... which determines if a node's operation-class has a specific
key, and can be used instead of gimp_gegl_node_get_key() when only
existence is important, to avoid compiler warnings.
Update the rest of the code to use the new function.
I don't reproduce these build warnings in my build but frogonia reported
them:
> warning: 'strcpy' accessing 1 byte at offsets 0 and [0, 2147483645] may overlap 1 byte at offset 0 [-Wrestrict]
And indeed the man of strcpy() clearly states that "The strings may not
overlap" (which is clearly not guaranted in current code as we don't
know the size of what's after the substring).
strncpy() might have done the trick as it doesn't say anything about
possible overlap or not, but let's just use memmove() instead where it
is clearly written that memory areas may overlap (note that bcopy() as
proposed by ankh would have also done the trick, but it is marked as
deprecated).
(cherry picked from commit 291af71e95)
… defaults to FALSE for PNG, TIFF.
The logics is that in most case, when alpha is 0, you don't expect color
to be retrieved and some people were bothered by this when erasing data
with eraser tool or selection deletion.
For the WIP image as well as for XCF format, we definitely want to keep
color data as there are workflows where you would want to erase/unerase
pixels subtly (selection also are not binary and should not be expected
to erase color data on layers with alpha channel). So we want to keep
the non-destructive behavior there.
Yet for export formats, going for "deleting what looks deleted" is a
valid approach, especially as it has been argued that experienced users
would know how to read and change the format options (and these
experienced users are likely the ones who may have valid reasons to want
pixels with full transparency to keep color values).
PNG and TIFF were the only 2 formats where I see such options, so
changing the default for these 2 plug-ins.
(cherry picked from commit 105c4efc8b for
the TIFF one and adapted to gimp-2-10 code for PNG)
… oscp-router0[12].gnome.org.
The file was opened asynchronously through g_file_read_async() but the
actually reading from a GFileInputStream was actually done in the
callback, in the main thread. So it may be possible that file is
accessible but connection is very slow or non-reliable and actual
reading would block the main process. I am assuming this is what may
have happened for this reporter.
Instead let's use g_file_load_contents_async() which does the file
opening, reading and closing all in the asynchronous task, hence would
not slow down anything even if the remote resource is unavailable or
connection is very slow.
(cherry picked from commit cc16ec5f9f)
Apparently, the "drag-begin" signal of dockbook tabs can be raised
multiple times when a drag begins (this seems to happen randomly,
and rarely -- possibly a GTK bug). In
gimp_paned_box_drag_callback(), which gets called in response, make
sure not to leak the corresponding idle source in this case, which
can lead to a segfault if the widget is destroyed before the idle
is run.
(cherry picked from commit 6bae5d8cf7)
It was unneeded on master, surely because there must be a
gtk_widget_show_all() run somewhere down the line.
Also reuse the button widget (not so important, but while I'm at it).
In GimpFilterTool, use gimp_drawable_filter_set_preview(), added in
the previous commit, to toggle the filter's preview, instead of
removing and re-adding the filter. This avoids destroying cached
results when disabling the preview, allowing to quickly toggle the
preview on and off to compare the result.
(cherry picked from commit 912b36a8f8)
... which allows temporarily turning the applicator into a NOP,
without destroying cached data.
This commit also improves gimp_applicator_set_{src,dest}_buffer().
(cherry picked from commit e324e2d4b3)
Cropping tools will no longer discard layer data that falls out of the
crop region if the layer's pixels are locked
(cherry picked from commit 9a9e2e9a19)
Adds a new "Delete cropped pixels" toggle to the crop tool options,
to allow turning off the deletion of layer data that falls out of the
crop region.
(cherry picked from commit ee97c25b87)
Revert "app: also localize LC_TIME to GUI language."
This reverts commit a4e65ac836.
LC_TIME format apparently does not accept the broader format with just
the lang part, for instance "fr". The region is necessary, like "fr_FR".
In our supported languages, we were mostly using lang only except for a
few langs where we support regional variants.
This makes sense as time format may be quite different depending on
regions (dates are quite different in GB vs US, though that was actually
one of the cases where we were using accurate locale format).
And the main issue is that it was not only breaking time localization.
Normal string localization (LANGUAGE) got somehow broken as well by
broken LC_TIME.
Anyway for now let's just stick to whatever OS-set time format (even
though it may differ from GIMP Preferences-set language defaults).
(cherry picked from commit 39890a49eb)
After discussing with Schumaml and Houz on IRC, let's just drop the
generic comment which may be confusing and is probably unecessary,
especially now that I write the revision number (present and new ones).
So let's just assume people understand that they should update, even
when GIMP version stays the same (this was the case I was the most wary
about, that people think the notification is an error as since they have
the last version, they don't have to update GIMP).
I hesitated to at least leave just the simple "It is recommended to
update." sentence but after much hesitation, let's go with no generic
comment at all. The download button by itself should suffice.
Of course, when a specific revision comment is set in
gimp_versions.json, it is still displayed.
(cherry picked from commit 56ffa2fd11)
- gtk_show_uri_on_window() replaced by gtk_show_uri()
- gtk_button_new_from_icon_name replaced by
gtk_image_new_from_icon_name() in a GtkButton.
- gtk_widget_set_margin_top() removed.
While this issue was unseen so far on common desktop machines, the CI
build encountered it, hence failing 6 of the unit tests.
A connection to the bus could not be established hence the dbus_manager
was never allocated, and finally it would crash at exit if we tried to
unref it unconditionnally. Use g_clear_object() instead.
Also add some stderr output for easier debugging, for when one of the 2
possible error cases might happen (as documented by g_bus_own_name()).
(cherry picked from commit 2c1efdedf0)
- Instead of the download link being just a link button (similar to
browser "blue text" links), make it a proper button encompassing the
update icon and the download text "Download GIMP X.Y.Z".
- Make also the revision number part of the download information.
- Frame title is now "Update available!" instead of "New version
available!" (because it's not necessarily a new version, it can also
be just a package revision. "Update" is more generic).
- When no update is available, the "Check for updates" button is big and
has a label, and also now an icon. When an update is available, hence
we display now a button for the download link, the "Check for updates"
button becomes small with icon only in order not to take visual
priority over the download itself (yet allowing to re-check for even
newer information).
(cherry picked from commit a55e898c96)
As noted on IRC, after the first update check on a given day, it looks
like the check button does nothing. Therefore also display the check
time to clearly show the button click was taken into account (simply
there are no updates, hence only displayed check time is updated).
(cherry picked from commit a05b085112)
Do it even on stable builds as it can help diagnose bugs on releases.
Note that the issue we had just know with casing on the build-id would
not have been seen by such stderr output anyway. Just in case for future
issues.
