summaryrefslogtreecommitdiff
path: root/libmate-desktop
AgeCommit message (Collapse)AuthorFilesLines
2015-07-24bg: objects without a filename are not slideshowsMonsta1-0/+3
from https://github.com/GNOME/gnome-desktop/commit/e19df898281f83aa6db3d8fe57c720b29d98abee
2015-07-24rr: fix crash when XRRSetCrtcConfig failsMonsta1-11/+12
from https://github.com/GNOME/gnome-desktop/commit/4a95c179de8920ee8f78ae5e1f05df173e9d5bd8
2015-07-24bg slideshow: fix crash when system date is earlier than start dateMonsta1-1/+1
from https://github.com/GNOME/gnome-desktop/commit/d7cc5dadc3831b149bab1cd3e4af50e441546c2e
2015-07-21bg slideshow: smarter check for NULL pointersMonsta1-1/+15
2015-07-20Revert "changed checking for NULL a bit"raveit651-6/+1
This reverts commit 48df7a0c71a2df4050abae40ec9ca2b35dee7988. The commit causes a regression with timing xml background. see https://github.com/mate-desktop/mate-control-center/issues/159
2015-07-03backport cinnamon-desktop fix for buggy image typesleigh123linux1-32/+54
2015-06-27Remove duplicated typedefOBATA Akio2-3/+0
Such typedefs are in corresponding header files.
2015-04-29color select dialog: drop CSD/headerbar usageMonsta2-21/+0
maybe it was relevant with GTK+ 3.12 where the developers forced CSD on all dialogs, but now it isn't. in GTK+ 3.14 they apparently realized they were wrong and stopped forcing CSD on dialogs: http://blogs.gnome.org/mclasen/2014/07/28/a-talk-in-9-images/
2015-04-05Correct FSF adressposophe1-3/+3
2015-04-05Fix missing return in non-void functionsposophe1-0/+1
2015-04-05mate-desktop-item: add a child watch if "do not reap child" flag setMonsta1-1/+16
similar to https://git.gnome.org/browse/gnome-panel/commit/?id=76acc5b avoids double forking with desktop files that have "exec pkexec ..." inside. Closes https://github.com/mate-desktop/mate-desktop/pull/166
2015-03-13mate-thumbnail: first connect to settings, then read them.Monsta1-3/+6
fixes the issue with GLib >= 2.43, https://git.gnome.org/browse/glib/commit/?id=8ff5668a458344da22d30491e3ce726d861b3619 Closes https://github.com/mate-desktop/mate-desktop/pull/162
2015-01-15Fix missing ifdef for introspectionmate-desktop-1.9.4Stefano Karapetsas1-0/+3
2014-12-23pnp.ids: Pull latest from gnome-desktopinfirit1-8/+227
2014-12-23removed some unused codeMonsta1-10/+5
2014-12-16Revert "Revert "Drop support for Glib < 2.36""infirit1-99/+0
This reverts commit a0a0a16fb72bb7b592d23ce29ca52c7d63c2e40b.
2014-12-16Revert "Drop support for Glib < 2.32"infirit1-0/+64
This reverts commit 82d352cb10ed140b618b2c00a3935a3e8c6d1ba0.
2014-12-16Drop support for Glib < 2.32infirit1-64/+0
2014-12-16Revert "Drop support for Glib < 2.36"infirit1-0/+99
This reverts commit 6a0c8a93e30243232d63487c09ccb15a93755265.
2014-12-09fix GI build - remove useless GConf referenceMonsta1-2/+2
2014-12-09fix GTK+3 build - use GDK_WINDOW_XID for inter-GTK compatibilitymonsta1-1/+1
2014-12-09fix GTK+3 build - add missing semicolonmonsta1-1/+1
2014-11-28Don't fall over if RANDR is missinginfirit1-1/+2
When RANDR is not present, the screen object never gets an info, and since screen_info_free asserts, we must not call it unconditionally in finalize. Based on gnome-desktop commit: aa93854f2ba6b0f72f1bdff9699b5724a72c89c9 From: Matthias Clasen <[email protected]> Gnome bug: tps://bugzilla.gnome.org/show_bug.cgi?id=649053
2014-11-28Avoid a possible crashinfirit1-1/+2
When the spice client goes to fullscreen it (temporarily) turns all xrandr outputs off (for some reason). This causes gnome-settings-daemon to crash (and respawn and reset the old screen size) due to a NULL deref in libgnome-desktop. Based on gnome-desktop commit: b8849c08adb1ba1b55d1a3511dd9eefd00504dc0 Patch by: Alex Larsson Gnome bug: https://bugzilla.gnome.org/show_bug.cgi?id=646714
2014-11-28Include Xlib.h so we can build when RANDR is not availableinfirit1-0/+2
The RANDR headers include the Xlib paraphernalia, but we don't include those headers when gnome-desktop is configured to not use RANDR. So, instead we include Xlib.h so that we'll have defined types for Display, Screen, etc. Based on gnome-desktop commit: 27fb2967f620cbc82da57b9b5eea7700218b6620 From: Federico Mena Quintero <[email protected]>
2014-11-28Fix warning and remove unused variableinfirit2-3/+1
2014-11-28Don't build some RANDR code if RANDR is not availableinfirit1-0/+4
Based on gnome-desktop commit: f89d8a582abbb79898a93d26a8c0f52f74b20bd2 From: Federico Mena Quintero <[email protected]> Gnome bug: https://bugzilla.gnome.org/show_bug.cgi?id=643940
2014-11-28introspection: Fix c-includes to actually include headersinfirit1-0/+1
Based on gnome-commit: 446e8ffcf16ccf909f45b293532f50996b0dd728 From: Benjamin Otte <[email protected]>
2014-11-28Properly get the workarea for positioning the labelsinfirit1-4/+39
Based on gnome-desktop commit: b4b3fbe614c245e5e2cc4ca431d0f5310eca6ed1 From: Matthias Clasen <[email protected]> Gnome bug: https://bugzilla.gnome.org/show_bug.cgi?id=641996
2014-11-28Fix refcount issues for ->configinfirit1-1/+1
If the GnomeRRLabeler object is going to be unref'ing the config object, then it better keeps its own reference to it, otherwise it could cause problems when destroyed and re-created. Based on gnome-desktop commit: 8ceca629c2db2c8da0c7d8e391a68367c6c705d9 From: Bastien Nocera <[email protected]>
2014-11-28Fix two problems that causes m-s-d to crash at startinfirit1-1/+2
We need to ref outputs when we are copying them in the parser, and we need to avoid confusing i and j, even if they both have dots. Based on gnome-desktop commit: 11880c8bfd7b3f1edde58b8e83e46ebd59a44311 From: Matthias Clasen <[email protected]>
2014-11-28MateRROutputInfo: rename boolean accessorsinfirit3-7/+7
Rename get_connected() to is_connected() and get_active() to is_active(), following the normal convention. Based on gnome-desktop commit: 46d48ab6ec256498d0dd0a5c97c0388381971cbf From: Giovanni Campagna <[email protected]>
2014-11-28Disallow null filenames passed to mate_rr_config_load_filename()infirit1-1/+10
Based on gnome-desktop commit: 7c2e51427fff29f1582eb5ed035f8c7637ea908f From: Federico Mena Quintero <[email protected]>
2014-11-28Make the signal be MateRRScreen::changed, not screen-changed, to match the ↵infirit1-1/+1
vmethod name Based on gnome-desktop commit: cce5c87297ff8c163d4a29e1fc4e7277cbb4705a From: Federico Mena Quintero <[email protected]>
2014-11-28Turn the RANDR version checks into macrosinfirit1-12/+15
This makes them less error-prone to use. Also, fixes the master check for RANDR 1.2 in gnome_rr_screen_initable_init(), which wouldn't handle RANDR versions above 1.x. Based on gnome-desktop commit: 4da0756ef2e42b8040d8043029057b08e85efcec From: Federico Mena Quintero <[email protected]>
2014-11-28Don't export private symbols in the shared libraryinfirit1-0/+1
Based on gnome-desktop commit: 81590ec66ee8507ae61b47d3d93a8685e970a629 From: Emilio Pozuelo Monfort <[email protected]>
2014-11-28Always use gdk_window_get_window_typeinfirit1-6/+2
2014-11-28Turn GnomeRRConfig and GnomeOutputInfo into GObjectsinfirit9-710/+1077
Based on gnome-desktop commits: 1d1f8ef64f404eccb29d384a75939701327e8e77 2c7a2e90fee98ff61e530098bc962a65dec670bc 7436f759bec00d87f597ef0b6945da8fa8969e85 From: Giovanni Campagna <[email protected]> Gnome Bug: https://bugzilla.gnome.org/show_bug.cgi?id=630913
2014-11-28rr: unset primary on displays that are offinfirit1-1/+3
Based on gnome-desktop commit: 4dd5a30269d71bbe1da884a849db2290df5544c9 From: William Jon McCann <[email protected]>
2014-11-28rr: add api for ensuring a primary display is set in configinfirit2-0/+65
Based on gnome-desktop commit: f632805c549053a548877785e3e624b1fe687f23 From: William Jon McCann <[email protected]>
2014-11-28Add support for GObjectIntrospectioninfirit8-30/+183
Turned all GnomeRR structures into boxed types, then added the needed annotations and the Makefile.am bits. Does not yet include API changes, but should bind (awfully) all of libgnome-desktop. Based on gnome commit: cad94246fb5be76482212407a380cd75f9e7b932 url: https://git.gnome.org/browse/gnome-desktop/commit/?id=cad94246fb5be76482212407a380cd75f9e7b932
2014-11-28Turn MateRRScreen into a GObjectinfirit4-132/+310
Rework MateRRScreen so that it is a full GObject, and all its data is moved to MateScreenPrivate. GObject's are more supported when it comes to introspection and bindings, in particular wrt constructors. Therefore, this is a necessary step for useful introspection generation. https://bugzilla.gnome.org/show_bug.cgi?id=630913 Based on gnome commit 4d2a27d55bcf9bb8de170ae6c6540724125aafa3 url: https://git.gnome.org/browse/gnome-desktop/commit/?id=4d2a27d55bcf9bb8de170ae6c6540724125aafa3
2014-11-28changed checking for NULL a bitMonsta1-1/+6
the code below clearly assumes that slide should never be NULL, so the check is now done via g_return_if_fail Closes https://github.com/mate-desktop/mate-desktop/pull/143
2014-11-28avoid dereferencing NULL pointerMonsta1-4/+2
Closes https://github.com/mate-desktop/mate-desktop/pull/145
2014-11-28don't leak memorymonsta1-2/+6
Closes https://github.com/mate-desktop/mate-desktop/pull/138
2014-11-28va_start should be paired with va_endmonsta2-0/+4
Closes https://github.com/mate-desktop/mate-desktop/pull/137
2014-11-07gtk3: rgba rr-labeler apiDenis Gorodnichev2-1/+62
2014-11-07gtk3: append missing includeDenis Gorodnichev1-0/+1
2014-11-07mate_desktop_gtk_style_get_light_color and mate_desktop_gtk_style_get_dark_colorDenis Gorodnichev2-0/+260
2014-07-12libmate-desktop: Remove some compatibility with glib < 2.32Stefano Karapetsas1-17/+0