commit bb41a89c42dc4b24dc0a302b406df78e7c273744
Author: Matthias Clasen <[email protected]>
Date:   Mon Jun 22 12:46:32 2015 -0400

   2.45.3

NEWS         | 2 ++
configure.ac | 2 +-
2 files changed, 3 insertions(+), 1 deletion(-)

commit 4e7d22e268a4e06beb1c09585a48288c31004da5
Author: Matthias Clasen <[email protected]>
Date:   Tue Jun 23 06:55:28 2015 -0400

   Fix distcheck

   The appmonitor test was sometimes leaving files behind, causing
   distcheck some heartburn.

gio/tests/appmonitor.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)

commit 3e29dadae46fe4d1e26ff91c30d9d3b84eb6d84e
Author: Chun-wei Fan <[email protected]>
Date:   Tue Jun 23 13:52:25 2015 +0800

   gio/ginetaddress.c: Fix Windows XP inet_pton() Emulation

   We need to be more careful when we try to assign values to gpointers,
   so
   that means we have to assign the value to the properly-dereference
   gpointer, so that the assigned value will be retained after the
   function
   returns.  This code will be dropped soon, but it is done for XP
   compatibility's sake for 2.44.

   Should fix the issue reported in bug 730352 comment #24.

gio/ginetaddress.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)

commit fd789f118741d89df348ba21ddc1d813da02fb76
Author: Tim-Philipp Müller <[email protected]>
Date:   Wed Jun 3 13:06:24 2015 +0100

   gsocket: avoid unnecessary select in _send_messages() and
   _receive_message()

   For performance reasons we should always try to send or
   receive our messages first and only wait for more space
   or data to become available if we get an EAGAIN (and
   are in blocking mode).

   https://bugzilla.gnome.org/show_bug.cgi?id=751122

gio/gsocket.c | 26 ++++++++++++++------------
1 file changed, 14 insertions(+), 12 deletions(-)

commit 9e85f60ec68eb0330e18254726facdc5d8983e35
Author: Paolo Borelli <[email protected]>
Date:   Thu Jun 18 09:36:12 2015 +0200

   socketclient: annotate the connection param of "event" as nullable

   When emitting the RESOLVING/RESOLVED events the connection param is
   set to NULL.

gio/gsocketclient.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

commit f5d4543139172f366afa0de36dd7e5eac08e2297
Author: Alexander Larsson <[email protected]>
Date:   Wed Jun 17 17:31:19 2015 +0200

   configure: test have_docbook_style != yes, not have_docbook_dtd

configure.ac | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

commit 4b02bfd6eee43aaaa0ad30955cc85558f402d90f
Author: Philip Withnall <[email protected]>
Date:   Wed Jun 17 09:25:49 2015 +0100

   gfile: Clarify that g_file_replace_contents() uses atomic renames

   It uses g_file_replace() internally, so is inherently safe.

   Though it might vomit .goutputstream-XXXXXX files all over the place
   occasionally.

gio/gfile.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)

commit 6e576503876759d18c94bbad3c20ebb83b82c631
Author: Matthias Clasen <[email protected]>
Date:   Tue Jun 16 18:38:27 2015 -0400

   key file: Clarify documentation around comments

   The documentation was not very clear about the handling
   of the '#' comment markers. State clearly how these are
   handled by the getter and the setter.

   https://bugzilla.gnome.org/show_bug.cgi?id=479730

glib/gkeyfile.c | 14 ++++++++++----
1 file changed, 10 insertions(+), 4 deletions(-)

commit 7a295f063fdef26ae87bfe3a942847c5ecd7f7c3
Author: Matthias Clasen <[email protected]>
Date:   Tue Jun 16 14:46:04 2015 -0400

   Updates

NEWS | 54 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 54 insertions(+)

commit 9f90ee5eeccd47f39c7a03dcd786b125a19c195d
Author: Michael Catanzaro <[email protected]>
Date:   Sat Jun 13 22:52:33 2015 -0500

   genmarshal: silence register storage class warnings

   Using the register keyword triggers warnings on noteworthy compilers
   (clang), since it's deprecated in C++ and at danger of being removed
   from the language. There is no reason to use it since it isn't 1980
   anymore.

   https://bugzilla.gnome.org/show_bug.cgi?id=750918

gobject/glib-genmarshal.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)

commit 0d6e200384c8c2fb172273dcfaa6a48bc357d4a3
Author: Cosimo Cecchi <[email protected]>
Date:   Thu Jun 11 15:56:25 2015 -0700

   gresource: fix a couple of typos in documentation

gio/gresource.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)

commit 1f0a11c59aa244a70766bdb4c3b06d3d79b25c7a
Author: Rico Tzschichholz <[email protected]>
Date:   Thu Jun 11 07:52:40 2015 +0200

   g_log_set_handler_full: Bump "Since" version accordingly

glib/gmessages.c | 2 +-
glib/gmessages.h | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)

commit 2471d9cf8697b07d4e86b6f143eda7b779be02a9
Author: Matthias Clasen <[email protected]>
Date:   Fri Nov 28 23:31:00 2014 -0500

   Add g_log_set_handler_full

   This is a bindable version of g_log_set_handler that takes
   a destroy notify for the user_data.

   https://bugzilla.gnome.org/show_bug.cgi?id=740516

docs/reference/glib/glib-sections.txt |  1 +
glib/gmessages.c                      | 38
++++++++++++++++++++++++++++++++---
glib/gmessages.h                      |  6 ++++++
3 files changed, 42 insertions(+), 3 deletions(-)

commit 1102e6f9ca89b54e4e1139cca80f97c7bfb72dc3
Author: Matthias Clasen <[email protected]>
Date:   Sun Mar 29 14:00:36 2015 -0400

   Allow property actions to invert booleans

   This can be handy when you want to change the sense of a toggle
   in the UI without rewriting the underlying logic. Currently, this
   is just exposed as a construct-only property. We may add a
   convenience wrapper or a special !property syntax for this later.

   https://bugzilla.gnome.org/show_bug.cgi?id=728489

gio/gpropertyaction.c | 45 ++++++++++++++++++++++++++++++++++++++++++++-
gio/tests/actions.c   | 31 +++++++++++++++++++++++++++++++
2 files changed, 75 insertions(+), 1 deletion(-)

commit fb1e5ff04b48b9bb93b99b127abf3ffe73a98e5c
Author: Matthias Clasen <[email protected]>
Date:   Tue Jun 9 19:17:58 2015 -0400

   Fix deprecation notice

   GSimpleAsyncResult has not been deprecated all that long.

gio/gsimpleasyncresult.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

commit b701c3c60824fca4c0056a7a46c627fe2977257d
Author: Simon McVittie <[email protected]>
Date:   Mon Apr 27 16:26:33 2015 +0100

   Regression test for falling back to autolaunch: and
   XDG_RUNTIME_DIR/bus

   Bug: https://bugzilla.gnome.org/show_bug.cgi?id=747941
   Signed-off-by: Simon McVittie <[email protected]>
   Reviewed-by: Philip Withnall <[email protected]>

gio/tests/Makefile.am            |   2 +
gio/tests/dbus-launch.c          |  77 ++++++++++++++++++
gio/tests/gdbus-unix-addresses.c | 170
+++++++++++++++++++++++++++++++++++++++
3 files changed, 249 insertions(+)

commit 32492c6ab0000c50564360c74acf069814d942d1
Author: Simon McVittie <[email protected]>
Date:   Wed Apr 15 17:57:29 2015 +0100

   GDBus: try XDG_RUNTIME_DIR/bus before resorting to dbus-launch

   This is the right thing to do for the "a session is a user-session"
   model implemented in dbus 1.9.14, which is described in
   <http://lists.freedesktop.org/archives/dbus/2015-January/016522.html>.

   It also resembles sd-bus' behaviour, although sd-bus will only try
   kdbus and XDG_RUNTIME_DIR/bus, and never runs dbus-launch.

   On systems following the more traditional "a session is a
   login-session"
   model, X_R_D/bus won't exist, so it is harmless to check for it before
   falling back to X11 autolaunching. Again, this matches the behaviour
   of current libdbus and sd-bus versions.

   Now that we do this, g_test_dbus_unset() needs to clear
   XDG_RUNTIME_DIR
   as well as everything else.

   Bug: https://bugzilla.gnome.org/show_bug.cgi?id=747941
   Signed-off-by: Simon McVittie <[email protected]>
   Reviewed-by: Philip Withnall <[email protected]>

gio/gdbusaddress.c | 67
++++++++++++++++++++++++++++++++++++++++++++++++++++++
gio/gtestdbus.c    |  2 ++
2 files changed, 69 insertions(+)

commit 0d3f56e31c9099653539fa2539e05b3ba336a441
Author: Simon McVittie <[email protected]>
Date:   Wed Apr 15 19:59:50 2015 +0100

   g_dbus_address_connect: specifically use dbus-launch for autolaunch:

   This only alters what happens if we specifically connect to
   "autolaunch:", for instance via
   "DBUS_SESSION_BUS_ADDRESS=autolaunch:".
   We will still potentially try other platform-specific things if
   DBUS_SESSION_BUS_ADDRESS is unset. There are currently no other
   platform-specific things, so there is no practical difference yet,
   but I'm about to add a more-preferred fallback path before autolaunch.

   This matches libdbus' behaviour and the D-Bus Specification, in which
   the autolaunch: transport specifically means X11 autolaunch
   (as implemented by "dbus-launch --autolaunch") on Unix, or a
   shared-memory-based protocol on Windows. Other platform-specific
   transports or default/fallback modes, including launchd on Mac OS X
   and XDG_RUNTIME_DIR/bus on Unix, are not part of "autolaunch:".

   It's rather unfortunate that the same name means two different
   platform-specific mechanisms, specific to different platforms -
   if they were added today I'd call them x11: and windows-shm: or
   something - but it's been like this since 2007 so it's too late now.

   Bug: https://bugzilla.gnome.org/show_bug.cgi?id=747941
   Signed-off-by: Simon McVittie <[email protected]>
   Reviewed-by: Philip Withnall <[email protected]>

gio/gdbusaddress.c | 38 +++++++++++++++++++++-----------------
1 file changed, 21 insertions(+), 17 deletions(-)

commit bf181a3ac78e824ca7e67ecfb2ba957e740594d7
Author: Simon McVittie <[email protected]>
Date:   Mon Apr 27 14:38:41 2015 +0100

   regex: if PCRE is 8.34 or later, disable auto-possessification for DFA

   Normally, recent PCRE behaves as if certain patterns were replaced
   by a more "possessive" pattern that gives the same answer for normal
   regex matching, but is more efficient. However, the modified pattern
   produces fewer results under DFA. If we want the full set of results
   we have to apply PCRE_NO_AUTO_POSSESS, and that's a compile-time flag.

   This currently only affects a system PCRE, but would also work
   fine for
   an internal PCRE 8.34 or later if the embedded copy is updated.

   Bug: https://bugzilla.gnome.org/show_bug.cgi?id=733325
   Reviewed-by: Christian Persch <[email protected]>

glib/gregex.c | 128
+++++++++++++++++++++++++++++++++++++++++++---------------
1 file changed, 95 insertions(+), 33 deletions(-)

commit f45ceb838dae8f96b9bf646998a26a954a6fbf3d
Author: Christophe Fergeau <[email protected]>
Date:   Wed Jun 3 11:30:43 2015 +0200

   gapplication: Make sure --help output is translated

   Currently, applications using g_application_add_main_option_entries()
   won't get translated entries in --help output. We need to call
   g_option_group_set_translation_domain() with a NULL domain to ensure
   that the
   default application gettext domain (ie the one passed to the
   textdomain() call) will be used for the main entries passed by the
   application.

   If we want to allow more flexibility on which gettext domain should be
   used for these entries, new API will be needed.

   https://bugzilla.gnome.org/show_bug.cgi?id=750322

gio/gapplication.c | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)

commit 6cd1f8b40ff66509eedfce9c3c46cbef799a7ffe
Author: Philip Withnall <[email protected]>
Date:   Fri Dec 19 20:43:06 2014 +0000

   gsettings: Document GSettings build system integration

   Add a new section to the main GSettings documentation which documents
   the best practices for integrating GSettings into an autoconf/automake
   build system using the GLIB_GSETTINGS macro.

   Some of this material was adapted from the migrating-gconf.xml guide.

   https://bugzilla.gnome.org/show_bug.cgi?id=741788

gio/gsettings.c | 54
++++++++++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 54 insertions(+)

commit 723961b749613aad05b6bd22d1f12479a0022bef
Author: Philip Withnall <[email protected]>
Date:   Fri Dec 19 20:42:21 2014 +0000

   gsettings: Expand documentation default value l10n

   Mention context, translation category, and the need for syntactic
   validity of the translated values.

   https://bugzilla.gnome.org/show_bug.cgi?id=741788

gio/gsettings.c | 30 +++++++++++++++++++++++++++---
1 file changed, 27 insertions(+), 3 deletions(-)

commit 73a71d6a435485714d8a96a156bb805d47438ac0
Author: Xavier Claessens <[email protected]>
Date:   Mon Jun 8 11:09:39 2015 -0400

   doc: Add missing GTlsDatabaseClass

   https://bugzilla.gnome.org/show_bug.cgi?id=750573

docs/reference/gio/gio-sections.txt |  1 +
gio/gtlsdatabase.c                  | 36
++++++++++++++++++++++++++++++++++++
2 files changed, 37 insertions(+)

commit 1405eeeed797470bed9d8062d4051bf5b011e42b
Author: Matthias Clasen <[email protected]>
Date:   Fri Jun 5 19:13:19 2015 -0400

   Add registry helpers to the docs

docs/reference/gio/gio-docs.xml     |  4 +++
docs/reference/gio/gio-sections.txt | 69
+++++++++++++++++++++++++++++++++++++
2 files changed, 73 insertions(+)

commit e5e6c25c88349bf9fe2456efeb3785655967ad71
Author: Руслан Ижбулатов <[email protected]>
Date:   Tue May 5 02:12:25 2015 +0000

   Bump W32 Registry API 'Since:' version

   https://bugzilla.gnome.org/show_bug.cgi?id=734888

gio/gwin32registrykey.c | 60 +++++++++++++++++++++---------------------
gio/gwin32registrykey.h | 70
++++++++++++++++++++++++-------------------------
2 files changed, 65 insertions(+), 65 deletions(-)

commit 6579c87bd2e9d61cac078165505e502f57b03623
Author: Руслан Ижбулатов <[email protected]>
Date:   Tue May 5 02:09:10 2015 +0000

   Make W32 registry API compatible with MSVC

   * Only check __OBJECT_ATTRIBUTES_DEFINED and __UNICODE_STRING_DEFINED
     on MinGW (MSVC doesn't have these)
   * MSVC: disable:4005 when including windows.h and ntstatus.h
   * Move NTAPI cconv into the parens with the NtQueryKeyFunc
   * Fix return values in some functions

   https://bugzilla.gnome.org/show_bug.cgi?id=734888

gio/gwin32registrykey.c | 31 +++++++++++++++++--------------
1 file changed, 17 insertions(+), 14 deletions(-)

commit 1ac5b92c2f16335d686268191e0605ae1f344b4a
Author: Руслан Ижбулатов <[email protected]>
Date:   Mon Aug 25 03:38:35 2014 +0000

   Add W32 Registry reading API to gio

   https://bugzilla.gnome.org/show_bug.cgi?id=734888

gio/Makefile.am         |    2 +
gio/gwin32registrykey.c | 2418
+++++++++++++++++++++++++++++++++++++++++++++++
gio/gwin32registrykey.h |  283 ++++++
3 files changed, 2703 insertions(+)

commit 2a71f187d731a71ef16f4368d49af68aa3fe158d
Author: Руслан Ижбулатов <[email protected]>
Date:   Sun May 17 11:42:55 2015 +0000

   Make GWin32AppInfo MSVC-compatible - use G_VA_COPY

   https://bugzilla.gnome.org/show_bug.cgi?id=666831

gio/gwin32appinfo.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

commit 4d800e4d86db6825dd3c508c83352b9a4cd24350
Author: Руслан Ижбулатов <[email protected]>
Date:   Fri Aug 29 08:53:35 2014 +0000

   GWin32AppInfo rewrite

   - On first call scan the registry, collect information about URI
   protocols,
     file extensions, applications and handlers, store that as a set of
     interconnected structures in several hash tables
   - Watch the registry keys, re-scan the registry when any one of
   them changes.

   https://bugzilla.gnome.org/show_bug.cgi?id=666831

gio/gwin32appinfo.c | 4987
+++++++++++++++++++++++++++++++++++++++++++++------
gio/gwin32appinfo.h |    4 +-
2 files changed, 4432 insertions(+), 559 deletions(-)

commit 5bc0bc2fdeb8a7c7bb56fabbd8886b56bf34c85b
Author: Mikhail Zabaluev <[email protected]>
Date:   Tue Oct 14 01:25:56 2014 +0300

   Added g_utf8_validate() to UTF-8 performance testing

   https://bugzilla.gnome.org/show_bug.cgi?id=738504

glib/tests/utf8-performance.c | 20 ++++++++++++++++++++
1 file changed, 20 insertions(+)

commit 007f6684c54a695a6c35315f12f9f306d5af83bb
Author: Matthias Clasen <[email protected]>
Date:   Fri Jun 5 15:27:59 2015 -0400

   Add index for new api

docs/reference/gio/gio-docs.xml | 4 ++++
1 file changed, 4 insertions(+)

commit 01e9302673630b20120505d92bbc245ba1bd027a
Author: Matthias Clasen <[email protected]>
Date:   Fri Jun 5 15:27:25 2015 -0400

   Add new API to docs

docs/reference/gio/gio-sections.txt | 1 +
1 file changed, 1 insertion(+)

commit cb7020af5e864eed81520e980d66ccd683697ed4
Author: Ryan Lortie <[email protected]>
Date:   Wed Nov 19 12:45:38 2014 -0500

   GSettings: deprecate g_settings_list_keys()

   This is now possible with g_settings_schema_list_keys().

   https://bugzilla.gnome.org/show_bug.cgi?id=740308

gio/gsettings.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

commit 6cf867fb2a5086d2c5c0ac6c71ac795008cd13ae
Author: Ryan Lortie <[email protected]>
Date:   Wed Nov 19 12:45:06 2014 -0500

   gsettings tests: use g_settings_schema_list_keys()

   Stop using g_settings_list_keys() because soon it will be deprecated.

   https://bugzilla.gnome.org/show_bug.cgi?id=740308

gio/tests/gsettings.c | 32 ++++++++++++++++++++++++++++++--
1 file changed, 30 insertions(+), 2 deletions(-)

commit bb8eea6148fab965969c74e80a936864fc65e671
Author: Ryan Lortie <[email protected]>
Date:   Wed Nov 19 12:42:10 2014 -0500

   gsettings tool: use schema for listing keys

   Use the newly added g_settings_schema_list_keys() API instead of
   g_settings_list_keys() in order to list keys.

   Doing this allows the 'list-keys' command to work without creating a
   GSettings object, which is more efficient.  It also means that
   we don't
   have to provide a (meaningless and ignored) path when listing keys on
   relocatable schemas.

   While we're at it, update the 'range' command not to require
   creation of
   a GSettings object, in a similar way.

   https://bugzilla.gnome.org/show_bug.cgi?id=740308

gio/gsettings-tool.c | 75
+++++++++++++++++++++++++++++++++++++++-------------
1 file changed, 57 insertions(+), 18 deletions(-)

commit 82fcfeb3b065d7cc4d53e80776f9b476df7c2137
Author: Ryan Lortie <[email protected]>
Date:   Wed Nov 19 12:40:22 2014 -0500

   GSettingsSchema: add g_settings_schema_list_keys()

   The list of keys in a GSettings object depends entirely on the schema,
   so it makes sense to expose this API there.

   Move the implementation out of gsettings.c and into gsettingsschema.c,
   replacing the earlier with a simple call to the new location.

   We don't do the same for children because the children can change.

   https://bugzilla.gnome.org/show_bug.cgi?id=740308

gio/gsettings.c       | 18 +-----------------
gio/gsettingsschema.c | 41 +++++++++++++++++++++++++++++++++++++++++
gio/gsettingsschema.h |  3 +++
3 files changed, 45 insertions(+), 17 deletions(-)

commit 36e093a31a9eb12021e7780b9e322c29763ffa58
Author: Patrick Griffis <[email protected]>
Date:   Wed Mar 25 15:51:29 2015 -0400

   Implement GNotification on OSX

   https://bugzilla.gnome.org/show_bug.cgi?id=747146

gio/Makefile.am                 |   5 +-
gio/gcocoanotificationbackend.c | 278
++++++++++++++++++++++++++++++++++++++++
gio/giomodule.c                 |   7 +
gio/gnotificationbackend.c      |   4 +-
4 files changed, 292 insertions(+), 2 deletions(-)

commit 89058e8a9b769ab223bc75739f5455dab18f7a3d
Author: Patrick Griffis <[email protected]>
Date:   Mon Apr 13 13:39:28 2015 -0400

   configure: Require OSX >= 10.9

   https://bugzilla.gnome.org/show_bug.cgi?id=747146

configure.ac | 9 +++++++++
1 file changed, 9 insertions(+)

commit 9e8f4d473604e291c0164965bf9bba5ac77cc2bf
Author: Ryan Lortie <[email protected]>
Date:   Thu Mar 26 16:35:06 2015 -0400

   kqueue: add a bit of extra paranoia on cancel

   Cancellation of GPollFileMonitor is now handled correctly (in
   the sense
   that no further signals will follow) but let's be extra paranoid and
   disconnect our handler anyway, for good measure.

   https://bugzilla.gnome.org/show_bug.cgi?id=739424

gio/kqueue/gkqueuefilemonitor.c | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)

commit a367921d445f290263b052ad3da55525c1a5deb3
Author: Ryan Lortie <[email protected]>
Date:   Thu Mar 26 16:31:48 2015 -0400

   gpollfilemonitor: send 'changes done' for creates

   The new rules of GFileMonitor says that users should expect to see a
   CHANGES_DONE_HINT following a CREATED as well as CHANGED.

   https://bugzilla.gnome.org/show_bug.cgi?id=739424

gio/gpollfilemonitor.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

commit 62e5ee5514f0838689bc7d2d2c1a43fa912eaf1a
Author: Ryan Lortie <[email protected]>
Date:   Thu Mar 26 16:29:19 2015 -0400

   gpollfilemonitor: don't emit after cancellation

   GPollFileMonitor emits CHANGES_DONE_HINT after CHANGED signals, but it
   doesn't check to ensure that the file monitor wasn't cancelled
   before it
   does that.

   If the original signal caused the monitor to be unreffed, cancelled
   and
   destroyed, we would still end up emitting an extra signal on it.

   Avoid that by checking first for cancellation.

   https://bugzilla.gnome.org/show_bug.cgi?id=739424

gio/gpollfilemonitor.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)

commit 16190d2dcd111b75b240baffe88653e5fa2822f6
Author: Mikhail Zabaluev <[email protected]>
Date:   Mon Oct 6 02:01:38 2014 +0300

   glib/genviron.c, GSubprocessLauncher: ain't no "filename encoding"

   Removed all mentions of GLib file name encoding referring to
   the environment strings. The env var content has no defined relation
   to GLib's notion of filename encoding, or any encoding whatsoever.
   It would be wrong to pass all UTF-8 strings through
   g_filename_from_utf8() in order to put them into the environment,
   for one thing.

   https://bugzilla.gnome.org/show_bug.cgi?id=738185

gio/gsubprocesslauncher.c | 22 +++++++++-------------
glib/genviron.c           | 19 ++++++++-----------
2 files changed, 17 insertions(+), 24 deletions(-)

commit b31a873fb3bd1b67af9f7cc88bf44f8b5a941cec
Author: Jan Safranek <[email protected]>
Date:   Tue Nov 4 11:54:00 2014 +0000

   GDBus: Add new call flag to allow interactive authorization

   DBus has recently introduced new message flag
   DBUS_HEADER_FLAG_ALLOW_INTERACTIVE_AUTHORIZATION, which tells that
   caller is willing to wait for unspecified amount of time for the call
   to return, as the service may perform interactive authorization (e.g.
   using polkit).

   https://bugzilla.gnome.org/show_bug.cgi?id=739616

gio/gdbusconnection.c |  8 +++++++-
gio/gioenums.h        | 11 +++++++++--
2 files changed, 16 insertions(+), 3 deletions(-)

commit 865ce79ce0e92326b247ef51e7e2c78aaad99889
Author: Ryan Lortie <[email protected]>
Date:   Thu May 21 12:32:20 2015 -0500

   GActionGroupExporter: flush queue on requests

   In order to maintain a logical stream of events, we need to make
   sure we
   flush and queued change notifications before responding to any
   requests
   for information from clients.

   If we don't do this, it's possible that we emit an 'add' event
   that was
   queued at the time of a 'DescribeAll' call _after_ the reply to that
   call (which already contained the description of the new action).

   In practice, this is not only logically incorrect, but it can
   also cause
   problems.  If a change to action 'state' or 'enabled' occurs after the
   DescribeAll but before the signal has been dispatched, it will be
   ignored because an 'add' signal is already pending.  When that add
   signal is sent, it will contain the correct data, but the receiver
   will
   ignore it because it already saw the action in the DescribeAll reply.

   https://bugzilla.gnome.org/show_bug.cgi?id=749693

gio/gactiongroupexporter.c | 13 +++++++++++++
1 file changed, 13 insertions(+)

commit eeae7950fca962e65a5e08cfceae372782f5cc7e
Author: Ryan Lortie <[email protected]>
Date:   Sun May 25 11:21:06 2014 +0200

   GActionGroup: fix an annotation

   .get_action_state_type() does not return a copy.

   We remove the annotation entirely because it is evident from the
   'const'
   on the return type.

   https://bugzilla.gnome.org/show_bug.cgi?id=730168

gio/gactiongroup.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)

commit 96df2727f4c120a9fec888a2840f60ac3c3edde9
Author: Matthias Clasen <[email protected]>
Date:   Fri Jun 5 12:30:15 2015 -0400

   win32: Make g_content_type_get_mime_type work for directories

   Now that we are using inode/directory for directories, handle
   this case in g_content_type_get_mime_type() as well.

gio/gcontenttype-win32.c | 3 +++
1 file changed, 3 insertions(+)

commit 5f0665cbbc1d8b2fafeeb1fba7569612832b7541
Author: Руслан Ижбулатов <[email protected]>
Date:   Thu Apr 30 23:09:30 2015 +0000

   W32: Special treatment for inode/directory mime/type

   This is a hack for GLocalFileInfo to correctly get icons for
   directories.
   Without this change content type for any W32 directory is NULL
   (because there's no registry entry for "inode/directory" by default,
   and in any way there's no file extension that means "directory"
   to put there),
   and GLocalFileInfo uses content type to grab icons.

   https://bugzilla.gnome.org/show_bug.cgi?id=748727

gio/gcontenttype-win32.c | 4 ++++
1 file changed, 4 insertions(+)

commit 9931336d2dfbaf44aeb54b7e31164b91718e7769
Author: Matthias Clasen <[email protected]>
Date:   Fri Jun 5 12:26:41 2015 -0400

   win32: Return proper icon names

   The code here was returning gtk-directory and similar names as
   fallback, with a comment claiming that these are 'builtin gtk'.
   But they aren't, anymore, so just return the standard names.

gio/gcontenttype-win32.c | 19 +++++++++----------
1 file changed, 9 insertions(+), 10 deletions(-)

commit c20f3b239cd37733df1b68f113fdc17621cc3157
Author: Matthias Clasen <[email protected]>
Date:   Fri Jun 5 11:30:17 2015 -0400

   Fix a markup confusion

   "0." at the beginning of a line is interpreted as a numeric list
   by the gtk-doc markdown parser, so be careful to avoid that.

   https://bugzilla.gnome.org/show_bug.cgi?id=750399

glib/gdate.c | 10 +++++-----
1 file changed, 5 insertions(+), 5 deletions(-)

commit 07b3595c230c42ef8a31406d00477c9f755e902d
Author: Matthias Clasen <[email protected]>
Date:   Fri Jun 5 11:21:57 2015 -0400

   Trivial: fix a typo

   Pointed out in https://bugzilla.gnome.org/show_bug.cgi?id=750399

glib/gerror.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

commit c94e4c6f03e39e3e7749980b8d19f1644aacb003
Author: Philip Withnall <[email protected]>
Date:   Fri Dec 19 21:03:46 2014 +0000

   gsettings: Add a documentation section on relocatable schemas

   https://bugzilla.gnome.org/show_bug.cgi?id=741788

gio/gsettings.c | 26 +++++++++++++++++++++++++-
1 file changed, 25 insertions(+), 1 deletion(-)

commit 338741fff5381d1a8d11b8f62c9e208af8b016fa
Author: Stefan Ekenberg <[email protected]>
Date:   Wed Jun 3 15:59:57 2015 +0200

   Prevent race condition in g_io_condition_get_type

   Prevents race condition in function g_io_condition_get_type by
   ensuring
   that the initialization section for 'etype' is executed only once
   during a program's life time, and that concurrent threads are blocked
   until initialization completes. This changes solves the problem that
   concurrent threads could execute the check 'etype == 0' before any of
   them had initialized it, which in turn meant that multiple threads
   would then attempt to register the "GIOCondition" type.

   https://bugzilla.gnome.org/show_bug.cgi?id=750386

gobject/gsourceclosure.c | 8 +++++---
1 file changed, 5 insertions(+), 3 deletions(-)

commit 6c43b6a21aca5cac33992c6b690e566ac1664997
Author: Matthias Clasen <[email protected]>
Date:   Thu Jun 4 19:25:20 2015 -0400

   Trivial: fix a comment typo

gio/gunixmounts.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

commit cac0a49ca031bbf0d3fa9902010b05e2ad6523b3
Author: Balázs Úr <[email protected]>
Date:   Thu Jun 4 21:03:10 2015 +0000

   Updated Hungarian translation

po/hu.po | 502
+++++++++++++++++++++++++++++++++------------------------------
1 file changed, 261 insertions(+), 241 deletions(-)

commit bdc3f149ec134081943fcd626a1c826cccd05171
Author: Garrett Regier <[email protected]>
Date:   Wed Jun 3 17:16:21 2015 -0700

   binding: Simplify the default transform func

   https://bugzilla.gnome.org/show_bug.cgi?id=750369

gobject/gbinding.c | 8 +++-----
1 file changed, 3 insertions(+), 5 deletions(-)

commit ace7f6861e180d8a9a3b6982e2cfa522cb0cb5a0
Author: Garrett Regier <[email protected]>
Date:   Wed Jun 3 17:15:17 2015 -0700

   binding: Remove conditional from the default transform function

   Avoiding checking for INVERT_BOOLEAN each and
   instead choose the correct function in constructed().

   https://bugzilla.gnome.org/show_bug.cgi?id=750369

gobject/gbinding.c | 49 +++++++++++++++++--------------------------------
1 file changed, 17 insertions(+), 32 deletions(-)

commit 36593a3aba8dded2781a2bdd9b2ad0050dc939f2
Author: Garrett Regier <[email protected]>
Date:   Wed Jun 3 17:13:53 2015 -0700

   binding: Remove GObject data usage

   It isn't actually doing anything, instead it is
   being managed without actually being used.
   This has the result that GBinding is now more
   thread-safe.

   https://bugzilla.gnome.org/show_bug.cgi?id=745013

gobject/gbinding.c | 46 ++--------------------------------------------
1 file changed, 2 insertions(+), 44 deletions(-)

commit f68582356dd94f34e85574852556ab591854da27
Author: Piotr Drąg <[email protected]>
Date:   Wed Jun 3 19:02:12 2015 +0200

   Updated POTFILES.in

po/POTFILES.in | 1 +
1 file changed, 1 insertion(+)

commit 0d8dd2cf5cb7ba2d75b71bf765b299727eb32e26
Author: Xavier Claessens <[email protected]>
Date:   Wed Jun 3 10:18:36 2015 -0400

   doc: add GTlsInteractionClass

   https://bugzilla.gnome.org/show_bug.cgi?id=750344

docs/reference/gio/gio-sections.txt | 2 +-
gio/gtlsinteraction.c               | 9 +++++++++
2 files changed, 10 insertions(+), 1 deletion(-)

commit 7cba800a84730c9c5843acdd775e42b8c1438edf
Author: Alexander Larsson <[email protected]>
Date:   Mon Jun 1 10:02:47 2015 +0200

   GNetworkMonitorNetlink: Fix check for non-kernel messages

   This code used to look at the SCM_CREDENTIALS and ignore every message
   not from uid 0. However, when user namespaces are in use this does not
   work, as if uid 0 is not mapped you get overflowuid instead. Right now
   this means we ignore all messages in such user namespaces and glib
   apps hang on startup.

   We can't look at pids either, as pid 0 is returned for processes
   outside your pid namespace.

   Instead the correct approach is to look at the sending sockaddr and
   if the port id (nl_pid) is zero, then its from the kernel.

   Source:
   http://lists.linuxfoundation.org/pipermail/containers/2015-May/036032.html

   https://bugzilla.gnome.org/show_bug.cgi?id=750203

gio/gnetworkmonitornetlink.c | 30 +++++++++++++++---------------
1 file changed, 15 insertions(+), 15 deletions(-)

commit f8273f39a1fa5e961c30e96fa2a82d728736be09
Author: Alexander Larsson <[email protected]>
Date:   Mon Jun 1 10:01:26 2015 +0200

   Add GNativeSocketAddress for handling "other" addresses

   Instead of just dropping address types that we're not specifically
   handling we return a GNativeSocketAddress which is just a dummy
   container for the stuct sockaddr.

   https://bugzilla.gnome.org/show_bug.cgi?id=750203

docs/reference/gio/gio-sections.txt |  20 +++++
gio/Makefile.am                     |   2 +
gio/giotypes.h                      |   1 +
gio/gnativesocketaddress.c          | 160
++++++++++++++++++++++++++++++++++++
gio/gnativesocketaddress.h          |  65 +++++++++++++++
gio/gsocketaddress.c                |   3 +-
6 files changed, 250 insertions(+), 1 deletion(-)

commit f7b13e05f9bc5bd2b54f589d16ad580f6d833173
Author: Mattias Ellert <[email protected]>
Date:   Fri Oct 24 12:29:00 2014 +0000

   thread: Check sysconf value before using it

   sysconf() is documented as returning -1 if it can't determine
   a minimum thread stack size. Check for this case before using
   the return value.

   https://bugzilla.gnome.org/show_bug.cgi?id=739122

glib/gthread-posix.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)

commit 916297be799ee001b4a214cc52c3b960bb0b5deb
Author: Matthias Clasen <[email protected]>
Date:   Sat Sep 20 01:08:32 2014 -0400

   Add a global signal handler table

   Add a global lookup table for signal handlers. We already give
   them a unique ID, so there is no good reason to pay for
   non-constant lookups when disconnecting handlers.

   https://bugzilla.gnome.org/show_bug.cgi?id=737009

gobject/gsignal.c | 50 ++++++++++++++++++++++++++++++++++++++++----------
1 file changed, 40 insertions(+), 10 deletions(-)

commit 8a97dc56528b45dcd84a8c46f94b9a25a26d5cc3
Author: Matthias Clasen <[email protected]>
Date:   Sat Sep 20 00:34:24 2014 -0400

   Add a performance test for signal connection

   This test checks the performance of connecting, disconnecting and
   blocking many handlers. Various cases are checked: disconnect in
   the same order, in the inverse order, at random. Connect to one
   signal on a single object, to two signals on the same object, or
   to the same signal on two different objects.

   https://bugzilla.gnome.org/show_bug.cgi?id=737009

gobject/tests/Makefile.am      |   1 +
gobject/tests/signal-handler.c | 299
+++++++++++++++++++++++++++++++++++++++++
2 files changed, 300 insertions(+)

commit ff1381e31e2c05d0e410c49e6a728ad175a9fc0a
Author: Daniel Mustieles <[email protected]>
Date:   Tue May 26 19:25:56 2015 +0200

   Updated Spanish translation

po/es.po | 504
+++++++++++++++++++++++++++++++++------------------------------
1 file changed, 262 insertions(+), 242 deletions(-)

You are viewing proxied material from ftp.icm.edu.pl. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.