summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Install test metadata to a fully versioned locationtelepathy-mission-control-5.99.11Simon McVittie2014-05-141-1/+1
|
* 5.99.11Simon McVittie2014-05-142-3/+13
|
* Stop putting TpDBusPropertiesMixinClass in class structsSimon McVittie2014-05-073-29/+12
| | | | | | | We could have included <telepathy-glib/telepathy-glib-dbus.h> instead, but this seems nicer. Bug: https://bugs.freedesktop.org/show_bug.cgi?id=77189
* Rewrite the protocol name for local-xmpp accounts created by Telepathy1Simon McVittie2014-04-081-0/+12
| | | | | Bug: https://bugs.freedesktop.org/show_bug.cgi?id=77029 Reviewed-by: Xavier Claessens
* Stop using tp_dbus_properties_mixin_iface_init which has been removedSimon McVittie2014-04-082-6/+2
| | | | Bug: https://bugs.freedesktop.org/show_bug.cgi?id=77144
* McpRequest: use GVariant for the requested propertiesSimon McVittie2014-04-085-18/+14
| | | | Bug: https://bugs.freedesktop.org/show_bug.cgi?id=77139
* McpDispatchOperation: use GVariant for the channel's propertiesSimon McVittie2014-04-085-55/+32
| | | | Bug: https://bugs.freedesktop.org/show_bug.cgi?id=77139
* McpAccountStorage: use GVariant instead of dbus-glib typesSimon McVittie2014-04-084-99/+68
| | | | Bug: https://bugs.freedesktop.org/show_bug.cgi?id=77139
* change terminology from "handle type" to "entity type"Simon McVittie2014-04-088-17/+19
| | | | Bug: https://bugs.freedesktop.org/show_bug.cgi?id=77139
* Adapt for removal of tp_connection_manager_param_get_defaultSimon McVittie2014-04-081-2/+19
| | | | | | This is far from the best possible implementation, but it unblocks us. Bug: https://bugs.freedesktop.org/show_bug.cgi?id=77139
* restart developmentSimon McVittie2014-04-032-1/+6
|
* 5.99.10telepathy-mission-control-5.99.10Simon McVittie2014-04-032-17/+17
|
* mc-debug-server: if we have several reasons to stop, don't crashSimon McVittie2014-04-021-4/+17
| | | | | | | | | Previously, when the session and (fake) system bus closed (simultaneously, because they are in fact the same dbus-daemon), we would call mcd_mission_abort() for the first one. That results in a call to tp_clear_object (&mcd), and then the second call to mcd_mission_abort (mcd) would crash with a critical warning, because MCD_IS_MISSION (NULL) is false.
* Remove references to TpDBusDaemonSimon McVittie2014-04-0111-129/+43
|
* McdManager: use tp_client_factory_ensure_connection_managerSimon McVittie2014-04-011-3/+2
|
* mc-tool: use the global singleton client factorySimon McVittie2014-04-011-12/+4
|
* McdClientRegistry: stop having a TpDBusDaemonSimon McVittie2014-04-014-36/+14
|
* McdClientRegistry: use GDBus to call ListNames, ListActivatableNamesSimon McVittie2014-04-011-27/+57
| | | | | Previously we were using the "weak object" mechanism, whereas now we're holding a ref for the duration of the calls.
* McdHandlerMap: have a TpClientFactory instead of a TpDBusDaemonSimon McVittie2014-04-013-18/+19
|
* McdClientProxy: have a factorySimon McVittie2014-04-013-4/+6
|
* McdClientRegistry: have a TpClientFactorySimon McVittie2014-04-013-8/+32
|
* Pass TpClientFactory to McdDispatcherSimon McVittie2014-04-013-9/+28
|
* mcp-plugin: avoid TpDBusDaemon, just use g_bus_get_sync()Simon McVittie2014-04-011-6/+15
|
* McdAccount: have a construct-only factory and demote dbus-daemon to readableSimon McVittie2014-04-011-6/+20
|
* mcd_account_manager_get_dbus_daemon: replace with ↵Simon McVittie2014-04-013-12/+8
| | | | mcd_account_manager_get_factory
* McdMaster: take a TpClientFactory at construct-timeSimon McVittie2014-04-012-7/+22
| | | | | | | We don't want the shared singleton (tp_client_factory_dup()) because it constructs specialized TpChannel subclasses, and we'd rather have the base class. Similarly, if a plugin asks for extra features on the shared singleton, MC actively doesn't want them.
* McdAccountManager: replace tp_list_connection_names with direct GDBus callsSimon McVittie2014-03-271-24/+46
| | | | | Also use a direct GDBus call to disconnect stray connections, rather than making a temporary proxy for the purpose.
* McdHandlerMap: add some assertions for claritySimon McVittie2014-03-271-0/+17
|
* McdHandlerMap: use g_bus_watch_name_on_connectionSimon McVittie2014-03-271-61/+46
|
* wait-for-name: use GDBus directlySimon McVittie2014-03-271-31/+57
|
* McdClient: replace tp_dbus_daemon_watch_name_owner with ↵Simon McVittie2014-03-271-13/+29
| | | | g_bus_watch_name_on_connection
* sync_dbus: support being called on a CM, tooSimon McVittie2014-03-271-0/+6
|
* McdDispatchOperation: don't run approvers if we already have a resultSimon McVittie2014-03-271-10/+11
| | | | | | | | | | | It seems that under GDBus' changed timing, we can have this sequence of events: * decide that we're ready to run approvers, scheduling it to happen in a high-priority idle * receive and process the Claim method call, which finishes the dispatch operation * run the idle
* tests: make sure to specify a D-Bus interface for method callsSimon McVittie2014-03-273-4/+4
| | | | | GDBus doesn't implement method calls that lack an explicit interface. This is A Good Thing™.
* mc-debug-server: hold a reference to the debug senderSimon McVittie2014-03-271-0/+4
| | | | | It seems we were instantiating it once per debug message. Oops. (The real Mission Control got this right, though.)
* mc-debug-server: export a GDBus interface rather than a libdbus filterSimon McVittie2014-03-272-58/+70
| | | | | | | This changes it from effectively being exported at every object path to being exported at one object path. I've chosen the AccountManager object path, since that's where I put the MC pseudo-object in the test suite in an earlier commit.
* mc-debug-server: use GDBus to watch for disconnectionSimon McVittie2014-03-271-12/+24
|
* TestPermissionPluginClass etc.: adapt for GDBusSimon McVittie2014-03-271-194/+61
| | | | | Consolidate PermissionContext and RequestPermissionContext rather than using two very similar structs and code paths.
* McdStorage: rewrite unique_name to avoid dbus_g_connection_lookup_g_objectSimon McVittie2014-03-271-12/+4
| | | | | If we don't have an account in the McdStorage, its object path won't be in use.
* McdMaster: don't force exit-on-disconnect onSimon McVittie2014-03-271-5/+0
| | | | | It's on by default in both libdbus and GDBus, so this was unnecessary even with dbus-glib, and there's no point in porting it.
* McdConnection: rewrite translate_g_error() using g_dbus_error_encode_gerror()Simon McVittie2014-03-271-31/+5
|
* McdConnection: tp_connection_init_known_interfaces() no longer existsSimon McVittie2014-03-271-2/+0
|
* McdConnection: use g_dbus_connection_call() for simple Disconnect callSimon McVittie2014-03-271-7/+3
|
* Use tp_dbus_daemon_register_object to register object pathsSimon McVittie2014-03-274-27/+10
| | | | | Also make sure to register object paths before their corresponding bus names.
* Use g_dbus_method_invocation_get_senderSimon McVittie2014-03-272-8/+5
|
* McdClientRegistry: use GDBus to watch for NameOwnerChangedSimon McVittie2014-03-272-63/+28
| | | | | Depend on D-Bus 1.6 for arg0namespace (I'm assuming here that the dbus-daemon version matches the libdbus version).
* Replace dbus_g_method_return_error with g_dbus_method_invocation_return_gerrorSimon McVittie2014-03-276-22/+22
|
* replace DBusGMethodInvocation with GDBusMethodInvocationSimon McVittie2014-03-2710-40/+40
|
* telepathy.am: update for out-of-tree releasesSimon McVittie2014-03-271-1/+1
|
* Use the AccountManager object path to call AbortSimon McVittie2014-03-271-1/+1
| | | | | | I moved the MC "object" to point to that path already, and in a later commit I'll change the debug API from being exposed at all object paths to only that one object path.
View Lorry Controller Status