14:25:58ZArnaud B.arnaud.bonatti@gmail.comChange check- and radio-menuitems behavior<small>(Split from #3383.)</small>
In Gtk3, popovers created *via* `gtk_menu_button_set_menu_model()` close when a “simple-entry” is clicked, but stay opened when it’s a “check-entry” or a “radio-entry”.
This behavior looks good (examp...<small>(Split from #3383.)</small>
In Gtk3, popovers created *via* `gtk_menu_button_set_menu_model()` close when a “simple-entry” is clicked, but stay opened when it’s a “check-entry” or a “radio-entry”.
This behavior looks good (examples taken from various games I maintain):
* a “Sound” check-entry will not give directly feedback, so having the popover staying opened allows to see whether you correctly enabled/disabled the sound, and didn’t clicked out of the popover for example;
* you can replace a Preferences dialog easily by a menu/submenu, because multiple booleans options could be toggled at the same time;
* in an “Appearance” menu, multiple radio-entries allows to test every game theme, for finding the one you need, before closing the popover once you choosed;
* a simple-entry will usually do something visible anyway, like open a dialog (think of “Help” or “About”), so closing is okay
![Capture_d_écran_vidéo_de_23-11-2020_14_34_53](/uploads/cdcc3b07d342b0dfff84d1c1ab4b4efb/Capture_d_écran_vidéo_de_23-11-2020_14_34_53.webm)
This behavior is not the one currently in Gtk4, because the created popover is now a “standard menu,” that (IIUC) comes from the time when we used menubars: for now, every menuitem clicked closes the menu. It would be quite better for every “standard menu” to use the Gtk3 behavior of the popover from the `GtkMenuButton`, because that’s working great with current UI paradigms.https://gitlab.gnome.org/GNOME/gtk/-/issues/3388Scrolling hangs under Nvidia/X11 if Window tiled to the side2021-04-03T11:09:28ZaeldemeryScrolling hangs under Nvidia/X11 if Window tiled to the sideSorry couldn't take a gif.
As the title said there is some issue with scrolling under proprietary NVIDIA driver and X11 specially if the window is attached to the edge of screen/tiled.
Manjaro-Linux
```
lspci | grep VGA
01:00.0 VGA co...Sorry couldn't take a gif.
As the title said there is some issue with scrolling under proprietary NVIDIA driver and X11 specially if the window is attached to the edge of screen/tiled.
Manjaro-Linux
```
lspci | grep VGA
01:00.0 VGA compatible controller: NVIDIA Corporation GP104BM [GeForce GTX 1070 Mobile] (rev a1)
```
```
lshw -C video
*-display
description: VGA compatible controller
product: GP104BM [GeForce GTX 1070 Mobile]
vendor: NVIDIA Corporation
physical id: 0
bus info: pci@0000:01:00.0
version: a1
width: 64 bits
clock: 33MHz
capabilities: pm msi pciexpress vga_controller bus_master cap_list rom
configuration: driver=nvidia latency=0
resources: irq:138 memory:bb000000-bbffffff memory:90000000-9fffffff memory:a0000000-a1ffffff ioport:4000(size=128) memory:c0000-dffff
```
installed package versions
```
xorg-server 1.20.9.r21.g5c400cae1-1
nvidia 455.45.01
mutter 3.38.1-1
```
![VID_20201122_204704](/uploads/f94e2efdd9e31e6762ea07a72fc8569e/VID_20201122_204704.mp4)https://gitlab.gnome.org/GNOME/gtk/-/issues/3378Part hidden text in "Modified" column when "save as" file chooser is opened2020-11-19T16:02:24ZSTOMPY05Part hidden text in "Modified" column when "save as" file chooser is opened## Steps to reproduce
Open save / save-as file dialog box in any application that opts to save in an empty directory or a directory that has no current file types as the one currently being saved.
## Current behavior
Modified column tex...## Steps to reproduce
Open save / save-as file dialog box in any application that opts to save in an empty directory or a directory that has no current file types as the one currently being saved.
## Current behavior
Modified column text is part hidden until window, columns, or sidebar are resized.
## Expected outcome
"Modified" column text should appear in full regardless.
## Version information
GTK Version 3.24.20-0ubuntu1
Ubuntu 20.04
## Additional information
![Screenshot](/uploads/c989c4115ced826737153bd3499dfe58/Screenshot.png)https://gitlab.gnome.org/GNOME/gtk/-/issues/3375GtkListBox .separators and the last row2024-02-29T17:57:06ZSophie HeroldGtkListBox .separators and the last rowThe gtk4-demo ListBox/Controls uses a `.separators` ListBox with a GtkFrame around it. The last row gets a `border-bottom` from `.separators` and the GtkFrame border below. This looks not intentional.
I think that not only the demo shou...The gtk4-demo ListBox/Controls uses a `.separators` ListBox with a GtkFrame around it. The last row gets a `border-bottom` from `.separators` and the GtkFrame border below. This looks not intentional.
I think that not only the demo should be fixed but there should be a nice solution for this design pattern, since it is quite common. Favorably without GtkFrame to support placeholder content without a frame.
_LibHandy fixes this for it's `.content` class by using `:last-child`. However, this approach breaks if the ListBox has a placeholder. I would have fixed it using `:last-of-type` but that's not available in gtk. (Uses of pseudo classes were removed anyways because they are expensive bfe5b0d1b7a13749642a94a8a75fd82b6401a9ae)_https://gitlab.gnome.org/GNOME/gtk/-/issues/3367Clipboard hang destination app2020-12-17T15:50:18ZNicolas GoyClipboard hang destination appThe firefox bug with more details is: https://bugzilla.mozilla.org/show_bug.cgi?id=1631061
I open an issue here as I am not sure if this is now a GTK bug, and many GTK devs know more and can help me debug this.The firefox bug with more details is: https://bugzilla.mozilla.org/show_bug.cgi?id=1631061
I open an issue here as I am not sure if this is now a GTK bug, and many GTK devs know more and can help me debug this.https://gitlab.gnome.org/GNOME/gtk/-/issues/3365Make gtk_tree_model_foreach return last walked path2020-11-15T22:59:40ZcryptogopherMake gtk_tree_model_foreach return last walked pathLink to code: https://gitlab.gnome.org/GNOME/gtk/-/blob/master/gtk/gtktreemodel.c#L2015
Right now `gtk_tree_model_foreach()` return type is void.
`gtk_tree_model_foreach()` calls the user defined `GtkTreeModelForeachFunc` callback/func...Link to code: https://gitlab.gnome.org/GNOME/gtk/-/blob/master/gtk/gtktreemodel.c#L2015
Right now `gtk_tree_model_foreach()` return type is void.
`gtk_tree_model_foreach()` calls the user defined `GtkTreeModelForeachFunc` callback/func:
```c
* Calls func on each node in model in a depth-first fashion.
*
* If @func returns %TRUE, then the tree ceases to be walked,
* and gtk_tree_model_foreach() returns.
```
It would be useful if `gtk_tree_model_foreach()` returned the tree path, for which `@func` returned TRUE.
Such functionality would allow using `gtk_tree_model_foreach()` to search through tree items until matching element is found.https://gitlab.gnome.org/GNOME/gtk/-/issues/3360Add PackType to flowbox2020-11-15T16:11:28ZaggelalexAdd PackType to flowboxCurrently there's no way to position the children in a flowbox at the end or at the start of the container.Currently there's no way to position the children in a flowbox at the end or at the start of the container.https://gitlab.gnome.org/GNOME/gtk/-/issues/3357Editing and Drag-and-Drop demo click criticals and warnings2021-04-03T11:09:28ZIvan Molodetskikhyalterz@gmail.comEditing and Drag-and-Drop demo click criticals and warnings## Steps to reproduce
1. Open Editing and Drag-and-Drop `gtk4-demo`.
2. Click on one of the things.
<!--
You should try and reproduce with the demos applications available
under the `demos` directory, or the test programs in the ...## Steps to reproduce
1. Open Editing and Drag-and-Drop `gtk4-demo`.
2. Click on one of the things.
<!--
You should try and reproduce with the demos applications available
under the `demos` directory, or the test programs in the `tests` directory.
Alternatively, please attach a *small and self-contained* example
*written in C* that exhibits the issue.
-->
## Current behavior
<!--
Please describe the current behaviour
-->
```
(gtk4-demo:301327): Gtk-WARNING **: 22:53:41.407: Allocating size to GtkEntry 0x2c72d00 without calling gtk_widget_measure(). How does the code know the size to allocate?
(gtk4-demo:301327): Gtk-CRITICAL **: 22:53:41.407: Allocation width too small. Tried to allocate 48x24, but GtkEntry 0x2c72d00 needs at least 51x24.
(gtk4-demo:301327): Gtk-CRITICAL **: 22:53:41.407: Allocation width too small. Tried to allocate 42x18, but GtkText 0x2774800 needs at least 45x18.
```
## Version information
<!--
- Which version of GTK you are using
- What operating system and version
- For Linux, which distribution
- If you built GTK yourself, the list of options used to configure the build
-->
Fedora 33, Wayland, 25e6988d95f951ca2464bdcda8eeac1bbb4657fd default options.https://gitlab.gnome.org/GNOME/gtk/-/issues/3351GtkTreeView: item selection with Shift key - Different behavior depending on ...2021-04-28T16:09:44ZKevin Ilphrin PelletGtkTreeView: item selection with Shift key - Different behavior depending on the "direction"GIMP version: 2.99.2
Operating System: Pop_OS 20.10 64 bits
Package: Flatpak beta
# Description of the bug
<!--Please describe your issue with details.
Add screenshot or other files if needed.(write it after the > symbol)-->
When I ...GIMP version: 2.99.2
Operating System: Pop_OS 20.10 64 bits
Package: Flatpak beta
# Description of the bug
<!--Please describe your issue with details.
Add screenshot or other files if needed.(write it after the > symbol)-->
When I have more than 2 layers (I have 5 layers in my example GIF), and layers 2 and 3 are selected there is two behavior:
- If layer 3 is the last I selected, and i select the layer 4 by using the shift key, layers 3 and 4 are selected
- If layer 2 is the last I selected, and I select the layer 1 (still using the Shift key method), layers 1, 2 and 3 are selected
![GimpMultiLayersSelectionBug](/uploads/41bfc19a758a695ee793ee5ebff3ad2f/GimpMultiLayersSelectionBug.gif)
# Reproduction
Is the bug reproducible? Always
Reproduction steps:
1. Create more than two layers
2. Select the first one
3. Select the second by keeping Shift key pressed
4. Select the third layer (still keeping the Shift key pressed)
The first layer is then deselected
…
Expected result: Should have the same behavior when selecting from bottom to top or from top to bottom
Actual result: From bottom to top, it selects the very first layer selected to the very last. From top to bottom, it selects the previous clicked layer to the very last layer
Note: French guy here, apologies if my English is poor :') Thank you all for your awesome work on GIMP, I'll keep trying the latest 2.99 and report new bugs :3https://gitlab.gnome.org/GNOME/gtk/-/issues/3349GtkCellRendererAccel accel edited signal prints “Gdk-CRITICAL **:”2020-11-12T18:57:03ZMichael BuzgaruGtkCellRendererAccel accel edited signal prints “Gdk-CRITICAL **:”I am monitoring the GtkCellRendererAccel’s `accel-edited` Signal and every time the `accel_edited_clbk` it is called I am getting this message in the console:
> (Demo:5644): Gdk-CRITICAL **: 13:37:41.156: gdk_window_get_window_type: ass...I am monitoring the GtkCellRendererAccel’s `accel-edited` Signal and every time the `accel_edited_clbk` it is called I am getting this message in the console:
> (Demo:5644): Gdk-CRITICAL **: 13:37:41.156: gdk_window_get_window_type: assertion 'GDK_IS_WINDOW (window)' failed
Here is the code which explains the situation:
`
#include <gtk/gtk.h>
enum
{
ACCEL_MODS_COL,
ACCEL_KEY_COL,
KEYCODE_COL,
N_COLS
};
struct entry
{
gint accel_mods;
guint accel_key;
guint keycode;
} entry;
static struct entry entries[] =
{
{ GDK_CONTROL_MASK, GDK_KEY_X, 0 }
};
GtkTreeModel *create_model ( void );
void create_renderer_column ( GtkWidget *tree_view, GtkTreeModel *tree_model );
static void accel_edited_clbk ( G_GNUC_UNUSED GtkCellRendererText *cell,
const gchar *path_string,
guint key_val,
GdkModifierType modifier,
guint keycode,
gpointer data )
{
GtkTreeIter iter;
GtkTreeModel *model;
GtkTreePath *path;
/// ***
model = GTK_TREE_MODEL ( data );
path = gtk_tree_path_new_from_string ( path_string );
gtk_tree_model_get_iter ( model, &iter, path );
/// ***
gtk_tree_store_set ( GTK_TREE_STORE ( model ), &iter,
ACCEL_MODS_COL, modifier,
ACCEL_KEY_COL, key_val,
KEYCODE_COL, keycode,
-1 );
/// ***
gtk_tree_path_free ( path );
/// ***
g_print ( "%u %d %u\n", key_val, modifier, keycode );
}
gint main ( void )
{
GtkWidget *window;
GtkWidget *tree_view;
GtkTreeModel *tree_model;
/// ***
gtk_init ( NULL, NULL );
/// *** Crate the Window
window = gtk_window_new ( GTK_WINDOW_TOPLEVEL );
gtk_window_set_default_size ( GTK_WINDOW ( window ), 200, 200 );
/// *** Create the Model
tree_model = create_model();
/// *** Create the View
tree_view = gtk_tree_view_new_with_model ( tree_model );
gtk_container_add ( GTK_CONTAINER ( window ), tree_view );
/// *** Create the Renderer and the Column
create_renderer_column ( tree_view, tree_model );
/// *** Showing all
gtk_widget_show_all ( window );
/// ***
gtk_main ();
}
GtkTreeModel *create_model ( void )
{
GtkTreeStore *tree_store;
GtkTreeIter iter;
/// ***
tree_store = gtk_tree_store_new ( N_COLS, G_TYPE_INT, G_TYPE_UINT, G_TYPE_UINT );
/// ***
for ( guint index = 0; index < G_N_ELEMENTS ( entries ) ; index++ )
{
gtk_tree_store_append ( tree_store, &iter, NULL );
gtk_tree_store_set ( tree_store, &iter,
ACCEL_MODS_COL, entries[index].accel_mods,
ACCEL_KEY_COL, entries[index].accel_key,
KEYCODE_COL, entries[index].keycode,
-1 );
}
/// ***
return GTK_TREE_MODEL ( tree_store );
}
void create_renderer_column ( GtkWidget *tree_view, GtkTreeModel *tree_model )
{
GtkCellRenderer *renderer;
GtkTreeViewColumn *column;
/// ***
renderer = gtk_cell_renderer_accel_new ();
g_object_set ( renderer,
"accel-mode", GTK_CELL_RENDERER_ACCEL_MODE_GTK,
"editable", TRUE,
NULL );
/// ***
g_signal_connect ( renderer, "accel-edited", G_CALLBACK ( accel_edited_clbk ), tree_model );
/// ***
column = gtk_tree_view_column_new_with_attributes ( "Accel", renderer,
"accel-mods", ACCEL_MODS_COL,
"accel-key", ACCEL_KEY_COL,
"keycode", KEYCODE_COL,
NULL );
/// ***
gtk_tree_view_append_column ( GTK_TREE_VIEW ( tree_view ), column );
}
`
I tried to debug the code with:
`G_DEBUG=fatal-criticals gdb --args ./Demo`
but with no result, because I cannot proceed with the debbuger once I try to input a new accelerator the screen freezes.https://gitlab.gnome.org/GNOME/gtk/-/issues/3346GtkFontChooserWidget causes massive spike in CPU usage of application, after ...2023-07-09T15:53:47ZChigozirim ChukwuGtkFontChooserWidget causes massive spike in CPU usage of application, after second sequential use## Steps to reproduce
1. Open any gtk app which uses `GtkFontChooserWidget` to choose fonts (geany, gedit, gnome-builder, etc)
2. Search for and choose a font
3. Open the font chooser again and try to search for more fonts
**Origina...## Steps to reproduce
1. Open any gtk app which uses `GtkFontChooserWidget` to choose fonts (geany, gedit, gnome-builder, etc)
2. Search for and choose a font
3. Open the font chooser again and try to search for more fonts
**Originally [reported](https://gitlab.xfce.org/xfce/xfce4-settings/-/issues/243) for xfce. Complete with demos**
The `gtk-demo` app uses a different kind of fontchooser which does not give the option to search for fonts, so I can't reproduce using it.
## Current behavior
Searching for fonts a second time, results in huge spike in CPU usage and the application making use of `GtkFontChooserWidget` has to be force-closed.
## Expected outcome
Searching for fonts should not affect the application rendering it unusable.
## Version information
- Gtk Version: `3.24.23` installed from [aur](https://aur.archlinux.org/packages/gtk3-classic/)
- System Info: `Linux ArcoB 5.9.6-arch1-1 #1 SMP PREEMPT Thu, 05 Nov 2020 21:00:46 +0000 x86_64 GNU/Linux`
## Additional information
From the original bug report:
#### This shows the behaviour
![](https://gitlab.xfce.org/xfce/xfce4-settings/uploads/d322e5170424a8bc29d807aabe33a52c/Peek_2020-11-10_12-07.gif)
#### This shows that we can search for fonts the first time around, but the second time results in the CPU spike
![](https://gitlab.xfce.org/xfce/xfce4-settings/uploads/0fe461f1beff65a408df816a29c76fce/Peek_2020-11-10_12-28.gif)
See the [original bug report](https://gitlab.xfce.org/xfce/xfce4-settings/-/issues/243) for more infohttps://gitlab.gnome.org/GNOME/gtk/-/issues/3338File Chooser: recent files from network folders "load in" displacing files al...2021-04-12T05:01:12ZIvan Molodetskikhyalterz@gmail.comFile Chooser: recent files from network folders "load in" displacing files already there## Steps to reproduce
1. Open Video Player `gtk4-demo`.
2. Populate recent files with a local video and then a few remote videos by opening them in that order.
3. Open the file chooser again.
<!--
You should try and reproduce with...## Steps to reproduce
1. Open Video Player `gtk4-demo`.
2. Populate recent files with a local video and then a few remote videos by opening them in that order.
3. Open the file chooser again.
<!--
You should try and reproduce with the demos applications available
under the `demos` directory, or the test programs in the `tests` directory.
Alternatively, please attach a *small and self-contained* example
*written in C* that exhibits the issue.
-->
## Current behavior
<!--
Please describe the current behaviour
-->
Network files "load in" a little later, displacing older local files and potentially messing up your selection.
## Expected outcome
<!--
Please describe the expected outcome
-->
Existing entries in the list should not jump around. Maybe all files that show after the first frame should be appended rather than inserted in the middle? Maybe show placeholders if they are instantly available?
## Version information
<!--
- Which version of GTK you are using
- What operating system and version
- For Linux, which distribution
- If you built GTK yourself, the list of options used to configure the build
-->
Fedora 33, Wayland, self-built gtk4 4e8a5f2ec8866265e5967b31415d0da83e7327fa with default settings.
## Additional information
<!--
- Screenshots or screen recordings are useful for visual errors
- Please report any warning or message printed on the terminal
-->
Imagine that I wanted to select the bottom (local) video.
![out](/uploads/7f114d73f2654fe12457be6ab94c95c9/out.mp4)
This gets really annoying with like 20 remote entries above the one that you need.https://gitlab.gnome.org/GNOME/gtk/-/issues/3331GTK3: gtk_expander_set_label_fill () is no-op2021-01-31T23:47:02ZcrviGTK3: gtk_expander_set_label_fill () is no-opThe resize code which uses the `'label-fill'` property, got nuked as part of https://gitlab.gnome.org/GNOME/gtk/-/commit/2e27c0af25c12642a65b5611ee9f08c7406d6171The resize code which uses the `'label-fill'` property, got nuked as part of https://gitlab.gnome.org/GNOME/gtk/-/commit/2e27c0af25c12642a65b5611ee9f08c7406d6171https://gitlab.gnome.org/GNOME/gtk/-/issues/3329Going back up one directory in save dialog prevents save or obscures view wit...2022-02-22T21:41:37ZJacob GustafsonGoing back up one directory in save dialog prevents save or obscures view with popup conditionally## Steps to reproduce
1. Press Ctrl+S in a GTK application while an untitled document is open (such as in Gedit or Geany).
2. Enter a writable directory inside another writable directory (such as `/home/user/Documents/Programming/Cpp`...## Steps to reproduce
1. Press Ctrl+S in a GTK application while an untitled document is open (such as in Gedit or Geany).
2. Enter a writable directory inside another writable directory (such as `/home/user/Documents/Programming/Cpp`).
3. Navigate "up" by one directory (such as by clicking "Programming" in the breadcrumbs at the top of the save window in the example given in 2 above).
4. Press "Save"
## Current behavior
The outcome is inconsistent:
A. The document doesn't save, and the window navigates back to the selected directory.
- Behavior A seems reproducible by initially entering the deepest directory using the directory list then clicking back by one using the breadcrumbs (rather than going back and forth using the breadcrumbs).
B. (1)A popup appears in the filename box--(2)if there are many files, it will obscure other widgets and prevent navigation. The dialog stays open.
- Behavior B seems reproducible by clicking back and forth between the shallow and deep breadcrumbs (rather than using directory list to enter the directory initially). The filename must be different than some existing file (Save "Untitled Document 1") then "Save As" then replace 1 with 2, then follow the steps 2-3 reproduce the issue.
C. The document saves in the clicked directory and the dialog closes as expected.
For B, the workaround is to go back up two directories (such as by clicking "Documents" in the breadcrumbs in the example scenario) then double-click the directory you want (such as by double-clicking "Programming" in the directory list).
## Expected outcome
1. Behavior C should occur: Pressing a breadcrumb should deselect the deeper directory the user tried to exit by clicking it, allowing the user to click save to save in the clicked directory.
2. Clicking a breadcrumb should never cause the autocomplete popup to appear.
## Version information
- GTK: gtk3-3.24.23-1.fc32.x86_64 gtk2-2.24.32-7.fc32.x86_64
- Fedora 32
## Additional information
Behavior A:
![image](/uploads/612878dfbb94661254cde5d59fd2d014/image.png)
Behavior B(1):
![image](/uploads/7eea88239e7ebc847afeec7f228eb379/image.png)
Behavior B(2):
![image](/uploads/4018ca895a4a31b84038d7ae35b4de05/image.png)https://gitlab.gnome.org/GNOME/gtk/-/issues/3327Geary hangs on start2020-11-07T18:12:51ZIvan Molodetskikhyalterz@gmail.comGeary hangs on start**Bug Summary**
After Ctrl-Q-ing nightly Flatpak Geary (from 2020-11-05 08:36:27 +0000) it no longer starts. The backtrace where it hangs is as follows:
```
#0 0x00007fb60292f4bf in __GI___poll (fds=0x55f613882ad0, nfds=1, timeout=-1)...**Bug Summary**
After Ctrl-Q-ing nightly Flatpak Geary (from 2020-11-05 08:36:27 +0000) it no longer starts. The backtrace where it hangs is as follows:
```
#0 0x00007fb60292f4bf in __GI___poll (fds=0x55f613882ad0, nfds=1, timeout=-1) at ../sysdeps/unix/sysv/linux/poll.c:29
#1 0x00007fb602c484ce in g_main_context_poll (priority=<optimized out>, n_fds=1, fds=0x55f613882ad0, timeout=<optimized out>, context=0x55f613812e40) at ../glib/gmain.c:4422
#2 g_main_context_iterate (context=0x55f613812e40, block=block@entry=1, dispatch=dispatch@entry=1, self=<optimized out>) at ../glib/gmain.c:4114
#3 0x00007fb602c48853 in g_main_loop_run (loop=0x55f6137edd30) at ../glib/gmain.c:4317
#4 0x00007fb602b200ef in g_dbus_connection_send_message_with_reply_sync
(connection=connection@entry=0x55f613820160 [GDBusConnection], message=message@entry=0x7fb5e8006800 [GDBusMessage], flags=flags@entry=G_DBUS_SEND_MESSAGE_FLAGS_NONE, timeout_msec=timeout_msec@entry=2147483647, out_serial=out_serial@entry=0x0, cancellable=cancellable@entry=0x0, error=0x7ffefdd82790) at ../gio/gdbusconnection.c:2164
#5 0x00007fb602b2050c in g_dbus_connection_call_sync_internal
(connection=0x55f613820160 [GDBusConnection], bus_name=bus_name@entry=0x55f613812f60 ":1.116", object_path=0x55f613812b70 "/org/freedesktop/portal/desktop", interface_name=interface_name@entry=0x55f613812ba0 "org.freedesktop.portal.Settings", method_name=method_name@entry=0x7fb601d8a7f8 "ReadAll", parameters=parameters@entry=0x7fb5e8008630, reply_type=0x7fb602b81666, flags=G_DBUS_CALL_FLAGS_NONE, timeout_msec=2147483647, fd_list=0x0, out_fd_list=0x0, cancellable=0x0, error=0x7ffefdd82920) at ../gio/gdbusconnection.c:5991
#6 0x00007fb602b22aa9 in g_dbus_connection_call_with_unix_fd_list_sync
(connection=<optimized out>, bus_name=bus_name@entry=0x55f613812f60 ":1.116", object_path=<optimized out>, interface_name=interface_name@entry=0x55f613812ba0 "org.freedesktop.portal.Settings", method_name=method_name@entry=0x7fb601d8a7f8 "ReadAll", parameters=parameters@entry=0x7fb5e8008630, reply_type=0x0, flags=G_DBUS_CALL_FLAGS_NONE, timeout_msec=2147483647, fd_list=0x0, out_fd_list=0x0, cancellable=0x0, error=0x7ffefdd82920) at ../gio/gdbusconnection.c:6365
#7 0x00007fb602b2d6c6 in g_dbus_proxy_call_sync_internal
(proxy=0x55f613833ba0 [GDBusProxy], method_name=method_name@entry=0x7fb601d8a7f8 "ReadAll", parameters=0x7fb5e8008630, flags=flags@entry=G_DBUS_CALL_FLAGS_NONE, timeout_msec=timeout_msec@entry=2147483647, fd_list=fd_list@entry=0x0, out_fd_list=0x0, cancellable=0x0, error=0x7ffefdd82920) at ../gio/gdbusproxy.c:2834
#8 0x00007fb602b2ec08 in g_dbus_proxy_call_sync
(proxy=<optimized out>, method_name=method_name@entry=0x7fb601d8a7f8 "ReadAll", parameters=<optimized out>, flags=flags@entry=G_DBUS_CALL_FLAGS_NONE, timeout_msec=timeout_msec@entry=2147483647, cancellable=cancellable@entry=0x0, error=0x7ffefdd82920) at ../gio/gdbusproxy.c:3026
#9 0x00007fb601d2f76b in init_settings (screen=0x55f6138013e0 [GdkWaylandScreen]) at ../gdk/wayland/gdkscreen-wayland.c:736
#10 _gdk_wayland_screen_new (display=display@entry=0x55f613878190 [GdkWaylandDisplay]) at ../gdk/wayland/gdkscreen-wayland.c:1319
#11 0x00007fb601d28563 in _gdk_wayland_display_open (display_name=<optimized out>) at ../gdk/wayland/gdkdisplay-wayland.c:617
#12 0x00007fb601cf3feb in gdk_display_manager_open_display (manager=<optimized out>, name=0x0) at ../gdk/gdkdisplaymanager.c:462
#13 0x00007fb601fe559b in gtk_init_check (argc=<optimized out>, argv=<optimized out>) at ../gtk/gtkmain.c:1109
#14 gtk_init_check (argc=<optimized out>, argv=<optimized out>) at ../gtk/gtkmain.c:1101
#15 0x00007fb601fe55cd in gtk_init (argc=argc@entry=0x0, argv=argv@entry=0x0) at ../gtk/gtkmain.c:1166
#16 0x00007fb601e94225 in gtk_application_startup (g_application=0x55f61380a1b0 [ApplicationClient]) at ../gtk/gtkapplication.c:304
#17 0x00007fb602e9e85f in application_client_real_startup (base=<optimized out>) at ../src/client/application/application-client.vala:379
#18 0x00007fb602d3cc36 in _g_closure_invoke_va (closure=<optimized out>, return_value=<optimized out>, instance=<optimized out>, args=<optimized out>, n_params=<optimized out>, param_types=<optimized out>)
at ../gobject/gclosure.c:873
#19 0x00007fb602d55c7f in g_signal_emit_valist (instance=0x55f61380a1b0, signal_id=<optimized out>, detail=0, var_args=var_args@entry=0x7ffefdd82ce0) at ../gobject/gsignal.c:3403
#20 0x00007fb602d55e63 in g_signal_emit (instance=instance@entry=0x55f61380a1b0, signal_id=<optimized out>, detail=detail@entry=0) at ../gobject/gsignal.c:3550
#21 0x00007fb602af1622 in g_application_register (application=application@entry=0x55f61380a1b0 [ApplicationClient], cancellable=cancellable@entry=0x0, error=error@entry=0x7ffefdd82e00)
at ../gio/gapplication.c:2204
#22 0x00007fb602af1a1e in g_application_real_local_command_line (application=0x55f61380a1b0 [ApplicationClient], arguments=0x7ffefdd82eb8, exit_status=0x7ffefdd82e54) at ../gio/gapplication.c:1106
#23 0x00007fb602e9da2e in application_client_real_local_command_line (base=<optimized out>, args=0x7ffefdd82eb8, exit_status=0x7ffefdd82eb4) at ../src/client/application/application-client.vala:349
#24 0x00007fb602af1d66 in g_application_run (application=0x55f61380a1b0 [ApplicationClient], argc=-36163916, argv=<optimized out>) at ../gio/gapplication.c:2528
#25 0x000055f61370b33a in _vala_main (args=0x7ffefdd83018, args_length1=1) at ../src/client/application/main.vala:15
#26 0x00007fb602860022 in __libc_start_main (main=0x55f61370b1c0 <main>, argc=1, argv=0x7ffefdd83018, init=<optimized out>, fini=<optimized out>, rtld_fini=<optimized out>, stack_end=0x7ffefdd83008)
at ../csu/libc-start.c:308
#27 0x000055f61370b20e in _start () at ../sysdeps/x86_64/start.S:120
```
I'm gonna guess it'll start fine after a re-log-in but still that this happens is annoying when I need to reset Geary state. I don't think it will happen every time when exiting and starting Geary back because I think I hit something like this before occasionally (although it tended to fix itself after a minute, but now it's stuck for good).
Fedora 33, Wayland.
<!-- Do not remove the following line -->https://gitlab.gnome.org/GNOME/gtk/-/issues/3324Feature: Add focus-follows-mouse capability to GTK32020-11-04T19:30:46ZJeffery SmallFeature: Add focus-follows-mouse capability to GTK3A developer on GIMP suggested that I make a request that might be more appropriate to the underlying toolkit.
I would like to see a focus-follows-mouse be implemented in GIMP as the cursor is dragged between various menus, tools and the...A developer on GIMP suggested that I make a request that might be more appropriate to the underlying toolkit.
I would like to see a focus-follows-mouse be implemented in GIMP as the cursor is dragged between various menus, tools and the canvas. The suggestion was that this would be easier for GIMP developers if the ability to switch between focus-follows-click vs. focus-follows-mouse was part of the toolkit.
Is this something that already exists, ore could it be added to GTK3?https://gitlab.gnome.org/GNOME/gtk/-/issues/3321Dragging widgets in headerbar with touch drags the window instead2023-02-18T17:10:08ZAlice MikhaylenkoDragging widgets in headerbar with touch drags the window insteadThe most noticeable case here is selecting text in entries, which fails as follows:
![Screencast_from_02-11-20_17_04_44](/uploads/340e0873475939fef754a634b1bfd3e0/Screencast_from_02-11-20_17_04_44.webm)
Other cases:
* Dragging a switch...The most noticeable case here is selecting text in entries, which fails as follows:
![Screencast_from_02-11-20_17_04_44](/uploads/340e0873475939fef754a634b1bfd3e0/Screencast_from_02-11-20_17_04_44.webm)
Other cases:
* Dragging a switch
* Dragging from a button (dragging the window here is expected, but it leaves the button in active state)
It seems it's caused by the touch pointer emulation synthetic events, which lack an event sequence and so aren't blocked when that sequence is denied.https://gitlab.gnome.org/GNOME/gtk/-/issues/3317Touch text selection popovers are misplaced on rotated text2021-04-03T11:11:41ZAlice MikhaylenkoTouch text selection popovers are misplaced on rotated text![Screenshot_from_2020-11-02_14-17-54](/uploads/20dfed549f7e2da0dd3a422a12280cc1/Screenshot_from_2020-11-02_14-17-54.png)
![Screenshot_from_2020-11-02_14-17-58](/uploads/e114eb8ad14289a0524155c734931fec/Screenshot_from_2020-11-02_14-17-...![Screenshot_from_2020-11-02_14-17-54](/uploads/20dfed549f7e2da0dd3a422a12280cc1/Screenshot_from_2020-11-02_14-17-54.png)
![Screenshot_from_2020-11-02_14-17-58](/uploads/e114eb8ad14289a0524155c734931fec/Screenshot_from_2020-11-02_14-17-58.png)
Also visible that one of the handles is in the wrong state.