platform/upstream/glib.git
13 years agoImprove option context test coverage
Matthias Clasen [Fri, 30 Jul 2010 19:35:47 +0000 (15:35 -0400)]
Improve option context test coverage

13 years agoImprove array test coverage
Matthias Clasen [Fri, 30 Jul 2010 03:08:07 +0000 (23:08 -0400)]
Improve array test coverage

13 years agoMove markup parse tests to the test framework
Matthias Clasen [Fri, 30 Jul 2010 00:52:31 +0000 (20:52 -0400)]
Move markup parse tests to the test framework

13 years agoImprove coverage of buffered output stream tests
Matthias Clasen [Thu, 29 Jul 2010 06:20:43 +0000 (02:20 -0400)]
Improve coverage of buffered output stream tests

13 years agoImprove coverage of credentials tests
Matthias Clasen [Thu, 29 Jul 2010 06:20:29 +0000 (02:20 -0400)]
Improve coverage of credentials tests

13 years agoImprove coverage of converter stream tests
Matthias Clasen [Thu, 29 Jul 2010 06:20:15 +0000 (02:20 -0400)]
Improve coverage of converter stream tests

13 years agoImprove coverage of memory stream tests
Matthias Clasen [Thu, 29 Jul 2010 06:19:46 +0000 (02:19 -0400)]
Improve coverage of memory stream tests

13 years agoRemove redundant include
Matthias Clasen [Thu, 29 Jul 2010 06:19:27 +0000 (02:19 -0400)]
Remove redundant include

13 years agoImprove coverage of icon tests
Matthias Clasen [Thu, 29 Jul 2010 06:19:01 +0000 (02:19 -0400)]
Improve coverage of icon tests

13 years agoImprove coverage of application tests
Matthias Clasen [Thu, 29 Jul 2010 06:18:45 +0000 (02:18 -0400)]
Improve coverage of application tests

13 years agoImprove coverage of stream tests
Matthias Clasen [Thu, 29 Jul 2010 04:54:23 +0000 (00:54 -0400)]
Improve coverage of stream tests

13 years agoReenable unix-streams test
Matthias Clasen [Thu, 29 Jul 2010 04:53:30 +0000 (00:53 -0400)]
Reenable unix-streams test

13 years agoWork around deadlock in unix-streams test
Matthias Clasen [Thu, 29 Jul 2010 04:52:34 +0000 (00:52 -0400)]
Work around deadlock in unix-streams test

13 years agoAdd tests about file associations
Matthias Clasen [Wed, 28 Jul 2010 20:59:11 +0000 (16:59 -0400)]
Add tests about file associations

13 years agoImprove regex test coverage
Matthias Clasen [Wed, 28 Jul 2010 20:05:19 +0000 (16:05 -0400)]
Improve regex test coverage

13 years agoRemove duplicate utf8 strreverse test
Matthias Clasen [Tue, 27 Jul 2010 22:21:15 +0000 (18:21 -0400)]
Remove duplicate utf8 strreverse test

Also add some tests for find next/prev char.

13 years agoImprove strfuncs test coverage a bit
Matthias Clasen [Tue, 27 Jul 2010 20:52:36 +0000 (16:52 -0400)]
Improve strfuncs test coverage a bit

13 years agoImprove option context test coverage a bit
Matthias Clasen [Tue, 27 Jul 2010 20:25:13 +0000 (16:25 -0400)]
Improve option context test coverage a bit

13 years agoMove markup collect tests to the test framework
Matthias Clasen [Tue, 27 Jul 2010 20:24:56 +0000 (16:24 -0400)]
Move markup collect tests to the test framework

13 years agoAdd some more GFile tests
Matthias Clasen [Tue, 27 Jul 2010 20:24:08 +0000 (16:24 -0400)]
Add some more GFile tests

13 years agoImprove compressor test coverage a bit
Matthias Clasen [Tue, 27 Jul 2010 18:01:48 +0000 (14:01 -0400)]
Improve compressor test coverage a bit

13 years agoTest g_settings_list_schemas
Matthias Clasen [Tue, 27 Jul 2010 05:00:55 +0000 (01:00 -0400)]
Test g_settings_list_schemas

13 years agoImprove GNode test coverage a bit
Matthias Clasen [Tue, 27 Jul 2010 04:17:20 +0000 (00:17 -0400)]
Improve GNode test coverage a bit

13 years agoAdd some logging tests
Matthias Clasen [Tue, 27 Jul 2010 03:24:31 +0000 (23:24 -0400)]
Add some logging tests

13 years agoSilence log handler tests
Matthias Clasen [Tue, 27 Jul 2010 03:06:33 +0000 (23:06 -0400)]
Silence log handler tests

13 years agoImprove tree test coverage
Matthias Clasen [Tue, 27 Jul 2010 02:13:25 +0000 (22:13 -0400)]
Improve tree test coverage

13 years agoMove markup escape tests to test framework
Matthias Clasen [Tue, 27 Jul 2010 02:12:59 +0000 (22:12 -0400)]
Move markup escape tests to test framework

13 years agoMore pattern tests
Matthias Clasen [Tue, 27 Jul 2010 01:19:04 +0000 (21:19 -0400)]
More pattern tests

13 years agoAdd tests for g_utf8_strreverse
Matthias Clasen [Tue, 27 Jul 2010 01:18:22 +0000 (21:18 -0400)]
Add tests for g_utf8_strreverse

13 years agoFix a warning
Matthias Clasen [Tue, 27 Jul 2010 01:17:56 +0000 (21:17 -0400)]
Fix a warning

13 years agoDon't leak in converter test
Matthias Clasen [Tue, 27 Jul 2010 01:16:03 +0000 (21:16 -0400)]
Don't leak in converter test

13 years agoFix an icon test
Matthias Clasen [Tue, 27 Jul 2010 00:22:33 +0000 (20:22 -0400)]
Fix an icon test

13 years agoAdd tests for GNetworkAddress
Matthias Clasen [Mon, 26 Jul 2010 23:07:16 +0000 (19:07 -0400)]
Add tests for GNetworkAddress

13 years agoDrop an unneeded if
Matthias Clasen [Mon, 26 Jul 2010 22:40:17 +0000 (18:40 -0400)]
Drop an unneeded if

13 years agoRemove some unused pcre code
Matthias Clasen [Mon, 26 Jul 2010 22:15:24 +0000 (18:15 -0400)]
Remove some unused pcre code

13 years agoAdd some volume monitor tests
Matthias Clasen [Mon, 26 Jul 2010 21:59:18 +0000 (17:59 -0400)]
Add some volume monitor tests

Although not much of this is easily testable.

13 years agoMove queue tests to the test framework
Matthias Clasen [Sun, 25 Jul 2010 23:12:51 +0000 (19:12 -0400)]
Move queue tests to the test framework

13 years agoBug 625628 – GDBusProxy: wrong property name sorting
David Zeuthen [Fri, 30 Jul 2010 20:33:29 +0000 (16:33 -0400)]
Bug 625628 – GDBusProxy: wrong property name sorting

https://bugzilla.gnome.org/show_bug.cgi?id=625628

Signed-off-by: David Zeuthen <davidz@redhat.com>
13 years agoFix gschema-compile test
David Zeuthen [Fri, 30 Jul 2010 20:30:34 +0000 (16:30 -0400)]
Fix gschema-compile test

This was broken by commit 1b5b87bfa0f5435dd6e973a4e36076db083dc5c7

Signed-off-by: David Zeuthen <davidz@redhat.com>
13 years agoBug 625584 – Crashes application on unref with signal subscription
David Zeuthen [Fri, 30 Jul 2010 20:01:03 +0000 (16:01 -0400)]
Bug 625584 – Crashes application on unref with signal subscription

Don't do too much work in the finalizer - in particular, there's no
need to send RemoveMatch() messages to the bus daemon since we're
going to sever the connection and the bus will garbage collect
anyway. In this case it crashed the process.

Also add a test case that checks that the appropriate GDestroyNotify
callbacks are called when unreffing a connection with either 1)
exported objects; 2) signal subscriptions or 3) filter functions
.. yes, ideally apps would unregister such callbacks before giving up
their ref but that's not how things work :-)

Signed-off-by: David Zeuthen <davidz@redhat.com>
13 years agoGSettings: Mark some strings for cat translation
Ryan Lortie [Fri, 30 Jul 2010 19:10:53 +0000 (21:10 +0200)]
GSettings: Mark some strings for cat translation

<vuntz> *miaou*

13 years agoGDBusConnection: Fix up debug output when completing calls
David Zeuthen [Fri, 30 Jul 2010 15:40:00 +0000 (11:40 -0400)]
GDBusConnection: Fix up debug output when completing calls

Signed-off-by: David Zeuthen <davidz@redhat.com>
13 years agoGCredentials: use effective uid/gid
David Zeuthen [Fri, 30 Jul 2010 15:26:43 +0000 (11:26 -0400)]
GCredentials: use effective uid/gid

Otherwise e.g. setuid root processes can't connect to the system
bus. This was discovered when porting PolicyKit's pkexec(1) command to
a PolicyKit library using GDBus.

Signed-off-by: David Zeuthen <davidz@redhat.com>
13 years ago*bump*
Ryan Lortie [Thu, 29 Jul 2010 20:04:43 +0000 (22:04 +0200)]
*bump*

13 years agoadd one more NEWS item 2.25.12
Ryan Lortie [Thu, 29 Jul 2010 14:52:34 +0000 (16:52 +0200)]
add one more NEWS item

13 years agoGIO: Add G-I annotations for several methods
Eduardo Lima Mitev [Mon, 26 Jul 2010 15:21:14 +0000 (17:21 +0200)]
GIO: Add G-I annotations for several methods

Specifically:
 - g_io_stream_get_input/output_stream()
 - g_socket_client_get_local_address()
 - g_socket_connection_get_socket()
 - g_socket_listener_add_address()
 - g_socket_listener_accept_socket()
 - g_socket_listener_accept()
 - g_io_extension_point_get_extensions()

13 years agoget a start on the NEWS
Ryan Lortie [Thu, 29 Jul 2010 10:01:46 +0000 (12:01 +0200)]
get a start on the NEWS

13 years ago[gschema-compile] fix segfault
Owen Taylor [Wed, 28 Jul 2010 19:15:57 +0000 (15:15 -0400)]
[gschema-compile] fix segfault

If files are passed on the command line and we don't set override_files
at all, don't try to read from the array.

13 years agoUse correct type for GDBusServer:active property
David Zeuthen [Wed, 28 Jul 2010 16:58:04 +0000 (12:58 -0400)]
Use correct type for GDBusServer:active property

It's a boolean, not a string.

Signed-off-by: David Zeuthen <davidz@redhat.com>
13 years agoGDBusConnection: Pass the GObject to g_simple_async_result_new()
David Zeuthen [Wed, 28 Jul 2010 16:48:17 +0000 (12:48 -0400)]
GDBusConnection: Pass the GObject to g_simple_async_result_new()

Signed-off-by: David Zeuthen <davidz@redhat.com>
13 years agoAdd support for vendor override files
Ryan Lortie [Tue, 27 Jul 2010 20:56:30 +0000 (22:56 +0200)]
Add support for vendor override files

You can drop a key-file in the schema directory that looks like:

[org.gtk.Example]
key='value'

to override the default value of 'key' in schema 'org.gtk.Example'.

13 years agoUpdated Galician translations
Fran Diéguez [Tue, 27 Jul 2010 15:47:22 +0000 (17:47 +0200)]
Updated Galician translations

13 years agoUpdated Spanish translation
Jorge González [Mon, 26 Jul 2010 18:58:59 +0000 (20:58 +0200)]
Updated Spanish translation

13 years agoUpdated Hebrew translation.
Yaron Shahrabani [Sat, 24 Jul 2010 10:31:31 +0000 (13:31 +0300)]
Updated Hebrew translation.

13 years ago[gio]: gioenums.h: Remove trailing comma again.
Murray Cumming [Fri, 23 Jul 2010 11:24:43 +0000 (13:24 +0200)]
[gio]: gioenums.h: Remove trailing comma again.

13 years agoRevert "notify: Remove unused g_object_notify_queue_clear()"
Benjamin Otte [Fri, 23 Jul 2010 08:51:14 +0000 (10:51 +0200)]
Revert "notify: Remove unused g_object_notify_queue_clear()"

This reverts commit 0201a81f04ae5fd46c6db25859e90dde1fc35d4b and adds
locks in the right place.

As this file is a public header (see last commit), we cannot just remove
functions from it.

13 years agonotify: Readd g_object_notify_queue_from_object()
Benjamin Otte [Fri, 23 Jul 2010 08:49:26 +0000 (10:49 +0200)]
notify: Readd g_object_notify_queue_from_object()

gobjectnotifyqueue.c is a public header, so we cannot remove files from
it. (Yes you read that right, this .c file is a public header. See your
installation directory.)

13 years agoComplete Simplified Chinese translation.
Tao Wang [Fri, 23 Jul 2010 05:18:43 +0000 (13:18 +0800)]
Complete Simplified Chinese translation.

13 years agoUpdate Simplified Chinese translation.
Tao Wang [Fri, 23 Jul 2010 05:05:37 +0000 (13:05 +0800)]
Update Simplified Chinese translation.

13 years agoCleanup the GSettingsBackend API
Ryan Lortie [Thu, 22 Jul 2010 22:39:50 +0000 (18:39 -0400)]
Cleanup the GSettingsBackend API

This is a substantial ABI break.  Rebuild dconf.

13 years agoGSettings: add g_settings_reset()
Ryan Lortie [Thu, 22 Jul 2010 21:49:40 +0000 (17:49 -0400)]
GSettings: add g_settings_reset()

Resets a key to its default value.

13 years agonotify: Add tests for threadsafe object notifies
Olivier Crête [Thu, 22 Jul 2010 19:52:54 +0000 (21:52 +0200)]
notify: Add tests for threadsafe object notifies

https://bugzilla.gnome.org/show_bug.cgi?id=166020

13 years agonotify: Make dedup code clearer
Benjamin Otte [Thu, 22 Jul 2010 19:51:28 +0000 (21:51 +0200)]
notify: Make dedup code clearer

A for loops seems easier to understand to me than gotos.

13 years agonotify: Make GObject property change notifications thread safe
Olivier Crête [Thu, 22 Jul 2010 19:26:32 +0000 (21:26 +0200)]
notify: Make GObject property change notifications thread safe

Adds locking around object property change notification handling. The
notifications are only emitted after all threads have called
g_object_thaw_notify().

https://bugzilla.gnome.org/show_bug.cgi?id=166020

13 years agonotify: Remove g_object_notify_queue_from_object()
Olivier Crête [Thu, 22 Jul 2010 19:20:35 +0000 (21:20 +0200)]
notify: Remove g_object_notify_queue_from_object()

The function will not be safe with object locking in place, so we remove
it. The workaround is somewhat ugly, but it works.

https://bugzilla.gnome.org/show_bug.cgi?id=166020

13 years agonotify: Refactor g_object_notify_queue_thaw()
Olivier Crête [Thu, 22 Jul 2010 18:25:00 +0000 (20:25 +0200)]
notify: Refactor g_object_notify_queue_thaw()

This adds better error reporting and simplifies the code for adding
thread safety.

https://bugzilla.gnome.org/show_bug.cgi?id=166020

13 years agonotify: Remove unused g_object_notify_queue_clear()
Olivier Crête [Thu, 22 Jul 2010 07:02:18 +0000 (09:02 +0200)]
notify: Remove unused g_object_notify_queue_clear()

https://bugzilla.gnome.org/show_bug.cgi?id=166020

13 years agonotify: Emit a g_critical() instead g_return_if_fail()ing
Benjamin Otte [Thu, 22 Jul 2010 06:52:25 +0000 (08:52 +0200)]
notify: Emit a g_critical() instead g_return_if_fail()ing

This does not change the code semantically in any way but avoids a
return in the middle of the code.

https://bugzilla.gnome.org/show_bug.cgi?id=166020

13 years agobinding: Cast to avoid compiler warning
Benjamin Otte [Thu, 22 Jul 2010 18:24:29 +0000 (20:24 +0200)]
binding: Cast to avoid compiler warning

13 years agotests: Use GPOINTER_TO_SIZE() instead of just casting to guint
Benjamin Otte [Thu, 22 Jul 2010 18:23:28 +0000 (20:23 +0200)]
tests: Use GPOINTER_TO_SIZE() instead of just casting to guint

a) gcc doesn't like this ("Cast to pointer of different size)
b) It compares only half of the pointer

13 years agoinotify: Constify function argument
Benjamin Otte [Thu, 22 Jul 2010 18:23:04 +0000 (20:23 +0200)]
inotify: Constify function argument

gcc complained...

13 years agoFix very small leak in the GSettings test
Ryan Lortie [Thu, 22 Jul 2010 15:49:30 +0000 (11:49 -0400)]
Fix very small leak in the GSettings test

13 years agoBug 624991 - GSettings mapping for G_TYPE_STRV
Ryan Lortie [Thu, 22 Jul 2010 03:10:31 +0000 (20:10 -0700)]
Bug 624991 - GSettings mapping for G_TYPE_STRV

Added default mapping for G_TYPE_STRV based on a patch from Garrett
Regier.

Add a test case.

13 years agoUpdated Galician translations
Fran Diéguez [Thu, 22 Jul 2010 14:23:18 +0000 (16:23 +0200)]
Updated Galician translations

13 years agoBug 624968 - div by zero in g_malloc_n family
Behdad Esfahbod [Wed, 21 Jul 2010 20:08:19 +0000 (16:08 -0400)]
Bug 624968  - div by zero in g_malloc_n family

13 years agoUpdated Hebrew translation.
Yaron Shahrabani [Wed, 21 Jul 2010 16:11:53 +0000 (19:11 +0300)]
Updated Hebrew translation.

13 years agoGDBusAuthObserver: Fix docs
David Zeuthen [Tue, 20 Jul 2010 19:02:36 +0000 (15:02 -0400)]
GDBusAuthObserver: Fix docs

Signed-off-by: David Zeuthen <davidz@redhat.com>
13 years agoBug 617483 – Credentials passing
David Zeuthen [Tue, 20 Jul 2010 18:02:14 +0000 (14:02 -0400)]
Bug 617483 – Credentials passing

 - Make GCredentials instance and class structures private so it can't
   be subclassed and we don't have to worry about ABI compat
   issues. This also allows us to get rid of the GCredentialsPrivate
   struct.

 - Add a GCredentialsType enumeration that is used whenever exchanging
   pointers with the user. This allows us to support OSes with
   multiple native credential types. In particular, it allows
   supporting OSes where the native credential evolves or even changes
   over time.

 - Add g_socket_get_credentials() method.

 - Add tests for g_socket_get_credentials(). Right now this is in the
   GDBus peer-to-peer test case but we can change that later.

 - Move GTcpConnection into a separate gtk-doc page as was already
   half-done with GUnixConnection. Also finish the GUnixConnection
   move and ensure send_credentials() and receive_credentials()
   methods are in the docs. Also nuke comment about GTcpConnection
   being empty compared to its superclass.

Signed-off-by: David Zeuthen <davidz@redhat.com>
13 years agoGDBusServer: Avoid handling incoming connections after stop() has been called
David Zeuthen [Tue, 20 Jul 2010 16:14:30 +0000 (12:14 -0400)]
GDBusServer: Avoid handling incoming connections after stop() has been called

Without this fix, we'd sometimes run code after stop() and finalize()
to handle incoming requests. This was observed in the gdbus-peer test
case occasionally crashing:

 $ ./gdbus-peer
 /gdbus/peer-to-peer: OK
 /gdbus/delayed-message-processing: OK
 /gdbus/nonce-tcp:
 GLib-GObject-WARNING **: invalid uninstantiatable type `(null)' in cast to `GDBusServer'
 aborting...

Signed-off-by: David Zeuthen <davidz@redhat.com>
13 years agoGDBus: Add support for D-Bus type 'h' (ie. G_VARIANT_TYPE_HANDLE)
David Zeuthen [Tue, 20 Jul 2010 15:38:23 +0000 (11:38 -0400)]
GDBus: Add support for D-Bus type 'h' (ie. G_VARIANT_TYPE_HANDLE)

This allows sending and receiving D-Bus messages with instances of the
'h' D-Bus type. Unlike libdbus-1's dbus_message_iter_get_basic()
method, g_variant_get_handle() does not return a duplicated unix file
descriptor (that must be closed with close(2)) - instead, it returns
an index that can be used to get/dup the file descriptor from a
GUnixFDList object that can be obtained from the GDBusMessage object.

Signed-off-by: David Zeuthen <davidz@redhat.com>
13 years agoMove #include <sys/wait.h> down
Ryan Lortie [Tue, 20 Jul 2010 14:45:52 +0000 (10:45 -0400)]
Move #include <sys/wait.h> down

Until after we include the glib stuff, so that we have G_OS_UNIX
defined.

For some reason <stdlib.h> pulls in <sys/wait.h> on Fedora so this
wasn't a problem, but many others have reported the issue.

13 years agoUpdated Spanish translation
Jorge González [Tue, 20 Jul 2010 06:57:35 +0000 (08:57 +0200)]
Updated Spanish translation

13 years agoCorrected and updated the new version of the Armenian translation - po file.
noch [Tue, 20 Jul 2010 05:33:32 +0000 (10:33 +0500)]
Corrected and updated the new version of the Armenian translation - po file.

13 years agoGDBus: Introduce G_DBUS_SEND_MESSAGE_FLAGS_PRESERVE_SERIAL
David Zeuthen [Mon, 19 Jul 2010 21:03:07 +0000 (17:03 -0400)]
GDBus: Introduce G_DBUS_SEND_MESSAGE_FLAGS_PRESERVE_SERIAL

This is preferable to the current magical solution whereby the serial
is only rewritten if non-zero. In particular, it makes it easier to
send the same message on multiple connections without having to reset
the serial number.

Signed-off-by: David Zeuthen <davidz@redhat.com>
13 years agoGDBus: Forgot to add g_dbus_send_message_flags_get_type() to gio.symbols
David Zeuthen [Mon, 19 Jul 2010 20:10:50 +0000 (16:10 -0400)]
GDBus: Forgot to add g_dbus_send_message_flags_get_type() to gio.symbols

Signed-off-by: David Zeuthen <davidz@redhat.com>
13 years agoGDBus: Add GDBusSendMessageFlags
David Zeuthen [Mon, 19 Jul 2010 20:07:57 +0000 (16:07 -0400)]
GDBus: Add GDBusSendMessageFlags

This is currently unused but might be useful in the future. For
example, it might be nice with a way to bypass the current queue of
outgoing messages - having a flag enumeration allows us to add a
G_DBUS_SEND_MESSAGE_FLAGS_BYPASS_QUEUE etc. etc.

This commit breaks ABI and API. Users of the (rarely used) API to send
messages will have to port to this new API.

Signed-off-by: David Zeuthen <davidz@redhat.com>
13 years agoGDBus: Remove constness from introspection data structures
David Zeuthen [Mon, 19 Jul 2010 19:45:27 +0000 (15:45 -0400)]
GDBus: Remove constness from introspection data structures

Signed-off-by: David Zeuthen <davidz@redhat.com>
13 years agoGDBus: Add GDBusSignalFlags and use it in g_dbus_connection_signal_subscribe()
David Zeuthen [Mon, 19 Jul 2010 19:03:00 +0000 (15:03 -0400)]
GDBus: Add GDBusSignalFlags and use it in g_dbus_connection_signal_subscribe()

This is currently unused but will probably be useful in the
future. For example, we could have a _ARG0_IS_PATH to specify that
arg0 should be used for arg0path.

This commit breaks API and ABI. Users of
g_dbus_connection_signal_subscribe() will need to port to this new
version.

Signed-off-by: David Zeuthen <davidz@redhat.com>
13 years agoUpdate the GDBus subtree example for the new API
Ryan Lortie [Mon, 19 Jul 2010 19:03:54 +0000 (15:03 -0400)]
Update the GDBus subtree example for the new API

13 years agoBug 624754 – gdbusaddress.c missing sys/wait.h
David Hoyt [Mon, 19 Jul 2010 17:52:17 +0000 (13:52 -0400)]
Bug 624754 – gdbusaddress.c missing sys/wait.h

https://bugzilla.gnome.org/show_bug.cgi?id=624754

Signed-off-by: David Zeuthen <davidz@redhat.com>
13 years agoUpdated Spanish translation
Jorge González [Sun, 18 Jul 2010 18:29:41 +0000 (20:29 +0200)]
Updated Spanish translation

13 years agoFixed headers in galician translation file
Fran Diéguez [Sun, 18 Jul 2010 17:58:33 +0000 (19:58 +0200)]
Fixed headers in galician translation file

13 years agoUpdated Hebrew translation.
Yaron Shahrabani [Sat, 17 Jul 2010 10:11:08 +0000 (13:11 +0300)]
Updated Hebrew translation.

13 years agoGDBus: Add nonce-tcp: test case
David Zeuthen [Fri, 16 Jul 2010 17:19:48 +0000 (13:19 -0400)]
GDBus: Add nonce-tcp: test case

Also fix a couple of TODO items in gdbusaddress.c

Signed-off-by: David Zeuthen <davidz@redhat.com>
14 years agoUpdated Hebrew translation.
Yaron Shahrabani [Fri, 16 Jul 2010 14:20:42 +0000 (17:20 +0300)]
Updated Hebrew translation.

14 years agog_warning on inconsistent subtree behaviour
Ryan Lortie [Thu, 15 Jul 2010 23:37:26 +0000 (19:37 -0400)]
g_warning on inconsistent subtree behaviour

If the subtree introspection function indicates that an interface exists
but then the dispatch function returns a NULL vtable for that interface,
issue a g_warning pointing programmers in the right direction.

14 years agogio: don't assume that SOCK_CLOEXEC is supported whenever it's defined
Julien Cristau [Thu, 15 Jul 2010 14:26:02 +0000 (15:26 +0100)]
gio: don't assume that SOCK_CLOEXEC is supported whenever it's defined

Just because SOCK_CLOEXEC was defined at build time doesn't mean the
kernel we're running on supports it.  So if socket() fails with EINVAL,
try again without the flag.

https://bugzilla.gnome.org/show_bug.cgi?id=624463

14 years agoBug 624484: GDBusSubtreeDispatchFunc clarification
Ryan Lortie [Thu, 15 Jul 2010 20:53:43 +0000 (16:53 -0400)]
Bug 624484: GDBusSubtreeDispatchFunc clarification

Pass NULL rather than "/" for the toplevel of a subtree.

Add clarification to the docs about how trees are flat.

14 years agoBug 624483: GDBusSubtreeEnumerateFunc clarification
Ryan Lortie [Thu, 15 Jul 2010 20:47:32 +0000 (16:47 -0400)]
Bug 624483: GDBusSubtreeEnumerateFunc clarification

Clarify a couple of things in the docs:

  1) you must return flat names (no slashes)

  2) g_strfreev() will be called on the result

  3) a benefit of using the DISPATCH_TO_UNENUMERATED flag

14 years agoBug 624473: GDBusSubtreeIntrospectFunc return type
Ryan Lortie [Thu, 15 Jul 2010 20:26:42 +0000 (16:26 -0400)]
Bug 624473: GDBusSubtreeIntrospectFunc return type

Return a NULL terminated C array instead of a GPtrArray

Also, document that %NULL is a permitted return value and clarify its
meaning.

Finally, avoid calling the enumeration function during dispatch when the
G_DBUS_SUBTREE_FLAGS_DISPATCH_TO_UNENUMERATED_NODES flag was given.