Age | Commit message (Collapse) | Author | Files | Lines |
|
and use xalign
fixes https://github.com/mate-desktop/atril/issues/275
|
|
Handle the case where ev_selection_render_selection() doesn't actually
render a surface - this happens when a backend doesn't implement
render_selection().
https://bugzilla.gnome.org/show_bug.cgi?id=723431
origin commit:
https://git.gnome.org/browse/evince/commit/?id=c5a338c
fixes https://github.com/mate-desktop/atril/issues/283
|
|
|
|
|
|
https://github.com/linuxmint/xreader/commit/c47a1f4
|
|
Based on Evince commits 8b965e769808815df9c75a10a0e4972c0904d26a
and a612f809e4e23d88a661a9604ce84f332189a658.
https://github.com/linuxmint/xreader/commit/4eccf6b
|
|
|
|
|
|
After activating or deactivating fullscreen mode, the sidebar
window is automatically moved to its start, while scroll bar
stays in its original position.
The sidebar window move is unwanted and unsolicited, and it's
most probably caused by GtkIconView or GtkScrolledWindow bug.
Workaround this by having the sidebar sync its window with the
current scroll position after a fullscreen operation, do that by
just emitting a "value-changed" on the current scroll adjustment.
https://bugzilla.gnome.org/show_bug.cgi?id=783404
origin commit:
https://git.gnome.org/browse/evince/commit/?id=2bfb8fa
|
|
Preload both before and after current visible scrolling range,
the same amount of thumbnails in it, to help prevent thumbnail
creation happening in the user's sight.
https://bugzilla.gnome.org/show_bug.cgi?id=342110#c15
origin commit:
https://git.gnome.org/browse/evince/commit/?id=1fc8c68
|
|
Evince renders thumbnails on-the-fly as they get into the
scrolling visible area, but at the same time it will remove
them as they get out of the visible scrolling area, so when
user scrolls back to same position he will notice thumbnails
be recreated.
In pro of a more icing user experience, let's adopt a mixed approach
and keep the thumbnails that the user has already navigated, so
when he scrolls back and forth in the same area no thumbnail
re-generation will be visible. This also matches behaviour with
other pdf readers.
https://bugzilla.gnome.org/show_bug.cgi?id=342110
origin commit:
https://git.gnome.org/browse/evince/commit/?id=121e4d9
|
|
Caused by GtkIconView doing an invalidate and relayout of *all*
items in the view anytime we update model data in any indiviual
item (which happens with all the items that are getting in and out
of the scrolling area while we scroll). This caused GtkIconView to
machine-gunned us with "size-allocate" signals, a signal we were
using to update thumbnails when the sidebar is resized.
Fixed by connecting to the GtkTreeModel "row-changed" signal before
GtkIconView does it, and stop emission from there.
As we don't depend now on "size-allocate" signals to show thumbnails
while we scroll, just queue a draw on the icon view when a
thumbnail finish rendering.
Thanks Jose Aliste for first spotting the problem.
https://bugzilla.gnome.org/show_bug.cgi?id=691448
origin commit:
https://git.gnome.org/browse/evince/commit/?id=6480c70
Fixes https://github.com/mate-desktop/atril/issues/255
|
|
Expand the scrolled window vertical so that you can actually
see the content of the widget. This also turned up a missing
gtk_style_context_save () in gimpcellrenderertoggle.c.
So fix that as well.
partial inspired from:
https://github.com/linuxmint/xreader/commit/6b6ae7f
|
|
This is a regression from commit
https://github.com/mate-desktop/atril/commit/f8aec90
see:
https://github.com/linuxmint/xreader/commit/46298df
https://github.com/linuxmint/xreader/issues/12
|
|
Whole the function fullscreen_toolbar_remove_shadow isn't needed
without that deprecation.
Setting instead the widget name to make theming possible.
GtkToolbar-shadow-type is deprecated since gtk+-3.6
|
|
Fixes https://github.com/mate-desktop/atril/issues/264
|
|
|
|
Fixes #257
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
taken from:
https://git.gnome.org/browse/evince/commit/?h=gnome-3-8&id=23048af
|
|
To avoid duplicated code.
Taken from:
https://git.gnome.org/browse/evince/commit/?h=gnome-3-8&id=0897a7d
|
|
taken from:
https://git.gnome.org/browse/evince/commit/?h=gnome-3-8&id=0291193
|
|
https://bugzilla.gnome.org/show_bug.cgi?id=639403
taken from:
https://git.gnome.org/browse/evince/commit/?h=gnome-3-8&id=6c1d1a2
|
|
taken from:
https://git.gnome.org/browse/evince/commit/?h=gnome-3-8&id=19db23a
|
|
taken from:
https://git.gnome.org/browse/evince/commit/?h=gnome-3-8&id=751df481af5d6148302c047de948dd514888c578
|
|
In ev_view_accessible_get_character_extents().
taken from:
https://git.gnome.org/browse/evince/commit/?h=gnome-3-8&id=ae1ace1
|
|
In ev_view_accessible_get_offset_at_point().
taken from:
https://git.gnome.org/browse/evince/commit/?h=gnome-3-8&id=28bd405
|
|
and make them public internally
taken from:
https://git.gnome.org/browse/evince/commit/?h=gnome-3-8&id=d4e458c
|
|
https://bugzilla.gnome.org/show_bug.cgi?id=685828
taken from:
https://git.gnome.org/browse/evince/commit/?id=37c58de
|
|
|
|
taken from:
https://git.gnome.org/browse/evince/commit/?h=gnome-3-6&id=5a7ad5b
|
|
|
|
|
|
replace it with gtk_widget_set_focus_on_click
|
|
|
|
and avoid to use GtkStock
|
|
If you stay on the last page when you close it, reopen it will
not be positioned until you visited it (reset to the first).
In the changed code:
Judgment only works on the last page.
This function explicitly introduces this problem, so it can be
removed directly.
I have been tested successfully.
|