The custom listboxes serve another purpose: to handle
row activation. Instead of using a custom listbox, we
can simply listen to the `activated` signal on the
rows that don't have a switch.
We were using this to move between groups and scroll
the page, but that's not necessary. All we need
to do is set `scroll-to-focus` on our GtkViewport.
This commit removes the keynav-failed callback
and all associated plumbing.
Only the shell can know how the description string will be used and if
or when it should be escaped. Previously the shell did not escape the
description before displaying it with markup to highlight search hits,
but now it does. For gnome-control-center this however means without
this change the string will be escaped twice causing markup to show up
in the description text.
Related: https://gitlab.gnome.org/GNOME/gnome-shell/-/merge_requests/2033
Make corners rounded.
Replace selected style with an overlay check, make the close button
smaller to match. Move slideshow indicator to the lower left corner.
Set margins in css.
Update the unused symbolic to match the svg with a baked-in shadow,
actually use it. Bundle it in gresources instead of installing, fix the
style class name. Fix margins - the close button was overriding them.
The timezone map tries to highlight regions of the world that keep the
same time, approximating this by their non-daylight-savings UTC offset.
There's no reasonable API for this, and it goes wrong in various cases,
such as Irish Standard Time which is legally defined as the country's
summer time with a negative DST offset in winter.
Hardcoding this is unpleasant, but there doesn't seem to be a better
solution, and in any case there's already similar hardcoding implied by
the segmented map images in panels/datetime/data/timezone_*.png. I've
tried to make it practical to fix other similar disagreements between
the detected offset and the groupings implied by map images, though for
now I've conservatively fixed only the case I'm familiar with.
Fixes: #1341
The code that set it could only ever have worked by luck in some
timezones on some systems, and there's no API to obtain the information
we were trying to store in it. Nothing uses it, so there's no point
trying to fix it.
tzname_normal is also unused, but slightly useful for debugging, so
rename it to tzname but otherwise keep it in place.
==31571== 120 bytes in 8 blocks are definitely lost in loss record 17,034 of 19,290
==31571== at 0x484086F: malloc (vg_replace_malloc.c:380)
==31571== by 0x4AF77A8: g_malloc (gmem.c:106)
==31571== by 0x4D9816A: gtk_accelerator_name (gtkaccelgroup.c:1577)
==31571== by 0x4A82D1: setup_custom_shortcut (cc-keyboard-shortcut-editor.c:364)
==31571== by 0x4A92DF: cc_keyboard_shortcut_editor_key_press_event (cc-keyboard-shortcut-editor.c:745)
==31571== 288 bytes in 24 blocks are definitely lost in loss record 18,138 of 19,290
==31571== at 0x484086F: malloc (vg_replace_malloc.c:380)
==31571== by 0x4AF77A8: g_malloc (gmem.c:106)
==31571== by 0x4A4119: variant_get_key_combos (cc-keyboard-item.c:475)
==31571== by 0x4A41FD: settings_get_key_combos (cc-keyboard-item.c:498)
==31571== by 0x4A46BE: cc_keyboard_item_load_from_gsettings (cc-keyboard-item.c:574)
==31571== by 0x4A5BBB: append_section (cc-keyboard-manager.c:315)
==31571== by 0x4A605D: append_sections_from_file (cc-keyboard-manager.c:431)
==31571== by 0x4A6766: reload_sections (cc-keyboard-manager.c:568)
==31571== by 0x4A6D68: cc_keyboard_manager_load_shortcuts (cc-keyboard-manager.c:707)
==31571== by 0x4A2FA4: cc_keyboard_shortcut_dialog_init (cc-keyboard-shortcut-dialog.c:841)
==31571== by 0x4A7A288: g_type_create_instance (gtype.c:1929)
==31571== by 0x4A61CAC: g_object_new_internal (gobject.c:1945)
==25729== 1,058 (40 direct, 1,018 indirect) bytes in 1 blocks are definitely lost in loss record 20,104 of 20,657
==25729== at 0x484086F: malloc (vg_replace_malloc.c:380)
==25729== by 0x4AF77A8: g_malloc (gmem.c:106)
==25729== by 0x4B0F1E4: g_slice_alloc (gslice.c:1072)
==25729== by 0x4B09060: UnknownInlinedFun (gsequence.c:1609)
==25729== by 0x4B09060: g_sequence_insert_before (gsequence.c:527)
==25729== by 0x4ED2FE7: gtk_list_store_insert (gtkliststore.c:1271)
==25729== by 0x4A5DCE: append_section (cc-keyboard-manager.c:363)
==25729== by 0x4A605D: append_sections_from_file (cc-keyboard-manager.c:431)
==25729== by 0x4A6766: reload_sections (cc-keyboard-manager.c:568)
==25729== by 0x4A6D33: cc_keyboard_manager_load_shortcuts (cc-keyboard-manager.c:706)
==25729== by 0x4A2FA4: cc_keyboard_shortcut_dialog_init (cc-keyboard-shortcut-dialog.c:841)
==25729== by 0x4A7A288: g_type_create_instance (gtype.c:1929)
==25729== by 0x4A61CAC: g_object_new_internal (gobject.c:1945)
The add_device() check for g_udev_device_get_parent() was leaking (see
below). As this is actually used in create_device(), I'm passing it as
argument to avoid a second call to that function.
> 34,175 (1,080 direct, 33,095 indirect) bytes in 27 blocks are definitely lost in loss record 19,729 of 19,758
> at 0x4A7A337: g_type_create_instance (gtype.c:1907)
> by 0x4A61CAC: g_object_new_internal (gobject.c:1945)
> by 0x4A62C7C: g_object_new_with_properties (gobject.c:2114)
> by 0x4A63780: g_object_new (gobject.c:1785)
> by 0x5EA5A6A: _g_udev_device_new (gudevdevice.c:132)
> by 0x4B6D24: add_device (gsd-device-manager.c:447)
> by 0x4B6FBF: gsd_device_manager_init (gsd-device-manager.c:517)
> by 0x4A7A288: g_type_create_instance (gtype.c:1929)
> by 0x4A61CAC: g_object_new_internal (gobject.c:1945)
> by 0x4A62C7C: g_object_new_with_properties (gobject.c:2114)
> by 0x4A63780: g_object_new (gobject.c:1785)
> by 0x4B7075: gsd_device_manager_get (gsd-device-manager.c:533)
I was seeing this critical when opening the VPN details page:
(gnome-control-center:78888): nm-CRITICAL **: 16:26:21.689: ((src/libnm-client-impl/nm-device.c:1299)): assertion '<dropped>' failed
It probably broke in 551d72280b. The code
above and below expects self->device may be NULL, so let's add a
NULL check here.
