D-BUS Tutorial Version 0.4.1 15 July 2005 Havoc Pennington Red Hat, Inc.
hp@pobox.com
David Wheeler John Palmieri Red Hat, Inc.
johnp@redhat.com
What is D-BUS? D-BUS is a system for interprocess communication (IPC). Architecturally, it has several layers: A library, libdbus, that allows two applications to connect to each other and exchange messages. A message bus daemon executable, built on libdbus, that multiple applications can connect to. The daemon can route messages from one application to zero or more other applications. Wrapper libraries based on particular application frameworks. For example, libdbus-glib and libdbus-qt. There are also bindings to languages such as Python. These wrapper libraries are the API most people should use, as they simplify the details of D-BUS programming. libdbus is intended to be a low-level backend for the higher level bindings. Much of the libdbus API is only useful for binding implementation. If you just want to use D-BUS and don't care how it works, jump directly to . Otherwise, read on. libdbus only supports one-to-one connections, just like a raw network socket. However, rather than sending byte streams over the connection, you send messages. Messages have a header identifying the kind of message, and a body containing a data payload. libdbus also abstracts the exact transport used (sockets vs. whatever else), and handles details such as authentication. The message bus daemon forms the hub of a wheel. Each spoke of the wheel is a one-to-one connection to an application using libdbus. An application sends a message to the bus daemon over its spoke, and the bus daemon forwards the message to other connected applications as appropriate. Think of the daemon as a router. The bus daemon has multiple instances on a typical computer. The first instance is a machine-global singleton, that is, a system daemon similar to sendmail or Apache. This instance has heavy security restrictions on what messages it will accept, and is used for systemwide communication. The other instances are created one per user login session. These instances allow applications in the user's session to communicate with one another. The systemwide and per-user daemons are separate. Normal within-session IPC does not involve the systemwide message bus process and vice versa. D-BUS applications There are many, many technologies in the world that have "Inter-process communication" or "networking" in their stated purpose: CORBA, DCE, DCOM, DCOP, XML-RPC, SOAP, MBUS, Internet Communications Engine (ICE), and probably hundreds more. Each of these is tailored for particular kinds of application. D-BUS is designed for two specific cases: Communication between desktop applications in the same desktop session; to allow integration of the desktop session as a whole, and address issues of process lifecycle (when do desktop components start and stop running). Communication between the desktop session and the operating system, where the operating system would typically include the kernel and any system daemons or processes. For the within-desktop-session use case, the GNOME and KDE desktops have significant previous experience with different IPC solutions such as CORBA and DCOP. D-BUS is built on that experience and carefully tailored to meet the needs of these desktop projects in particular. D-BUS may or may not be appropriate for other applications; the FAQ has some comparisons to other IPC systems. The problem solved by the systemwide or communication-with-the-OS case is explained well by the following text from the Linux Hotplug project:
A gap in current Linux support is that policies with any sort of dynamic "interact with user" component aren't currently supported. For example, that's often needed the first time a network adapter or printer is connected, and to determine appropriate places to mount disk drives. It would seem that such actions could be supported for any case where a responsible human can be identified: single user workstations, or any system which is remotely administered. This is a classic "remote sysadmin" problem, where in this case hotplugging needs to deliver an event from one security domain (operating system kernel, in this case) to another (desktop for logged-in user, or remote sysadmin). Any effective response must go the other way: the remote domain taking some action that lets the kernel expose the desired device capabilities. (The action can often be taken asynchronously, for example letting new hardware be idle until a meeting finishes.) At this writing, Linux doesn't have widely adopted solutions to such problems. However, the new D-Bus work may begin to solve that problem.
D-BUS may happen to be useful for purposes other than the one it was designed for. Its general properties that distinguish it from other forms of IPC are: Binary protocol designed to be used asynchronously (similar in spirit to the X Window System protocol). Stateful, reliable connections held open over time. The message bus is a daemon, not a "swarm" or distributed architecture. Many implementation and deployment issues are specified rather than left ambiguous. Semantics are similar to the existing DCOP system, allowing KDE to adopt it more easily. Security features to support the systemwide mode of the message bus.
Concepts Some basic concepts apply no matter what application framework you're using to write a D-BUS application. The exact code you write will be different for GLib vs. Qt vs. Python applications, however. Here is a diagram (png svg) that may help you visualize the concepts that follow. Objects and Object Paths Each application using D-BUS contains objects, which generally map to GObject, QObject, C++ objects, or Python objects (but need not). An object is an instance rather than a type. When messages are received over a D-BUS connection, they are sent to a specific object, not to the application as a whole. To allow messages to specify their destination object, there has to be a way to refer to an object. In your favorite programming language, this is normally called a pointer or reference. However, these references are implemented as memory addresses relative to the address space of your application, and thus can't be passed from one application to another. To solve this, D-BUS introduces a name for each object. The name looks like a filesystem path, for example an object could be named /org/kde/kspread/sheets/3/cells/4/5. Human-readable paths are nice, but you are free to create an object named /com/mycompany/c5yo817y0c1y1c5b if it makes sense for your application. Namespacing object paths is smart, by starting them with the components of a domain name you own (e.g. /org/kde). This keeps different code modules in the same process from stepping on one another's toes. Interfaces Each object supports one or more interfaces. Think of an interface as a named group of methods and signals, just as it is in GLib or Qt or Java. Interfaces define the type of an object instance. Message Types Messages are not all the same; in particular, D-BUS has 4 built-in message types: Method call messages ask to invoke a method on an object. Method return messages return the results of invoking a method. Error messages return an exception caused by invoking a method. Signal messages are notifications that a given signal has been emitted (that an event has occurred). You could also think of these as "event" messages. A method call maps very simply to messages, then: you send a method call message, and receive either a method return message or an error message in reply. Bus Names Object paths, interfaces, and messages exist on the level of libdbus and the D-BUS protocol; they are used even in the 1-to-1 case with no message bus involved. Bus names, on the other hand, are a property of the message bus daemon. The bus maintains a mapping from names to message bus connections. These names are used to specify the origin and destination of messages passing through the message bus. When a name is mapped to a particular application's connection, that application is said to own that name. On connecting to the bus daemon, each application immediately owns a special name called the unique connection name. A unique name begins with a ':' (colon) character; no other names are allowed to begin with that character. Unique names are special because they are created dynamically, and are never re-used during the lifetime of the same bus daemon. You know that a given unique name will have the same owner at all times. An example of a unique name might be :34-907. The numbers after the colon have no meaning other than their uniqueness. Applications may ask to own additional well-known names. For example, you could write a specification to define a name called com.mycompany.TextEditor. Your definition could specify that to own this name, an application should have an object at the path /com/mycompany/TextFileManager supporting the interface org.freedesktop.FileHandler. Applications could then send messages to this bus name, object, and interface to execute method calls. You could think of the unique names as IP addresses, and the well-known names as domain names. So com.mycompany.TextEditor might map to something like :34-907 just as mycompany.com maps to something like 192.168.0.5. Names have a second important use, other than routing messages. They are used to track lifecycle. When an application exits (or crashes), its connection to the message bus will be closed by the operating system kernel. The message bus then sends out notification messages telling remaining applications that the application's names have lost their owner. By tracking these notifications, your application can reliably monitor the lifetime of other applications. Addresses Applications using D-BUS are either servers or clients. A server listens for incoming connections; a client connects to a server. Once the connection is established, it is a symmetric flow of messages; the client-server distinction only matters when setting up the connection. A D-BUS address specifies where a server will listen, and where a client will connect. For example, the address unix:path=/tmp/abcdef specifies that the server will listen on a UNIX domain socket at the path /tmp/abcdef and the client will connect to that socket. An address can also specify TCP/IP sockets, or any other transport defined in future iterations of the D-BUS specification. When using D-BUS with a message bus, the bus daemon is a server and all other applications are clients of the bus daemon. libdbus automatically discovers the address of the per-session bus daemon by reading an environment variable. It discovers the systemwide bus daemon by checking a well-known UNIX domain socket path (though you can override this address with an environment variable). If you're using D-BUS without a bus daemon, it's up to you to define which application will be the server and which will be the client, and specify a mechanism for them to agree on the server's address. Big Conceptual Picture Pulling all these concepts together, to specify a particular method call on a particular object instance, a number of nested components have to be named: Address -> [Bus Name] -> Path -> Interface -> Method The bus name is in brackets to indicate that it's optional -- you only provide a name to route the method call to the right application when using the bus daemon. If you have a direct connection to another application, bus names aren't used; there's no bus daemon. The interface is also optional, primarily for historical reasons; DCOP does not require specifying the interface, instead simply forbidding duplicate method names on the same object instance. D-BUS will thus let you omit the interface, but if your method name is ambiguous it is undefined which method will be invoked. GLib API: Using Remote Objects The GLib binding is defined in the header file <dbus/dbus-glib.h>. D-BUS - GLib type mappings The heart of the GLib bindings for D-BUS is the mapping it provides between D-BUS "type signatures" and GLib types (GType). The D-BUS type system is composed of a number of "basic" types, along with several "container" types. Basic type mappings Below is a list of the basic types, along with their associated mapping to a GType. D-BUS basic type GType Free function Notes BYTE G_TYPE_UCHAR BOOLEAN G_TYPE_BOOLEAN INT16 G_TYPE_INT Will be changed to a G_TYPE_INT16 once GLib has it UINT16 G_TYPE_UINT Will be changed to a G_TYPE_UINT16 once GLib has it INT32 G_TYPE_INT Will be changed to a G_TYPE_INT32 once GLib has it UINT32 G_TYPE_UINT Will be changed to a G_TYPE_UINT32 once GLib has it INT64 G_TYPE_GINT64 UINT64 G_TYPE_GUINT64 DOUBLE G_TYPE_DOUBLE STRING G_TYPE_STRING g_free OBJECT_PATH DBUS_TYPE_G_PROXY g_object_unref The returned proxy does not have an interface set; use dbus_g_proxy_set_interface to invoke methods As you can see, the basic mapping is fairly straightforward. Container type mappings The D-BUS type system also has a number of "container" types, such as DBUS_TYPE_ARRAY and DBUS_TYPE_STRUCT. The D-BUS type system is fully recursive, so one can for example have an array of array of strings (i.e. type signature aas). However, not all of these types are in common use; for example, at the time of this writing the author knows of no one using DBUS_TYPE_STRUCT, or a DBUS_TYPE_ARRAY containing any non-basic type. The approach the GLib bindings take is pragmatic; try to map the most common types in the most obvious way, and let using less common and more complex types be less "natural". First, D-BUS type signatures which have an "obvious" corresponding builtin GLib type are mapped using that type: D-BUS type signature Description GType C typedef Free function Notes as Array of strings G_TYPE_STRV char ** g_strfreev v Generic value container G_TYPE_VALUE GValue * g_value_unset The calling conventions for values expect that method callers have allocated return values; see below. The next most common recursive type signatures are arrays of basic values. The most obvious mapping for arrays of basic types is a GArray. Now, GLib does not provide a builtin GType for GArray. However, we actually need more than that - we need a "parameterized" type which includes the contained type. Why we need this we will see below. The approach taken is to create these types in the D-BUS GLib bindings; however, there is nothing D-BUS specific about them. In the future, we hope to include such "fundamental" types in GLib itself. D-BUS type signature Description GType C typedef Free function Notes ay Array of bytes DBUS_TYPE_G_BYTE_ARRAY GArray * g_array_free au Array of uint DBUS_TYPE_G_UINT_ARRAY GArray * g_array_free ai Array of int DBUS_TYPE_G_INT_ARRAY GArray * g_array_free ax Array of int64 DBUS_TYPE_G_INT64_ARRAY GArray * g_array_free at Array of uint64 DBUS_TYPE_G_UINT64_ARRAY GArray * g_array_free ad Array of double DBUS_TYPE_G_DOUBLE_ARRAY GArray * g_array_free ab Array of boolean DBUS_TYPE_G_BOOLEAN_ARRAY GArray * g_array_free D-BUS also includes a special type DBUS_TYPE_DICT_ENTRY which is only valid in arrays. It's intended to be mapped to a "dictionary" type by bindings. The obvious GLib mapping here is GHashTable. Again, however, there is no builtin GType for a GHashTable. Moreover, just like for arrays, we need a parameterized type so that the bindings can communiate which types are contained in the hash table. At present, only strings are supported. Work is in progress to include more types. D-BUS type signature Description GType C typedef Free function Notes a{ss} Dictionary mapping strings to strings DBUS_TYPE_G_STRING_STRING_HASHTABLE GHashTable * g_hash_table_destroy Arbitrarily recursive type mappings Finally, it is possible users will want to write or invoke D-BUS methods which have arbitrarily complex type signatures not directly supported by these bindings. For this case, we have a DBusGValue which acts as a kind of special variant value which may be iterated over manually. The GType associated is DBUS_TYPE_G_VALUE. TODO insert usage of DBUS_TYPE_G_VALUE here. A sample program Here is a D-BUS program using the GLib bindings. int main (int argc, char **argv) { DBusGConnection *connection; GError *error; DBusGProxy *proxy; char **name_list; char **name_list_ptr; g_type_init (); error = NULL; connection = dbus_g_bus_get (DBUS_BUS_SESSION, &error); if (connection == NULL) { g_printerr ("Failed to open connection to bus: %s\n", error->message); g_error_free (error); exit (1); } /* Create a proxy object for the "bus driver" (name "org.freedesktop.DBus") */ proxy = dbus_g_proxy_new_for_name (connection, DBUS_SERVICE_DBUS, DBUS_PATH_DBUS, DBUS_INTERFACE_DBUS); /* Call ListNames method, wait for reply */ error = NULL; if (!dbus_g_proxy_call (proxy, "ListNames", &error, G_TYPE_INVALID, G_TYPE_STRV, &name_list, G_TYPE_INVALID)) { /* Just do demonstrate remote exceptions versus regular GError */ if (error->domain == DBUS_GERROR && error->code == DBUS_GERROR_REMOTE_EXCEPTION) g_printerr ("Caught remote method exception %s: %s", dbus_g_error_get_name (error), error->message); else g_printerr ("Error: %s\n", error->message); g_error_free (error); exit (1); } /* Print the results */ g_print ("Names on the message bus:\n"); for (name_list_ptr = name_list; *name_list_ptr; name_list_ptr++) { g_print (" %s\n", *name_list_ptr); } g_strfreev (name_list); g_object_unref (proxy); return 0; } Program initalization A connection to the bus is acquired using dbus_g_bus_get. Next, a proxy is created for the object "/org/freedesktop/DBus" with interface org.freedesktop.DBus on the service org.freedesktop.DBus. This is a proxy for the message bus itself. Understanding method invocation You have a number of choices for method invocation. First, as used above, dbus_g_proxy_call sends a method call to the remote object, and blocks until reply is recieved. The outgoing arguments are specified in the varargs array, terminated with G_TYPE_INVALID. Next, pointers to return values are specified, followed again by G_TYPE_INVALID. To invoke a method asynchronously, use dbus_g_proxy_begin_call. This returns a DBusGPendingCall object; you may then set a notification function using dbus_g_pending_call_set_notify. Connecting to object signals You may connect to signals using dbus_g_proxy_add_signal and dbus_g_proxy_connect_signal. You must invoke dbus_g_proxy_add_signal to specify the signature of your signal handlers; you may then invoke dbus_g_proxy_connect_signal multiple times. Note that it will often be the case that there is no builtin marshaller for the type signature of a remote signal. In that case, you must generate a marshaller yourself by using glib-genmarshal, and then register it using dbus_g_object_register_marshaller. Error handling and remote exceptions All of the GLib binding methods such as dbus_g_proxy_end_call return a GError. This GError can represent two different things: An internal D-BUS error, such as an out-of-memory condition, an I/O error, or a network timeout. Errors generated by the D-BUS library itself have the domain DBUS_GERROR, and a corresponding code such as DBUS_GERROR_NO_MEMORY. It will not be typical for applications to handle these errors specifically. A remote D-BUS exception, thrown by the peer, bus, or service. D-BUS remote exceptions have both a textual "name" and a "message". The GLib bindings store this information in the GError, but some special rules apply. The set error will have the domain DBUS_GERROR as above, and will also have the code DBUS_GERROR_REMOTE_EXCEPTION. In order to access the remote exception name, you must use a special accessor, such as dbus_g_error_has_name or dbus_g_error_get_name. The remote exception detailed message is accessible via the regular GError message member. More examples of method invocation Sending an integer and string, receiving an array of bytes GArray *arr; error = NULL; if (!dbus_g_proxy_call (proxy, "Foobar", &error, G_TYPE_INT, 42, G_TYPE_STRING, "hello", G_TYPE_INVALID, DBUS_TYPE_G_UCHAR_ARRAY, &arr, G_TYPE_INVALID)) { /* Handle error */ } g_assert (arr != NULL); printf ("got back %u values", arr->len); Sending a GHashTable GHashTable *hash = g_hash_table_new (g_str_hash, g_str_equal); guint32 ret; g_hash_table_insert (hash, "foo", "bar"); g_hash_table_insert (hash, "baz", "whee"); error = NULL; if (!dbus_g_proxy_call (proxy, "HashSize", &error, DBUS_TYPE_G_STRING_STRING_HASH, hash, G_TYPE_INVALID, G_TYPE_UINT, &ret, G_TYPE_INVALID)) { /* Handle error */ } g_assert (ret == 2); g_hash_table_destroy (hash); Receiving a boolean and a string gboolean boolret; char *strret; error = NULL; if (!dbus_g_proxy_call (proxy, "GetStuff", &error, G_TYPE_INVALID, G_TYPE_BOOLEAN, &boolret, G_TYPE_STRING, &strret, G_TYPE_INVALID)) { /* Handle error */ } printf ("%s %s", boolret ? "TRUE" : "FALSE", strret); g_free (strret); Sending two arrays of strings /* NULL terminate */ char *strs_static[] = {"foo", "bar", "baz", NULL}; /* Take pointer to array; cannot pass array directly */ char **strs_static_p = strs_static; char **strs_dynamic; strs_dynamic = g_new (char *, 4); strs_dynamic[0] = g_strdup ("hello"); strs_dynamic[1] = g_strdup ("world"); strs_dynamic[2] = g_strdup ("!"); /* NULL terminate */ strs_dynamic[3] = NULL; error = NULL; if (!dbus_g_proxy_call (proxy, "TwoStrArrays", &error, G_TYPE_STRV, strs_static_p, G_TYPE_STRV, strs_dynamic, G_TYPE_INVALID, G_TYPE_INVALID)) { /* Handle error */ } g_strfreev (strs_dynamic); Sending a boolean, receiving an array of strings char **strs; char **strs_p; gboolean blah; error = NULL; blah = TRUE; if (!dbus_g_proxy_call (proxy, "GetStrs", &error, G_TYPE_BOOLEAN, blah, G_TYPE_INVALID, G_TYPE_STRV, &strs, G_TYPE_INVALID)) { /* Handle error */ } for (strs_p = strs; *strs_p; strs_p++) printf ("got string: \"%s\"", *strs_p); g_strfreev (strs); Sending a variant GValue val = {0, }; g_value_init (&val, G_TYPE_STRING); g_value_set_string (&val, "hello world"); error = NULL; if (!dbus_g_proxy_call (proxy, "SendVariant", &error, G_TYPE_VALUE, &val, G_TYPE_INVALID, G_TYPE_INVALID)) { /* Handle error */ } g_assert (ret == 2); g_value_unset (&val); Receiving a variant GValue val = {0, }; error = NULL; if (!dbus_g_proxy_call (proxy, "GetVariant", &error, G_TYPE_INVALID, G_TYPE_VALUE, &val, G_TYPE_INVALID)) { /* Handle error */ } if (G_VALUE_TYPE (&val) == G_TYPE_STRING) printf ("%s\n", g_value_get_string (&val)); else if (G_VALUE_TYPE (&val) == G_TYPE_INT) printf ("%d\n", g_value_get_int (&val)); else ... g_value_unset (&val); GLib API: Implementing Objects At the moment, to expose a GObject via D-BUS, you must write XML by hand which describes the methods exported by the object. In the future, this manual step will be obviated by the upcoming GLib introspection support. Here is a sample XML file which describes an object that exposes one method, named ManyArgs. <?xml version="1.0" encoding="UTF-8" ?> <node name="/com/example/MyObject"> <interface name="com.example.MyObject"> <annotation name="org.freedesktop.DBus.GLib.CSymbol" value="my_object"/> <method name="ManyArgs"> <!-- This is optional, and in this case is redunundant --> <annotation name="org.freedesktop.DBus.GLib.CSymbol" value="my_object_many_args"/> <arg type="u" name="x" direction="in" /> <arg type="s" name="str" direction="in" /> <arg type="d" name="trouble" direction="in" /> <arg type="d" name="d_ret" direction="out" /> <arg type="s" name="str_ret" direction="out" /> </method> </interface> </node> This XML is in the same format as the D-BUS introspection XML format. Except we must include an "annotation" which give the C symbols corresponding to the object implementation prefix (my_object). In addition, if particular methods symbol names deviate from C convention (i.e. ManyArgs -> many_args), you may specify an annotation giving the C symbol. Once you have written this XML, run dbus-binding-tool --mode=glib-server FILENAME > HEADER_NAME. to generate a header file. For example: dbus-binding-tool --mode=glib-server my-objet.xml > my-object-glue.h. Next, include the generated header in your program, and invoke dbus_g_object_class_install_info, passing the object class and "object info" included in the header. For example: dbus_g_object_type_install_info (COM_FOO_TYPE_MY_OBJECT, &com_foo_my_object_info); This should be done exactly once per object class. To actually implement the method, just define a C function named e.g. my_object_many_args in the same file as the info header is included. At the moment, it is required that this function conform to the following rules: The function must return a value of type gboolean; TRUE on success, and FALSE otherwise. The first parameter is a pointer to an instance of the object. Following the object instance pointer are the method input values. Following the input values are pointers to return values. The final parameter must be a GError **. If the function returns FALSE for an error, the error parameter must be initalized with g_set_error. Finally, you can export an object using dbus_g_connection_register_g_object. For example: dbus_g_connection_register_g_object (connection, "/com/foo/MyObject", obj); Python API: Using Remote Objects The Python bindings provide a simple to use interface for talking over D-BUS. Where possible much of the inner-workings of D-BUS are hidden behind what looks like normal Python objects. D-BUS - Python type mappings While python itself is a largely untyped language D-BUS provides a simple type system for talking with other languages which may be strongly typed. Python for the most part tries automatically map python objects to types on the bus. It is none the less good to know what the type mappings are so one can better utilize services over the bus. Basic type mappings Below is a list of the basic types, along with their associated mapping to a Python object. D-BUS basic type Python object Notes BYTE dbus.Byte BOOLEAN dbus.Boolean Any variable assigned a True or False boolean value will automatically be converted into a BOOLEAN over the bus INT16 dbus.Int16 UINT16 dbus.UInt16 INT32 dbus.Int32 This is the default mapping for Python integers UINT32 dbus.UInt32 INT64 dbus.Int64 UINT64 dbus.UInt64 DOUBLE dbus.Double Any variable assigned a floating point number will automatically be converted into a DOUBLE over the bus STRING dbus.String Any variable assigned a quoted string will automatically be converted into a STRING over the bus OBJECT_PATH dbus.ObjectPath Container type mappings The D-BUS type system also has a number of "container" types, such as DBUS_TYPE_ARRAY and DBUS_TYPE_STRUCT. The D-BUS type system is fully recursive, so one can for example have an array of array of strings (i.e. type signature aas). D-BUS container types have native corresponding built-in Python types so it is easy to use them. D-BUS type Python type Notes ARRAY Python lists Python lists, denoted by square brackets [], are converted into arrays and visa versa. The one restriction is that when sending a Python list each element of the list must be of the same type. This is because D-BUS arrays can contain only one element type. Use Python tuples for mixed types. STRUCT Python tuple Python tuples, denoted by parentheses (,), are converted into structs and visa versa. Tuples can have mixed types. DICTIONARY Python dictionary D-BUS doesn't have an explicit dictionary type. Instead it uses LISTS of DICT_ENTRIES to represent a dictionary. A DICT_ENTRY is simply a two element struct containing a key/value pair. Python dictionaries are automatically converted to a LIST of DICT_ENTRIES and visa versa. VARIANT any type A variant is a container for any type. Python exports its methods to accept only variants since we are an untyped language and can demarshal into any Python type. Invoking Methods Here is a D-BUS program using the Python bindings to get a listing of all names on the session bus. import dbus bus = dbus.SessionBus() proxy_obj = bus.bus.get_object('org.freedesktop.DBus', '/org/freedesktop/DBus') dbus_iface = dbus.Interface(proxy_obj, 'org.freedesktop.DBus') print dbus_iface.ListNames() Notice I get an interface on the proxy object and use that to make the call. While the specifications state that you do not need to specify an interface if the call is unambiguous (i.e. only one method implements that name) due to a bug on the bus that drops messages which don't have an interface field you need to specify interfaces at this time. In any event it is always good practice to specify the interface of the method you wish to call to avoid any side effects should a method of the same name be implemented on another interface. You can specify the interface for a single call using the dbus_interface keyword. proxy_obj.ListNames(dbus_interface = 'org.freedesktop.DBus') This is all fine and good if all you want to do is call methods on the bus and then exit. In order to do more complex things such as use a GUI or make asynchronous calls you will need a mainloop. You would use asynchronous calls because in GUI applications it is very bad to block for any long period of time. This cause the GUI to seem to freeze. Since replies to D-BUS messages can take an indeterminate amount of time using async calls allows you to return control to the GUI while you wait for the reply. This is exceedingly easy to do in Python. Here is an example using the GLib/GTK+ mainloop. import gtk import dbus if getattr(dbus, 'version', (0,0,0)) >= (0,41,0): import dbus.glib def print_list_names_reply(list): print str(list) def print_error(e): print str(e) bus = dbus.SessionBus() proxy_obj = bus.bus.get_object('org.freedesktop.DBus', '/org/freedesktop/DBus') dbus_iface = dbus.Interface(proxy_obj, 'org.freedesktop.DBus') dbus_iface.ListNames(reply_handler=print_list_names_reply, error_handler=print_error) gtk.main() In the above listing you will notice the reply_handler and error_handler keywords. These tell the method that it should be called async and to call print_list_names_reply or print_error depending if you get a reply or an error. The signature for replys depends on the number of arguments being sent back. Error handlers always take one parameter which is the error object returned. You will also notice that I check the version of the dbus bindings before importing dbus.glib. In older versions glib was the only available mainloop. As of version 0.41.0 we split out the glib dependency to allow for other mainloops to be implemented. Notice also the python binding version does not match up with the D-BUS version. Once we reach 1.0 this should change with Python changes simply tracking the D-BUS changes. While the glib mainloop is the only mainloop currently implemented, integrating other mainloops should be very easy to do. There are plans for creating a a generic mainloop to be the default for non gui programs. Listening for Signals Signals are emitted by objects on the bus to notify listening programs that an event has occurred. There are a couple of ways to register a signal handler on the bus. One way is to attach to an already created proxy using the connect_to_signal method which takes a signal name and handler as arguments. Let us look at an example of connecting to the HAL service to receive signals when devices are added and removed and when devices register a capability. This example assumes you have HAL already running. import gtk import dbus if getattr(dbus, 'version', (0,0,0)) >= (0,41,0): import dbus.glib def device_added_callback(udi): print 'Device with udi %s was added' % (udi) def device_removed_callback(udi): print 'Device with udi %s was added' % (udi) def device_capability_callback(udi, capability): print 'Device with udi %s added capability %s' % (udi, capability) bus = dbus.SystemBus() hal_manager_obj = bus.get_object('org.freedesktop.Hal', '/org/freedesktop/Hal/Manager') hal_manager = dbus.Interface(hal_manager_obj, 'org.freedesktop.Hal.Manager') hal_manager.connect_to_signal('DeviceAdded', device_added_callback) hal_manager.connect_to_signal('DeviceRemoved', device_removed_callback) hal_manager.connect_to_signal('NewCapability', device_capability_callback) gtk.main() The drawback of using this method is that the service that you are connecting to has to be around when you register your signal handler. While HAL is guaranteed to be around on systems that use it this is not always the case for every service on the bus. Say our program started up before HAL, we could connect to the signal by adding a signal receiver directly to the bus. bus.add_signal_receiver(device_added_callback, 'DeviceAdded', 'org.freedesktop.Hal.Manager', 'org.freedesktop.Hal', '/org/freedesktop/Hal/Manager') bus.add_signal_receiver(device_removed_callback, 'DeviceRemoved', 'org.freedesktop.Hal.Manager', 'org.freedesktop.Hal', '/org/freedesktop/Hal/Manager') bus.add_signal_receiver(device_capability_callback, 'DeviceAdded', 'org.freedesktop.Hal.Manager', 'org.freedesktop.Hal', '/org/freedesktop/Hal/Manager') All this can be done without creating the proxy object if one wanted to but in most cases you would want to have a reference to the object so once a signal was received operations could be executed on the object. Cost of Creating a Proxy Object Note that creating proxy objects can have an associated processing cost. When introspection is implemented a proxy may wait for introspection data before processing any requests. It is generally good practice to create proxies once and reuse the proxy when calling into the object. Constantly creating the same proxy over and over again can become a bottleneck for your program. TODO: example of getting information about devices from HAL Python API: Implementing Objects Implementing object on the bus is just as easy as invoking methods or listening for signals on the bus. Version Alert The Python D-BUS bindings require version 2.4 or greater of Python when creating D-BUS objects. Inheriting From dbus.service.Object In order to export a Python object over the bus one must first get a bus name and then create a Python object that inherits from dbus.service.Object. The following is the start of an example HelloWorld object that we want to export over the session bus. import dbus import dbus.service if getattr(dbus, 'version', (0,0,0)) >= (0,41,0): import dbus.glib class HelloWorldObject(dbus.service.Object): def __init__(self, bus_name, object_path='/org/freedesktop/HelloWorldObject'): dbus.service.Object.__init__(self, bus_name, object_path) session_bus = dbus.SessionBus() bus_name = dbus.service.BusName('org.freedesktop.HelloWorld', bus=session_bus) object = HelloWorldObject(bus_name) gtk.main() Here we got the session bus, then created a BusName object which requests a name on the bus. We pass that bus name to the HelloWorldObject object which inherits from dbus.service.Object. We now have an object on the bus but it is pretty useless. Exporting Methods Over The Bus Let's make this object do something and export a method over the bus. import dbus import dbus.service if getattr(dbus, 'version', (0,0,0)) >= (0,41,0): import dbus.glib class HelloWorldObject(dbus.service.Object): def __init__(self, bus_name, object_path='/org/freedesktop/HelloWorldObject'): dbus.service.Object.__init__(self, bus_name, object_path) @dbus.service.method('org.freedesktop.HelloWorldIFace') def hello(self): return 'Hello from the HelloWorldObject' session_bus = dbus.SessionBus() bus_name = dbus.service.BusName('org.freedesktop.HelloWorld', bus=session_bus) object = HelloWorldObject(bus_name) gtk.main() Python Decorators Notice the @ symbol on the line before the hello method. This is a new directive introduced in Python 2.4. It is called a decorator and it "decorates" methods. All you have to know is that it provides metadata that can then be used to alter the behavior of the method being decorated. In this case we are telling the bindings that the hello method should be exported as a D-BUS method over the bus. As you can see we exported the hello method as part of the org.freedesktop.HelloWorldIFace interface. It takes no arguments and returns a string to the calling program. Let's create a proxy and invoke this method. import dbus bus = dbus.SessionBus() proxy_obj = bus.bus.get_object('org.freedesktop.HelloWorld', '/org/freedesktop/HelloWorldObject') iface = dbus.Interface(proxy_obj, 'org.freedesktop.HelloWorldIFace') print iface.hello() When invoking methods exported over the bus the bindings automatically know how many parameters the method exports. You can even make a method that exports an arbitrary number of parameters. Also, whatever you return will automatically be transfered as a reply over the bus. Some examples. @dbus.service.method('org.freedesktop.HelloWorldIFace') def one_arg(self, first_arg): return 'I got arg %s' % first_arg @dbus.service.method('org.freedesktop.HelloWorldIFace') def two_args(self, first_arg, second_arg): return ('I got 2 args', first_arg, second_arg) @dbus.service.method('org.freedesktop.HelloWorldIFace') def return_list(self): return [1, 2, 3, 4, 5, 6] @dbus.service.method('org.freedesktop.HelloWorldIFace') def return_dict(self): return {one: '1ne', two: '2wo', three: '3ree'} Emitting Signals Setting up signals to emit is just as easy as exporting methods. It uses the same syntax as methods. import dbus import dbus.service if getattr(dbus, 'version', (0,0,0)) >= (0,41,0): import dbus.glib class HelloWorldObject(dbus.service.Object): def __init__(self, bus_name, object_path='/org/freedesktop/HelloWorldObject'): dbus.service.Object.__init__(self, bus_name, object_path) @dbus.service.method('org.freedesktop.HelloWorldIFace') def hello(self): return 'Hello from the HelloWorldObject' @dbus.service.signal('org.freedesktop.HelloWorldIFace') def hello_signal(self, message): pass session_bus = dbus.SessionBus() bus_name = dbus.service.BusName('org.freedesktop.HelloWorld', bus=session_bus) object = HelloWorldObject(bus_name) object.hello_signal('I sent a hello signal') gtk.main() Adding a @dbus.service.signal decorator to a method turns it into a signal emitter. You can put code in this method to do things like keep track of how many times you call the emitter or to print out debug messages but for the most part a pass noop will do. Whenever you call the emitter a signal will be emitted with the parameters you passed in as arguments. In the above example we send the message 'I sent a hello signal' with the signal. Inheriting from HelloWorldObject One of the cool things you can do in Python is inherit from another D-BUS object. We use this trick in the bindings to provide a default implementation for the org.freedesktop.DBus.Introspectable interface. Let's inherit from the HelloWorldObject example above and overide the hello method to say goodbye. class HelloWorldGoodbyeObject(HelloWorldObject): def __init__(self, bus_name, object_path='/org/freedesktop/HelloWorldGoodbyeObject'): HelloWorldObject.__init__(self, bus_name, object_path) @dbus.service.method('org.freedesktop.HelloWorldGoodbyeIFace') def hello(self): return 'Goodbye' goodbye_object = HelloWorldGoodbyeObject(bus_name) Let's now call both methods with a little help from interfaces. import dbus bus = dbus.SessionBus() proxy_obj = bus.bus.get_object('org.freedesktop.HelloWorld', '/org/freedesktop/HelloWorldGoodbyeObject') print proxy_obj.hello(dbus_interface='org.freedesktop.HelloWorldIFace') print proxy_obj.hello(dbus_interface='org.freedesktop.HelloWorldGoodbyeIFace') This should print out 'Hello from the HelloWorldObject' followed by a 'Goodbye'. Conclusion As you can see, using D-BUS from Python is an extremely easy proposition. Hopefully the tutorial has been helpful in getting you started. If you need anymore help please feel free to post on the mailing list. The Python bindings are still in a state of flux and there may be API changes in the future. This tutorial will be updated if such changes occur. Qt API: Using Remote Objects The Qt bindings are not yet documented. Qt API: Implementing Objects The Qt bindings are not yet documented.