Matthias Clasen [Tue, 28 Aug 2012 10:45:30 +0000 (06:45 -0400)]
Add an atomic compare-and-exchange operation for object data
This is useful when using object data in thread-safe libraries.
https://bugzilla.gnome.org/show_bug.cgi?id=682849
Matthias Clasen [Tue, 28 Aug 2012 10:44:00 +0000 (06:44 -0400)]
Add compare-and exchange for data lists
Also, make it possible to get a 'new ref' on a datalist member
in a race-free way.
This is useful when using object data in thread-safe libraries.
https://bugzilla.gnome.org/show_bug.cgi?id=682849
Rūdolfs Mazurs [Sun, 2 Sep 2012 13:02:50 +0000 (16:02 +0300)]
Updated Latvian translation
Aurimas Černius [Sun, 2 Sep 2012 12:39:16 +0000 (15:39 +0300)]
Updated Lithuanian translation
Paolo Borelli [Sun, 2 Sep 2012 10:25:50 +0000 (12:25 +0200)]
Fix gtk-doc for g_app_info_get_supported_types
Aleksander Morgado [Fri, 31 Aug 2012 09:41:53 +0000 (11:41 +0200)]
gdbus-codegen: avoid error when wrong interface is provided to --annotate
If the interface given cannot be matched, `iface_obj' was left uninitialized and
the iface_obj == None check would end up crashing:
Traceback (most recent call last):
File "/usr/bin/gdbus-codegen", line 41, in <module>
sys.exit(codegen_main.codegen_main())
File "/usr/lib64/gdbus-2.0/codegen/codegen_main.py", line 175, in codegen_main
apply_annotations(all_ifaces, opts.annotate)
File "/usr/lib64/gdbus-2.0/codegen/codegen_main.py", line 146, in apply_annotations
apply_annotation(iface_list, iface, None, None, None, None, key, value)
File "/usr/lib64/gdbus-2.0/codegen/codegen_main.py", line 64, in apply_annotation
if iface_obj == None:
UnboundLocalError: local variable 'iface_obj' referenced before assignment
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=683088
Matthias Clasen [Sat, 1 Sep 2012 03:22:16 +0000 (23:22 -0400)]
Fix shell tests
Kjartan Maraas [Sat, 1 Sep 2012 08:42:05 +0000 (10:42 +0200)]
Updated Norwegian bokmål translation
Nguyễn Thái Ngọc Duy [Sat, 1 Sep 2012 06:17:05 +0000 (13:17 +0700)]
Updated Vietnamese translation
Nguyễn Thái Ngọc Duy [Sat, 1 Sep 2012 06:16:36 +0000 (13:16 +0700)]
po/vi: import from Damned Lies
Fran Diéguez [Fri, 31 Aug 2012 19:00:52 +0000 (21:00 +0200)]
Updated Galician translations
Martin Pitt [Mon, 20 Aug 2012 04:02:31 +0000 (06:02 +0200)]
Bump max time in test_method_calls_on_proxy() test
On slower platforms, the overhead of the 240 D-BUS Sleep calls is larger than
the current maximum of 6 seconds. A run on a Panda board sometimes fails with
ERROR:/build/buildd/glib2.0-2.33.8/./gio/tests/gdbus-threading.c:409:test_method_calls_on_proxy:
assertion failed (elapsed_msec < 6000): (7365 < 6000)
Bump maximum time to 8 seconds to be more resilient to this.
https://bugzilla.gnome.org/show_bug.cgi?id=682222
Tom Tryfonidis [Fri, 31 Aug 2012 08:08:42 +0000 (11:08 +0300)]
Updated Greek translation
Piotr Drąg [Thu, 30 Aug 2012 15:59:33 +0000 (17:59 +0200)]
Updated Polish translation
William Jon McCann [Tue, 28 Aug 2012 23:22:01 +0000 (19:22 -0400)]
Add ability to get symbolic icon for content type
https://bugzilla.gnome.org/show_bug.cgi?id=682101
William Jon McCann [Tue, 28 Aug 2012 22:56:53 +0000 (18:56 -0400)]
Add symbolic icon support to gfileinfo
https://bugzilla.gnome.org/show_bug.cgi?id=682101
William Jon McCann [Tue, 28 Aug 2012 19:37:43 +0000 (15:37 -0400)]
Add symbolic icon support to drive, volume, and mount
We need symbolic icon support for display in Nautilus.
https://bugzilla.gnome.org/show_bug.cgi?id=682101
Rui Matos [Wed, 29 Aug 2012 17:18:45 +0000 (19:18 +0200)]
gdbus-tool: Check return value of strrchr()
Fixes a crash when invoking gdbus like:
$ gdbus emit --session -o / -s Foo
https://bugzilla.gnome.org/show_bug.cgi?id=682965
Chun-wei Fan [Thu, 30 Aug 2012 09:14:13 +0000 (17:14 +0800)]
Update gspawn-win*helper* Visual C++ projects
Add GLIB_COMPILATION to the preprocessor definitions to fix the build.
Dominique Leuenberger [Fri, 17 Aug 2012 09:10:41 +0000 (11:10 +0200)]
gshell: Fix parsing of comments in command lines.
Fixes bug 562907
Cosimo Cecchi [Wed, 29 Aug 2012 18:46:11 +0000 (14:46 -0400)]
appinfo: fix annotations for g_app_info_launch and g_app_info_launch_uris
The files/uris parameters for these functions can be NULL, so we need an
allow-none annotation here.
Colin Walters [Mon, 27 Aug 2012 22:30:06 +0000 (18:30 -0400)]
Switch all open() calls to use g_open()
Because it now handles EINTR. And we should do so. While most people
use Linux, which tries very hard to avoid propagating EINTR back up
into userspace, it can still happen.
https://bugzilla.gnome.org/show_bug.cgi?id=682833
Colin Walters [Mon, 27 Aug 2012 22:22:26 +0000 (18:22 -0400)]
xdgmime: Handle EINTR in open()
https://bugzilla.gnome.org/show_bug.cgi?id=682833
Cosimo Cecchi [Tue, 28 Aug 2012 17:15:29 +0000 (13:15 -0400)]
gio: don't quote quark names for G_DEFINE_QUARK
Cosimo Cecchi [Tue, 28 Aug 2012 17:15:56 +0000 (13:15 -0400)]
glib: don't quote quark names for G_DEFINE_QUARK
Cosimo Cecchi [Tue, 28 Aug 2012 17:14:35 +0000 (13:14 -0400)]
gquark: clarify G_DEFINE_QUARK docs about the use of quotes
Use the term "name" instead of "string" in the documentation, and add a
note explaining double quotes shouldn't be used for the quark name.
Matthias Clasen [Tue, 28 Aug 2012 09:52:24 +0000 (05:52 -0400)]
Add missing includes
Matthias Clasen [Tue, 28 Aug 2012 03:42:12 +0000 (23:42 -0400)]
Use G_DEFINE_QUARK for quarks in GIO
Matthias Clasen [Tue, 28 Aug 2012 03:37:41 +0000 (23:37 -0400)]
Fix a typo in a doc comment
Matthias Clasen [Tue, 28 Aug 2012 03:36:42 +0000 (23:36 -0400)]
Use G_DEFINE_QUARK for GLib's own quarks
This commit just deals with glib/.
gobject/ and gio/ will be handled in separate commits.
Matthias Clasen [Tue, 28 Aug 2012 03:34:30 +0000 (23:34 -0400)]
Add a G_DEFINE_QUARK macro
https://bugzilla.gnome.org/show_bug.cgi?id=627240
Matthias Clasen [Tue, 28 Aug 2012 03:16:25 +0000 (23:16 -0400)]
Put quarks in their own source file
Just because I'm too lazy to grep every single time for the other
source file in which the quark code is hiding.
Daiki Ueno [Fri, 24 Aug 2012 06:43:02 +0000 (15:43 +0900)]
gsettings-tool: make list-recursively really recurse
https://bugzilla.gnome.org/show_bug.cgi?id=682586
Matthias Clasen [Mon, 20 Aug 2012 03:50:09 +0000 (23:50 -0400)]
GFile: Cosmetic doc changes
Matthias Clasen [Sun, 19 Aug 2012 22:36:18 +0000 (18:36 -0400)]
Fix the GDBusProxy async test
Patch by Mike Gorse, bug
https://bugzilla.gnome.org/show_bug.cgi?id=674805
Matthias Clasen [Sun, 19 Aug 2012 22:35:56 +0000 (18:35 -0400)]
Make async permission tests work
Mike Gorse [Thu, 26 Apr 2012 21:06:34 +0000 (16:06 -0500)]
Fix the gdbus-proxy async test
The async test had several problems:
- It created a proxy and did not launch a main loop, meaning that its
callback would usually not get called, or, if it did get called, the
test harness would have taken down the connection already, causing an
assertion failure when the proxy had an error.
- It was dependent on the proxy test to set up the server and would fail
because some properties were modified by that test.
https://bugzilla.gnome.org/show_bug.cgi?id=674805
Colin Walters [Mon, 27 Aug 2012 18:37:21 +0000 (14:37 -0400)]
gstdio: Harden g_open() against EINTR
Noticed by code inspection, when auditing some of my code for EINTR
handling.
https://bugzilla.gnome.org/show_bug.cgi?id=682819
Colin Walters [Wed, 28 Mar 2012 17:53:37 +0000 (13:53 -0400)]
gvariant: Add tests for exact SHA256 checksums
I'd like to use GVariant as a data format in my userspace filesystem,
and having the actual bits be stable means I can reliably compute
cryptographic checksums.
This updated patch removes vardict checks, because Ryan wants the
flexibility to change them in the future.
https://bugzilla.gnome.org/show_bug.cgi?id=673012
Ryan Lortie [Mon, 27 Aug 2012 20:12:30 +0000 (16:12 -0400)]
GVariantType: gut g_variant_type_check()
This function is causing an insane amount of wasted time on some
real-world profiles and it's pretty useless since we already have
GVariantType (as a type different from a string) for the purpose of
static type safety.
Disable it for now. We can possibly turn this back on again if we solve
bug #544026.
https://bugzilla.gnome.org/show_bug.cgi?id=679835
Nilamdyuti Goswami [Mon, 27 Aug 2012 18:09:52 +0000 (23:39 +0530)]
Implemented FUEL entries to Assamese translation
Dan Winship [Thu, 23 Aug 2012 16:48:49 +0000 (12:48 -0400)]
glib/tests: fix leaks
https://bugzilla.gnome.org/show_bug.cgi?id=682560
Dan Winship [Thu, 23 Aug 2012 16:29:36 +0000 (12:29 -0400)]
gtestutils: add g_test_add_data_func_full()
Like g_test_add_data_func(), but includes a GDestroyNotify for the
test data.
https://bugzilla.gnome.org/show_bug.cgi?id=682560
Dan Winship [Thu, 23 Aug 2012 16:38:17 +0000 (12:38 -0400)]
gmain: remove unix signal watch if its GSourceFunc returns FALSE
g_unix_signal_watch_dispatch() was ignore the callback's return value.
Fix that.
https://bugzilla.gnome.org/show_bug.cgi?id=682560
Dan Winship [Thu, 23 Aug 2012 16:35:20 +0000 (12:35 -0400)]
gmain: don't leak child sources that are destroyed before their parents
A parent source holds refs on its children, so if the child source is
destroyed, we need to drop that ref. Fix, and reorganize to make this
all more obvious.
https://bugzilla.gnome.org/show_bug.cgi?id=682560
Dan Winship [Thu, 23 Aug 2012 16:33:43 +0000 (12:33 -0400)]
gmain: free source_lists when freeing GMainContext
If a context was freed with sources still attached, those sources
correctly got destroyed, but the corresponding GSourceList structs
were being leaked.
https://bugzilla.gnome.org/show_bug.cgi?id=682560
Dan Winship [Sun, 26 Aug 2012 18:40:13 +0000 (14:40 -0400)]
Update .gitignores
Colin Walters [Sat, 25 Aug 2012 21:13:43 +0000 (17:13 -0400)]
docs: Ensure CLEANFILES is set before we use +=
We need to be defensive about this in the case where gtk-doc.make is
empty.
William Jon McCann [Fri, 24 Aug 2012 17:32:25 +0000 (13:32 -0400)]
Don't crash if set_app_info is called before adding metadata
set_mime_type, set_is_private, add_group, set_groups, set_icon, etc
all added metadata before using it. If set_app_info was called before
any of those it would crash when trying to access the metadata.
OKANO Takayoshi [Thu, 23 Aug 2012 10:15:09 +0000 (19:15 +0900)]
[l10n] Update Japanese translation
A S Alam [Thu, 23 Aug 2012 02:27:51 +0000 (07:57 +0530)]
update Punjabi Translation
Piotr Drąg [Wed, 22 Aug 2012 20:36:15 +0000 (22:36 +0200)]
Updated Polish translation
Colin Walters [Wed, 22 Aug 2012 18:48:41 +0000 (14:48 -0400)]
Fix regression when TMPDIR/TMP are unset
We should just be returning /tmp as a default, not calling g_getenv
("/tmp") which makes no sense.
https://bugzilla.gnome.org/show_bug.cgi?id=672329
Kjartan Maraas [Wed, 22 Aug 2012 08:29:16 +0000 (10:29 +0200)]
Updated Norwegian bokmål translation
Nguyễn Thái Ngọc Duy [Wed, 22 Aug 2012 02:15:00 +0000 (09:15 +0700)]
Updated Vietnamese translation
Nguyễn Thái Ngọc Duy [Wed, 22 Aug 2012 01:51:44 +0000 (08:51 +0700)]
po/vi: imported from Damned Lies
Duarte Loreto [Tue, 21 Aug 2012 23:54:09 +0000 (00:54 +0100)]
Updated Portuguese translation
Ryan Lortie [Tue, 21 Aug 2012 20:32:17 +0000 (16:32 -0400)]
*bump*
thanks rico :)
Dieter Verfaillie [Tue, 21 Aug 2012 19:10:16 +0000 (21:10 +0200)]
win32: fix tests/sources.c
"make check" was failing on sys/resource.h not being available on
win32. Seeing that this test just spews some numbers on stdout
whithout really testing anything we can safely replace them with
similar enough numbers by relying on g_get_monotonic_time ().
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=682386
Stef Walter [Tue, 21 Aug 2012 11:26:27 +0000 (13:26 +0200)]
Revert "gtlscertificate: Add certificate-bytes and private-key-bytes props"
* In order to add contstruct properties to an abstract base
calls, and retain ABI stability, the base class must add a
default implementation of those properties.
* We cannot add a default implementation of certificate-bytes
or private-key-bytes since certificate and private-key properties
are writable on construct-only.
This reverts commit
541c985869fe9f2c0a858c0a91b4eb60f99d19f0.
https://bugzilla.gnome.org/show_bug.cgi?id=682081
Stef Walter [Fri, 17 Aug 2012 08:55:10 +0000 (10:55 +0200)]
gdbus: Fix double free and use after free of object path
* This occurs when the 'g-object-path' property is read
on a GDBusObjectSkeleton
https://bugzilla.gnome.org/show_bug.cgi?id=682075
Chao-Hsiung Liao [Tue, 21 Aug 2012 11:16:27 +0000 (19:16 +0800)]
Updated Traditional Chinese translation(Hong Kong and Taiwan)
Cosimo Cecchi [Mon, 20 Aug 2012 17:46:33 +0000 (19:46 +0200)]
mount-op: use gint64 instead of guint64 for time_left and bytes_left
Since those can be -1, they need to be signed.
https://bugzilla.gnome.org/show_bug.cgi?id=682284
Andika Triwidada [Tue, 21 Aug 2012 06:53:46 +0000 (13:53 +0700)]
Updated Indonesian translation
Ryan Lortie [Mon, 20 Aug 2012 22:32:25 +0000 (18:32 -0400)]
NEWS, versions, etc.
Ryan Lortie [Mon, 20 Aug 2012 22:32:46 +0000 (18:32 -0400)]
Revert the GMarkup attribute collect changes
We need to have some more discussion on this topic.
This reverts commits
86329ba44fc7662c0bad37955f0ec980a24be495 and
8d40389d15544bdc612989157f80380badce52f7.
https://bugzilla.gnome.org/show_bug.cgi?id=665634
Ryan Lortie [Mon, 20 Aug 2012 21:57:13 +0000 (17:57 -0400)]
gmarkup test: fix use of message expect API
Dan Winship [Mon, 20 Aug 2012 21:26:45 +0000 (17:26 -0400)]
tests/protocol: update the messages to reflect G_LOG_DOMAIN being set now
Ryan Lortie [Mon, 20 Aug 2012 20:25:09 +0000 (16:25 -0400)]
g_test_expect_message: add some g_return_if_fail
NULL domain is not supported and some might expect this to work, so make
sure they don't get the wrong idea...
Lars Uebernickel [Mon, 20 Aug 2012 12:32:14 +0000 (14:32 +0200)]
GMenuModel: add test for the new GMenuItem API
Ryan Lortie [Sat, 18 Aug 2012 18:24:46 +0000 (14:24 -0400)]
GMenuItem: add constructor to copy from model
Add g_menu_item_new_from_model() for constructing a GMenuItem that is a
copy of a menu item that's in a GMenuModel.
Ryan Lortie [Sat, 18 Aug 2012 18:20:59 +0000 (14:20 -0400)]
GMenuItem: add getter APIs
GMenuItem has been write-only up to this point. Add some APIs for
reading back values as well.
Ryan Lortie [Sat, 18 Aug 2012 18:14:21 +0000 (14:14 -0400)]
GMenuModel: remove a type safety bug
There was a /* XXX */ in the code here to do proper typechecking of the
GVariant in the menu model when using g_menu_model_get_item_attribute().
We have g_variant_check_format_string() now, so use it.
Ryan Lortie [Sat, 18 Aug 2012 18:12:55 +0000 (14:12 -0400)]
GVariant: add g_variant_check_format_string()
For some time now people have been asking for a way to check for type
compatibility between GVariant instances and format strings. There are
several APIs inside of GLib itself that would benefit from this.
This patch introduces a way to do that.
Aurimas Černius [Mon, 20 Aug 2012 20:10:25 +0000 (23:10 +0300)]
Updated Lithuanian translation
Dan Winship [Mon, 30 Jul 2012 20:38:30 +0000 (16:38 -0400)]
glib/tests: use g_test_expect_message()
Replace some tests that used to use g_test_trap_fork() with
g_test_expect_message() instead.
https://bugzilla.gnome.org/show_bug.cgi?id=679556
Dan Winship [Mon, 30 Jul 2012 20:05:08 +0000 (16:05 -0400)]
gtestutils: add g_text_expect_message()
Add g_test_expect_message() and g_test_assert_expected_messages(), to
allow tests of warnings, error messages, return-if-fails, etc.
https://bugzilla.gnome.org/show_bug.cgi?id=679556
Dan Winship [Mon, 30 Jul 2012 19:32:31 +0000 (15:32 -0400)]
g_logv: only expand the message once
Hoist the g_strdup_printf()'ing out of the loop, since the message is
the same for every handler that gets called.
https://bugzilla.gnome.org/show_bug.cgi?id=679556
Dan Winship [Wed, 1 Aug 2012 13:49:10 +0000 (09:49 -0400)]
glib/tests/testing, logging: don't use g_test_undefined()
We are verifying the behavior of the test/logging functions here; this
is not undefined behavior.
https://bugzilla.gnome.org/show_bug.cgi?id=679556
Marc-André Lureau [Tue, 3 Jul 2012 00:04:08 +0000 (02:04 +0200)]
win32: add pipe-io-cancel-test
Test that win32 streams can be cancelled.
It can even be tested with wine on Linux!
https://bugzilla.gnome.org/show_bug.cgi?id=679288
Marc-André Lureau [Tue, 7 Aug 2012 23:02:40 +0000 (01:02 +0200)]
win32: add pipe-io-concurrent
Implement test case suggested by Ryan Lortie on bug:
https://bugzilla.gnome.org/show_bug.cgi?id=679288
"There is a potential race here that's really unlikely to happen, but
here we go: We are trying to read from the same socket in two threads.
Some data comes. That causes the poll() in both threads (above) to
finish running. Then the cancellable is checked above. We now find
ourselves here. Only one thread will read the data. The other will
block on this function. Then the user may cancel the cancellable while
we are blocked here, but we will stay blocked...."
Marc-André Lureau [Mon, 20 Aug 2012 13:47:15 +0000 (15:47 +0200)]
win32: add pipe-io-overlap-test
Marc-André Lureau [Thu, 5 Jul 2012 22:46:32 +0000 (00:46 +0200)]
win32: make gio stream cancellable
v2:
- fix cancellation of concurrent readers
- replace g_assert() usage with g_warn_if_fail()
v3:
- fix indentation
- fix loop code to not leak (silly me)
https://bugzilla.gnome.org/show_bug.cgi?id=679288
Marc-André Lureau [Tue, 7 Aug 2012 23:08:15 +0000 (01:08 +0200)]
win32: handle ERROR_MORE_DATA
If a named pipe is being read in message mode and the next message is
longer than the nNumberOfBytesToRead parameter specifies, ReadFile
returns FALSE and GetLastError returns ERROR_MORE_DATA.
Since the API doesn't allow to return both a GError and the number of
bytes read so far, it makes more sense to return nread, and let the
client call GetLastError() himself to check if ERROR_MORE_DATA.
The current alternative loses the nread information.
https://bugzilla.gnome.org/show_bug.cgi?id=679288
Marc-André Lureau [Mon, 2 Jul 2012 19:45:41 +0000 (21:45 +0200)]
win32: use overlapped events for streams
Any file handle created with FLAG_OVERLAPPED must have
ReadFile()/WriteFile() called with an OVERLAPPED structure.
Failing to do so will give unspecified results, invalid read/write or
corruption.
Without FLAG_OVERLAPPED, it is not possible to read and write
concurrently, even with two seperate threads, created by 2 input and
output gio streams. Also, only with FLAG_OVERLAPPED may an IO
operation be asynchronous and thus be cancellable.
We may want to call ReOpenFile() to make sure the FLAG is set, but
this API is only available since Vista+.
According to MSDN doc, adding the OVERLAPPED argument for IO operation
on handles without FLAG_OVERLAPPED is allowed, and indeed the existing
test still passes.
v2:
- update GetLastError() after _g_win32_overlap_wait_result ()
- split the unrelated ERROR_MORE_DATA handling
https://bugzilla.gnome.org/show_bug.cgi?id=679288
Thomas Hindoe Paaboel Andersen [Mon, 20 Aug 2012 14:41:42 +0000 (16:41 +0200)]
gmem.c: array is only paritally filled by memcpy
The size of the local_data arrray is too large. It should not be
multiplied by the sizeof guint.
The memcpy of profile_data to local_data later will only fill a part of the
array.
Spotted with the PVS-Studio static analyzer
https://bugzilla.gnome.org/show_bug.cgi?id=681501
Daniel Mustieles [Mon, 20 Aug 2012 12:15:39 +0000 (14:15 +0200)]
Updated Spanish translation
Matthias Clasen [Sun, 19 Aug 2012 06:27:04 +0000 (02:27 -0400)]
Remove some more 'the the'
Matthias Clasen [Sun, 19 Aug 2012 06:26:42 +0000 (02:26 -0400)]
Improve test coverage for GSocketConnection
Matthias Clasen [Sun, 19 Aug 2012 06:26:05 +0000 (02:26 -0400)]
Improve test coverage for unix socket addresses
Matthias Clasen [Sun, 19 Aug 2012 06:25:37 +0000 (02:25 -0400)]
Improve test coverage for resources
Matthias Clasen [Sun, 19 Aug 2012 06:25:21 +0000 (02:25 -0400)]
Improve test coverage for filter streams
Matthias Clasen [Sun, 19 Aug 2012 06:25:02 +0000 (02:25 -0400)]
Improve test coverage for unix streams
Matthias Clasen [Sun, 19 Aug 2012 06:24:44 +0000 (02:24 -0400)]
Improve test coverage for pollable streams
Matthias Clasen [Sun, 19 Aug 2012 06:24:04 +0000 (02:24 -0400)]
Improve GSettings test coverage
Matthias Clasen [Sun, 19 Aug 2012 06:23:32 +0000 (02:23 -0400)]
Improve GAppInfo test coverage
Matthias Clasen [Sun, 19 Aug 2012 06:22:59 +0000 (02:22 -0400)]
Add some GInetAddress tests
Matthias Clasen [Sun, 19 Aug 2012 06:22:33 +0000 (02:22 -0400)]
Add some GPermission tests
Matthias Clasen [Sun, 19 Aug 2012 06:21:02 +0000 (02:21 -0400)]
Cosmetic change
gtk-doc doesn't need <!-- --> anymore to recognize plural
forms of links.
Matthias Clasen [Sun, 19 Aug 2012 06:19:27 +0000 (02:19 -0400)]
Make GPermission more forgiving
It is not great if calling g_permission_acquire on a simple
permission object just segfaults. This commit arranges for
this to return a G_IO_ERROR_NOT_SUPPORTED error.