1 <?xml version="1.0" standalone="no"?>
2 <!DOCTYPE article PUBLIC "-//OASIS//DTD DocBook XML V4.1.2//EN"
3 "http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd"
9 <title>D-BUS Specification</title>
10 <releaseinfo>Version 0.10</releaseinfo>
11 <date>28 January 2005</date>
14 <firstname>Havoc</firstname>
15 <surname>Pennington</surname>
17 <orgname>Red Hat, Inc.</orgname>
19 <email>hp@pobox.com</email>
24 <firstname>Anders</firstname>
25 <surname>Carlsson</surname>
27 <orgname>CodeFactory AB</orgname>
29 <email>andersca@codefactory.se</email>
34 <firstname>Alexander</firstname>
35 <surname>Larsson</surname>
37 <orgname>Red Hat, Inc.</orgname>
39 <email>alexl@redhat.com</email>
46 <sect1 id="introduction">
47 <title>Introduction</title>
49 D-BUS is a system for low-latency, low-overhead, easy to use
50 interprocess communication (IPC). In more detail:
54 D-BUS is <emphasis>low-latency</emphasis> because it is designed
55 to avoid round trips and allow asynchronous operation, much like
61 D-BUS is <emphasis>low-overhead</emphasis> because it uses a
62 binary protocol, and does not have to convert to and from a text
63 format such as XML. Because D-BUS is intended for potentially
64 high-resolution same-machine IPC, not primarily for Internet IPC,
65 this is an interesting optimization.
70 D-BUS is <emphasis>easy to use</emphasis> because it works in terms
71 of <firstterm>messages</firstterm> rather than byte streams, and
72 automatically handles a lot of the hard IPC issues. Also, the D-BUS
73 library is designed to be wrapped in a way that lets developers use
74 their framework's existing object/type system, rather than learning
75 a new one specifically for IPC.
82 The base D-BUS protocol is a one-to-one (peer-to-peer or client-server)
83 protocol, specified in <xref linkend="message-protocol"/>. That is, it is
84 a system for one application to talk to a single other
85 application. However, the primary intended application of the protocol is the
86 D-BUS <firstterm>message bus</firstterm>, specified in <xref
87 linkend="message-bus"/>. The message bus is a special application that
88 accepts connections from multiple other applications, and forwards
93 Uses of D-BUS include notification of system changes (notification of when
94 a camera is plugged in to a computer, or a new version of some software
95 has been installed), or desktop interoperability, for example a file
96 monitoring service or a configuration service.
100 D-BUS is designed for two specific use cases:
104 A "system bus" for notifications from the system to user sessions,
105 and to allow the system to request input from user sessions.
110 A "session bus" used to implement desktop environments such as
115 D-BUS is not intended to be a generic IPC system for any possible
116 application, and intentionally omits many features found in other
117 IPC systems for this reason. D-BUS may turn out to be useful
118 in unanticipated applications, but future versions of this
119 spec and the reference implementation probably will not
120 incorporate features that interfere with the core use cases.
125 <sect1 id="message-protocol">
126 <title>Message Protocol</title>
129 A <firstterm>message</firstterm> consists of a
130 <firstterm>header</firstterm> and a <firstterm>body</firstterm>. If you
131 think of a message as a package, the header is the address, and the body
132 contains the package contents. The message delivery system uses the header
133 information to figure out where to send the message and how to interpret
134 it; the recipient inteprets the body of the message.
138 The body of the message is made up of zero or more
139 <firstterm>arguments</firstterm>, which are typed values, such as an
140 integer or a byte array.
144 Both header and body use the same type system and format for
145 serializing data. Each type of value has a wire format.
146 Converting a value from some other representation into the wire
147 format is called <firstterm>marshaling</firstterm> and converting
148 it back from the wire format is <firstterm>unmarshaling</firstterm>.
151 <sect2 id="message-protocol-signatures">
152 <title>Type Signatures</title>
155 The D-BUS protocol does not include type tags in the marshaled data; a
156 block of marshaled values must have a known <firstterm>type
157 signature</firstterm>. The type signature is made up of <firstterm>type
158 codes</firstterm>. A type code is an ASCII character representing the
159 type of a value. Because ASCII characters are used, the type signature
160 will always form a valid ASCII string. A simple string compare
161 determines whether two type signatures are equivalent.
165 As a simple example, the type code for 32-bit integer (<literal>INT32</literal>) is
166 the ASCII character 'i'. So the signature for a block of values
167 containing a single <literal>INT32</literal> would be:
171 A block of values containing two <literal>INT32</literal> would have this signature:
178 All <firstterm>basic</firstterm> types work like
179 <literal>INT32</literal> in this example. To marshal and unmarshal
180 basic types, you simply read one value from the data
181 block corresponding to each type code in the signature.
182 In addition to basic types, there are three <firstterm>container</firstterm>
183 types: <literal>STRUCT</literal>, <literal>ARRAY</literal>, and <literal>VARIANT</literal>.
187 <literal>STRUCT</literal> has a type code, ASCII character 'r', but this type
188 code does not appear in signatures. Instead, ASCII characters
189 '(' and ')' are used to mark the beginning and end of the struct.
190 So for example, a struct containing two integers would have this
195 Structs can be nested, so for example a struct containing
196 an integer and another struct:
200 The value block storing that struct would contain three integers; the
201 type signature allows you to distinguish "(i(ii))" from "((ii)i)" or
206 The <literal>STRUCT</literal> type code 'r' is not currently used in the D-BUS protocol,
207 but is useful in code that implements the protocol. This type code
208 is specified to allow such code to interoperate in non-protocol contexts.
212 <literal>ARRAY</literal> has ASCII character 'a' as type code. The array type code must be
213 followed by a <firstterm>single complete type</firstterm>. The single
214 complete type following the array is the type of each array element. So
215 the simple example is:
219 which is an array of 32-bit integers. But an array can be of any type,
220 such as this array-of-struct-with-two-int32-fields:
224 Or this array of array of integer:
231 The phrase <firstterm>single complete type</firstterm> deserves some
232 definition. A single complete type is a basic type code, a variant type code,
233 an array with its element type, or a struct with its fields.
234 So the following signatures are not single complete types:
244 And the following signatures contain multiple complete types:
254 Note however that a single complete type may <emphasis>contain</emphasis>
255 multiple other single complete types.
259 <literal>VARIANT</literal> has ASCII character 'v' as its type code. A marshaled value of
260 type <literal>VARIANT</literal> will have the signature of a single complete type as part
261 of the <emphasis>value</emphasis>. This signature will be followed by a
262 marshaled value of that type.
266 A <literal>DICT_ENTRY</literal> works exactly like a struct, but rather
267 than parentheses it uses curly braces, and it has more restrictions.
268 The restrictions are: it occurs only as an array element type; and it
269 has exactly two single complete types inside the curly
270 braces. Implementations must not accept dict entries outside of arrays,
271 and must not accept dict entries with zero, one, or more than two
272 fields. A dict entry is always a key-value pair.
276 The first field in the <literal>DICT_ENTRY</literal> is always the key.
277 A message is considered corrupt if the same key occurs twice in the same
278 array of <literal>DICT_ENTRY</literal>. However, for performance reasons
279 implementations are not required to reject dicts with duplicate keys.
283 In most languages, an array of dict entry would be represented as a
284 map, hash table, or dict object.
288 The following table summarizes the D-BUS types.
293 <entry>Conventional Name</entry>
295 <entry>Description</entry>
300 <entry><literal>INVALID</literal></entry>
301 <entry>0 (ASCII NUL)</entry>
302 <entry>Not a valid type code, used to terminate signatures</entry>
304 <entry><literal>BYTE</literal></entry>
305 <entry>121 (ASCII 'y')</entry>
306 <entry>8-bit unsigned integer</entry>
308 <entry><literal>BOOLEAN</literal></entry>
309 <entry>98 (ASCII 'b')</entry>
310 <entry>Boolean value, 0 is <literal>FALSE</literal> and 1 is <literal>TRUE</literal>. Everything else is invalid.</entry>
312 <entry><literal>INT16</literal></entry>
313 <entry>110 (ASCII 'n')</entry>
314 <entry>16-bit signed integer</entry>
316 <entry><literal>UINT16</literal></entry>
317 <entry>113 (ASCII 'q')</entry>
318 <entry>16-bit unsigned integer</entry>
320 <entry><literal>INT32</literal></entry>
321 <entry>105 (ASCII 'i')</entry>
322 <entry>32-bit signed integer</entry>
324 <entry><literal>UINT32</literal></entry>
325 <entry>117 (ASCII 'u')</entry>
326 <entry>32-bit unsigned integer</entry>
328 <entry><literal>INT64</literal></entry>
329 <entry>120 (ASCII 'x')</entry>
330 <entry>64-bit signed integer</entry>
332 <entry><literal>UINT64</literal></entry>
333 <entry>116 (ASCII 't')</entry>
334 <entry>64-bit unsigned integer</entry>
336 <entry><literal>DOUBLE</literal></entry>
337 <entry>100 (ASCII 'd')</entry>
338 <entry>IEEE 754 double</entry>
340 <entry><literal>STRING</literal></entry>
341 <entry>115 (ASCII 's')</entry>
342 <entry>UTF-8 string (<emphasis>must</emphasis> be valid UTF-8). Must be nul terminated.</entry>
344 <entry><literal>OBJECT_PATH</literal></entry>
345 <entry>111 (ASCII 'o')</entry>
346 <entry>Name of an object instance</entry>
348 <entry><literal>SIGNATURE</literal></entry>
349 <entry>103 (ASCII 'g')</entry>
350 <entry>A type signature</entry>
352 <entry><literal>ARRAY</literal></entry>
353 <entry>97 (ASCII 'a')</entry>
356 <entry><literal>STRUCT</literal></entry>
357 <entry>114 (ASCII 'r'), 40 (ASCII '('), 41 (ASCII ')')</entry>
358 <entry>Struct</entry>
360 <entry><literal>VARIANT</literal></entry>
361 <entry>118 (ASCII 'v') </entry>
362 <entry>Variant type (the type of the value is part of the value itself)</entry>
364 <entry><literal>DICT_ENTRY</literal></entry>
365 <entry>101 (ASCII 'e'), 123 (ASCII '{'), 125 (ASCII '}') </entry>
366 <entry>Entry in a dict or map (array of key-value pairs)</entry>
375 <sect2 id="message-protocol-marshaling">
376 <title>Marshaling (Wire Format)</title>
379 Given a type signature, a block of bytes can be converted into typed
380 values. This section describes the format of the block of bytes. Byte
381 order and alignment issues are handled uniformly for all D-BUS types.
385 A block of bytes has an associated byte order. The byte order
386 has to be discovered in some way; for D-BUS messages, the
387 byte order is part of the message header as described in
388 <xref linkend="message-protocol-messages"/>. For now, assume
389 that the byte order is known to be either little endian or big
394 Each value in a block of bytes is aligned "naturally," for example
395 4-byte values are aligned to a 4-byte boundary, and 8-byte values to an
396 8-byte boundary. To properly align a value, <firstterm>alignment
397 padding</firstterm> may be necessary. The alignment padding must always
398 be the minimum required padding to properly align the following value;
399 and it must always be made up of nul bytes. The alignment padding must
400 not be left uninitialized (it can't contain garbage), and more padding
401 than required must not be used.
405 Given all this, the types are marshaled on the wire as follows:
410 <entry>Conventional Name</entry>
411 <entry>Encoding</entry>
412 <entry>Alignment</entry>
417 <entry><literal>INVALID</literal></entry>
418 <entry>Not applicable; cannot be marshaled.</entry>
421 <entry><literal>BYTE</literal></entry>
422 <entry>A single 8-bit byte.</entry>
425 <entry><literal>BOOLEAN</literal></entry>
426 <entry>As for <literal>UINT32</literal>, but only 0 and 1 are valid values.</entry>
429 <entry><literal>INT16</literal></entry>
430 <entry>16-bit signed integer in the message's byte order.</entry>
433 <entry><literal>UINT16</literal></entry>
434 <entry>16-bit unsigned integer in the message's byte order.</entry>
437 <entry><literal>INT32</literal></entry>
438 <entry>32-bit signed integer in the message's byte order.</entry>
441 <entry><literal>UINT32</literal></entry>
442 <entry>32-bit unsigned integer in the message's byte order.</entry>
445 <entry><literal>INT64</literal></entry>
446 <entry>64-bit signed integer in the message's byte order.</entry>
449 <entry><literal>UINT64</literal></entry>
450 <entry>64-bit unsigned integer in the message's byte order.</entry>
453 <entry><literal>DOUBLE</literal></entry>
454 <entry>64-bit IEEE 754 double in the message's byte order.</entry>
457 <entry><literal>STRING</literal></entry>
458 <entry>A <literal>UINT32</literal> indicating the string's
459 length in bytes excluding its terminating nul, followed by
460 string data of the given length, followed by a terminating nul
467 <entry><literal>OBJECT_PATH</literal></entry>
468 <entry>Exactly the same as <literal>STRING</literal> except the
469 content must be a valid object path (see below).
475 <entry><literal>SIGNATURE</literal></entry>
476 <entry>The same as <literal>STRING</literal> except the length is a single
477 byte (thus signatures have a maximum length of 255)
478 and the content must be a valid signature (see below).
484 <entry><literal>ARRAY</literal></entry>
486 A <literal>UINT32</literal> giving the length of the array data in bytes, followed by
487 alignment padding to the alignment boundary of the array element type,
488 followed by each array element. The array length is from the
489 end of the alignment padding to the end of the last element,
490 i.e. it does not include the padding after the length,
491 or any padding after the last element.
492 Arrays have a maximum length defined to be 2 to the 26th power or
493 67108864. Implementations must not send or accept arrays exceeding this
500 <entry><literal>STRUCT</literal></entry>
502 A struct must start on an 8-byte boundary regardless of the
503 type of the struct fields. The struct value consists of each
504 field marshaled in sequence starting from that 8-byte
511 <entry><literal>VARIANT</literal></entry>
513 A variant type has a marshaled <literal>SIGNATURE</literal>
514 followed by a marshaled value with the type
515 given in the signature.
516 Unlike a message signature, the variant signature
517 can contain only a single complete type.
518 So "i" is OK, "ii" is not.
521 1 (alignment of the signature)
524 <entry><literal>DICT_ENTRY</literal></entry>
537 <sect3 id="message-protocol-marshaling-object-path">
538 <title>Valid Object Paths</title>
541 An object path is a name used to refer to an object instance.
542 Conceptually, each participant in a D-BUS message exchange may have
543 any number of object instances (think of C++ or Java objects) and each
544 such instance will have a path. Like a filesystem, the object
545 instances in an application form a hierarchical tree.
549 The following rules define a valid object path. Implementations must
550 not send or accept messages with invalid object paths.
554 The path may be of any length.
559 The path must begin with an ASCII '/' (integer 47) character,
560 and must consist of elements separated by slash characters.
565 Each element must only contain the ASCII characters
571 No element may be the empty string.
576 Multiple '/' characters cannot occur in sequence.
581 A trailing '/' character is not allowed unless the
582 path is the root path (a single '/' character).
591 <sect3 id="message-protocol-marshaling-signature">
592 <title>Valid Signatures</title>
594 An implementation must not send or accept invalid signatures.
595 Valid signatures will conform to the following rules:
599 The signature ends with a nul byte.
604 The signature is a list of single complete types.
605 Arrays must have element types, and structs must
606 have both open and close parentheses.
611 Only type codes and open and close parentheses are
612 allowed in the signature. The <literal>STRUCT</literal> type code
613 is not allowed in signatures, because parentheses
619 The maximum depth of container type nesting is 32 array type
620 codes and 32 open parentheses. This implies that the maximum
621 total depth of recursion is 64, for an "array of array of array
622 of ... struct of struct of struct of ..." where there are 32
628 The maximum length of a signature is 255.
633 Signatures must be nul-terminated.
642 <sect2 id="message-protocol-messages">
643 <title>Message Format</title>
646 A message consists of a header and a body. The header is a block of
647 values with a fixed signature and meaning. The body is a separate block
648 of values, with a signature specified in the header.
652 The length of the header must be a multiple of 8, allowing the body to
653 begin on an 8-byte boundary when storing the entire message in a single
654 buffer. If the header does not naturally end on an 8-byte boundary
655 up to 7 bytes of nul-initialized alignment padding must be added.
659 The message body need not end on an 8-byte boundary.
663 The maximum length of a message, including header, header alignment padding,
664 and body is 2 to the 27th power or 134217728. Implementations must not
665 send or accept messages exceeding this size.
669 The signature of the header is:
673 Written out more readably, this is:
675 BYTE, BYTE, BYTE, BYTE, UINT32, UINT32, ARRAY of STRUCT of (BYTE,VARIANT)
680 These values have the following meanings:
686 <entry>Description</entry>
691 <entry>1st <literal>BYTE</literal></entry>
692 <entry>Endianness flag; ASCII 'l' for little-endian
693 or ASCII 'B' for big-endian. Both header and body are
694 in this endianness.</entry>
697 <entry>2nd <literal>BYTE</literal></entry>
698 <entry><firstterm>Message type</firstterm>. Unknown types MUST be ignored.
699 Currently-defined types are described below.
703 <entry>3rd <literal>BYTE</literal></entry>
704 <entry>Bitwise OR of flags. Unknown flags
705 MUST be ignored. Currently-defined flags are described below.
709 <entry>4th <literal>BYTE</literal></entry>
710 <entry>Major protocol version of the sending application. If
711 the major protocol version of the receiving application does not
712 match, the applications will not be able to communicate and the
713 D-BUS connection MUST be disconnected. The major protocol
714 version for this version of the specification is 0.
715 FIXME this field is stupid and pointless to put in
720 <entry>1st <literal>UINT32</literal></entry>
721 <entry>Length in bytes of the message body, starting
722 from the end of the header. The header ends after
723 its alignment padding to an 8-boundary.
727 <entry>2nd <literal>UINT32</literal></entry>
728 <entry>The serial of this message, used as a cookie
729 by the sender to identify the reply corresponding
734 <entry><literal>ARRAY</literal> of <literal>STRUCT</literal> of (<literal>BYTE</literal>,<literal>VARIANT</literal>)</entry>
735 <entry>An array of zero or more <firstterm>header
736 fields</firstterm> where the byte is the field code, and the
737 variant is the field value. The message type determines
738 which fields are required.
746 <firstterm>Message types</firstterm> that can appear in the second byte
752 <entry>Conventional name</entry>
753 <entry>Decimal value</entry>
754 <entry>Description</entry>
759 <entry><literal>INVALID</literal></entry>
761 <entry>This is an invalid type, if seen in a message
762 the connection should be dropped immediately.</entry>
765 <entry><literal>METHOD_CALL</literal></entry>
767 <entry>Method call.</entry>
770 <entry><literal>METHOD_RETURN</literal></entry>
772 <entry>Method reply with returned data.</entry>
775 <entry><literal>ERROR</literal></entry>
777 <entry>Error reply. If the first argument exists and is a
778 string, it is an error message.</entry>
781 <entry><literal>SIGNAL</literal></entry>
783 <entry>Signal emission.</entry>
790 Flags that can appear in the third byte of the header:
795 <entry>Conventional name</entry>
796 <entry>Hex value</entry>
797 <entry>Description</entry>
802 <entry><literal>NO_REPLY_EXPECTED</literal></entry>
804 <entry>This message does not expect method return replies or
805 error replies; the reply can be omitted as an
806 optimization. However, it is compliant with this specification
807 to return the reply despite this flag.</entry>
810 <entry><literal>NO_AUTO_START</literal></entry>
812 <entry>This message should not automatically launch an owner
813 for the destination name.
821 <sect3 id="message-protocol-header-fields">
822 <title>Header Fields</title>
825 The array at the end of the header contains <firstterm>header
826 fields</firstterm>, where each field is a 1-byte field code followed
827 by a field value. A header must contain the required header fields for
828 its message type, and zero or more of any optional header
829 fields. Future versions of this protocol specification may add new
830 fields. Implementations must ignore fields they do not
831 understand. Implementations must not invent their own header fields;
832 only changes to this specification may introduce new header fields.
836 Again, if an implementation sees a header field code that it does not
837 expect, it MUST ignore that field, as it will be part of a new
838 (but compatible) version of this specification. This also applies
839 to known header fields appearing in unexpected messages, for
840 example if a signal has a reply serial that should be ignored
841 even though it has no meaning as of this version of the spec.
845 However, implementations must not send or accept known header fields
846 with the wrong type stored in the field value. So for example
847 a message with an <literal>INTERFACE</literal> field of type <literal>UINT32</literal> would be considered
852 Here are the currently-defined header fields:
857 <entry>Conventional Name</entry>
858 <entry>Decimal Code</entry>
860 <entry>Required In</entry>
861 <entry>Description</entry>
866 <entry><literal>INVALID</literal></entry>
869 <entry>not allowed</entry>
870 <entry>Not a valid field name (error if it appears in a message)</entry>
873 <entry><literal>PATH</literal></entry>
875 <entry><literal>OBJECT_PATH</literal></entry>
876 <entry><literal>METHOD_CALL</literal>, <literal>SIGNAL</literal></entry>
877 <entry>The object to send a call to,
878 or the object a signal is emitted from.
882 <entry><literal>INTERFACE</literal></entry>
884 <entry><literal>STRING</literal></entry>
885 <entry><literal>SIGNAL</literal></entry>
887 The interface to invoke a method call on, or
888 that a signal is emitted from. Optional for
889 method calls, required for signals.
893 <entry><literal>MEMBER</literal></entry>
895 <entry><literal>STRING</literal></entry>
896 <entry><literal>METHOD_CALL</literal>, <literal>SIGNAL</literal></entry>
897 <entry>The member, either the method name or signal name.</entry>
900 <entry><literal>ERROR_NAME</literal></entry>
902 <entry><literal>STRING</literal></entry>
903 <entry><literal>ERROR</literal></entry>
904 <entry>The name of the error that occurred, for errors</entry>
907 <entry><literal>REPLY_SERIAL</literal></entry>
909 <entry><literal>UINT32</literal></entry>
910 <entry><literal>ERROR</literal>, <literal>METHOD_RETURN</literal></entry>
911 <entry>The serial number of the message this message is a reply
912 to. (The serial number is the second <literal>UINT32</literal> in the header.)</entry>
915 <entry><literal>DESTINATION</literal></entry>
917 <entry><literal>STRING</literal></entry>
918 <entry>optional</entry>
919 <entry>The name of the connection this message should be routed to.
920 Only used in combination with the message bus, see
921 <xref linkend="message-bus"/>.</entry>
924 <entry><literal>SENDER</literal></entry>
926 <entry><literal>STRING</literal></entry>
927 <entry>optional</entry>
928 <entry>Unique name of the sending connection.
929 The message bus fills in this field so it is reliable; the field is
930 only meaningful in combination with the message bus.</entry>
933 <entry><literal>SIGNATURE</literal></entry>
935 <entry><literal>SIGNATURE</literal></entry>
936 <entry>optional</entry>
937 <entry>The signature of the message body.
938 If omitted, it is assumed to be the
939 empty signature "" (i.e. the body must be 0-length).</entry>
948 <sect2 id="message-protocol-names">
949 <title>Valid Names</title>
951 The various names in D-BUS messages have some restrictions.
954 There is a <firstterm>maximum name length</firstterm>
955 of 255 which applies to bus names, interfaces, and members.
957 <sect3 id="message-protocol-names-interface">
958 <title>Interface names</title>
960 Interfaces have names with type <literal>STRING</literal>, meaning that
961 they must be valid UTF-8. However, there are also some
962 additional restrictions that apply to interface names
965 <listitem><para>They are composed of 1 or more elements separated by
966 a period ('.') character. All elements must contain at least
970 <listitem><para>Each element must only contain the ASCII characters
971 "[A-Z][a-z][0-9]_" and must not begin with a digit.
975 <listitem><para>They must contain at least one '.' (period)
976 character (and thus at least two elements).
979 <listitem><para>They must not begin with a '.' (period) character.</para></listitem>
980 <listitem><para>They must not exceed the maximum name length.</para></listitem>
984 <sect3 id="message-protocol-names-bus">
985 <title>Bus names</title>
987 Bus names have the same restrictions as interface names, with a
988 special exception for unique connection names. A unique name's first
989 element must start with a colon (':') character. After the colon, any
990 characters in "[A-Z][a-z][0-9]_" may appear. Elements after
991 the first must follow the usual rules, except that they may start with
992 a digit. Bus names not starting with a colon have none of these
993 exceptions and follow the same rules as interface names.
996 <sect3 id="message-protocol-names-member">
997 <title>Member names</title>
999 Member (i.e. method or signal) names:
1001 <listitem><para>Must only contain the ASCII characters
1002 "[A-Z][a-z][0-9]_" and may not begin with a
1003 digit.</para></listitem>
1004 <listitem><para>Must not contain the '.' (period) character.</para></listitem>
1005 <listitem><para>Must not exceed the maximum name length.</para></listitem>
1006 <listitem><para>Must be at least 1 byte in length.</para></listitem>
1010 <sect3 id="message-protocol-names-error">
1011 <title>Error names</title>
1013 Error names have the same restrictions as interface names.
1018 <sect2 id="message-protocol-types">
1019 <title>Message Types</title>
1021 Each of the message types (<literal>METHOD_CALL</literal>, <literal>METHOD_RETURN</literal>, <literal>ERROR</literal>, and
1022 <literal>SIGNAL</literal>) has its own expected usage conventions and header fields.
1023 This section describes these conventions.
1025 <sect3 id="message-protocol-types-method">
1026 <title>Method Calls</title>
1028 Some messages invoke an operation on a remote object. These are
1029 called method call messages and have the type tag <literal>METHOD_CALL</literal>. Such
1030 messages map naturally to methods on objects in a typical program.
1033 A method call message is expected to have a <literal>MEMBER</literal> header field
1034 indicating the name of the method. Optionally, the message has an
1035 <literal>INTERFACE</literal> field giving the interface the method is a part of. In the
1036 absence of an <literal>INTERFACE</literal> field, if two interfaces on the same object have
1037 a method with the same name, it is undefined which of the two methods
1038 will be invoked. Implementations may also choose to return an error in
1039 this ambiguous case. However, if a method name is unique
1040 implementations must not require an interface field.
1043 Method call messages also include a <literal>PATH</literal> field
1044 indicating the object to invoke the method on. If the call is passing
1045 through a message bus, the message will also have a
1046 <literal>DESTINATION</literal> field giving the name of the connection
1047 to receive the message.
1050 When an application handles a method call message, it is expected to
1051 return a reply. The reply is identified by a <literal>REPLY_SERIAL</literal> header field
1052 indicating the serial number of the <literal>METHOD_CALL</literal> being replied to. The
1053 reply can have one of two types; either <literal>METHOD_RETURN</literal> or <literal>ERROR</literal>.
1056 If the reply has type <literal>METHOD_RETURN</literal>, the arguments to the reply message
1057 are the return value(s) or "out parameters" of the method call.
1058 If the reply has type <literal>ERROR</literal>, then an "exception" has been thrown,
1059 and the call fails; no return value will be provided. It makes
1060 no sense to send multiple replies to the same method call.
1063 Even if a method call has no return values, a <literal>METHOD_RETURN</literal>
1064 reply is expected, so the caller will know the method
1065 was successfully processed.
1068 The <literal>METHOD_RETURN</literal> or <literal>ERROR</literal> reply message must have the <literal>REPLY_SERIAL</literal>
1072 If a <literal>METHOD_CALL</literal> message has the flag <literal>NO_REPLY_EXPECTED</literal>,
1073 then as an optimization the application receiving the method
1074 call may choose to omit the reply message (regardless of
1075 whether the reply would have been <literal>METHOD_RETURN</literal> or <literal>ERROR</literal>).
1076 However, it is also acceptable to ignore the <literal>NO_REPLY_EXPECTED</literal>
1077 flag and reply anyway.
1080 Unless a message has the flag <literal>NO_AUTO_START</literal>, if the
1081 destination name does not exist then a program to own the destination
1082 name will be started before the message is delivered. The message
1083 will be held until the new program is successfully started or has
1084 failed to start; in case of failure, an error will be returned. This
1085 flag is only relevant in the context of a message bus, it is ignored
1086 during one-to-one communication with no intermediate bus.
1088 <sect4 id="message-protocol-types-method-apis">
1089 <title>Mapping method calls to native APIs</title>
1091 APIs for D-BUS may map method calls to a method call in a specific
1092 programming language, such as C++, or may map a method call written
1093 in an IDL to a D-BUS message.
1096 In APIs of this nature, arguments to a method are often termed "in"
1097 (which implies sent in the <literal>METHOD_CALL</literal>), or "out" (which implies
1098 returned in the <literal>METHOD_RETURN</literal>). Some APIs such as CORBA also have
1099 "inout" arguments, which are both sent and received, i.e. the caller
1100 passes in a value which is modified. Mapped to D-BUS, an "inout"
1101 argument is equivalent to an "in" argument, followed by an "out"
1102 argument. You can't pass things "by reference" over the wire, so
1103 "inout" is purely an illusion of the in-process API.
1106 Given a method with zero or one return values, followed by zero or more
1107 arguments, where each argument may be "in", "out", or "inout", the
1108 caller constructs a message by appending each "in" or "inout" argument,
1109 in order. "out" arguments are not represented in the caller's message.
1112 The recipient constructs a reply by appending first the return value
1113 if any, then each "out" or "inout" argument, in order.
1114 "in" arguments are not represented in the reply message.
1117 Error replies are normally mapped to exceptions in languages that have
1121 This specification doesn't require anything of native API bindings;
1122 the preceding is only a suggested convention for consistency
1129 <sect3 id="message-protocol-types-signal">
1130 <title>Signal Emission</title>
1132 Unlike method calls, signal emissions have no replies.
1133 A signal emission is simply a single message of type <literal>SIGNAL</literal>.
1134 It must have three header fields: <literal>PATH</literal> giving the object
1135 the signal was emitted from, plus <literal>INTERFACE</literal> and <literal>MEMBER</literal> giving
1136 the fully-qualified name of the signal.
1140 <sect3 id="message-protocol-types-errors">
1141 <title>Errors</title>
1143 Messages of type <literal>ERROR</literal> are most commonly replies
1144 to a <literal>METHOD_CALL</literal>, but may be returned in reply
1145 to any kind of message. The message bus for example
1146 will return an <literal>ERROR</literal> in reply to a signal emission if
1147 the bus does not have enough memory to send the signal.
1150 An <literal>ERROR</literal> may have any arguments, but if the first
1151 argument is a <literal>STRING</literal>, it must be an error message.
1152 The error message may be logged or shown to the user
1157 <sect3 id="message-protocol-types-notation">
1158 <title>Notation in this document</title>
1160 This document uses a simple pseudo-IDL to describe particular method
1161 calls and signals. Here is an example of a method call:
1163 org.freedesktop.DBus.StartServiceByName (in STRING name, in UINT32 flags,
1164 out UINT32 resultcode)
1166 This means <literal>INTERFACE</literal> = org.freedesktop.DBus, <literal>MEMBER</literal> = StartServiceByName,
1167 <literal>METHOD_CALL</literal> arguments are <literal>STRING</literal> and <literal>UINT32</literal>, <literal>METHOD_RETURN</literal> argument
1168 is <literal>UINT32</literal>. Remember that the <literal>MEMBER</literal> field can't contain any '.' (period)
1169 characters so it's known that the last part of the name in
1170 the "IDL" is the member name.
1173 In C++ that might end up looking like this:
1175 unsigned int org::freedesktop::DBus::StartServiceByName (const char *name,
1176 unsigned int flags);
1178 or equally valid, the return value could be done as an argument:
1180 void org::freedesktop::DBus::StartServiceByName (const char *name,
1182 unsigned int *resultcode);
1184 It's really up to the API designer how they want to make
1185 this look. You could design an API where the namespace wasn't used
1186 in C++, using STL or Qt, using varargs, or whatever you wanted.
1189 Signals are written as follows:
1191 org.freedesktop.DBus.NameLost (STRING name)
1193 Signals don't specify "in" vs. "out" because only
1194 a single direction is possible.
1197 It isn't especially encouraged to use this lame pseudo-IDL in actual
1198 API implementations; you might use the native notation for the
1199 language you're using, or you might use COM or CORBA IDL, for example.
1206 <sect1 id="auth-protocol">
1207 <title>Authentication Protocol</title>
1209 Before the flow of messages begins, two applications must
1210 authenticate. A simple plain-text protocol is used for
1211 authentication; this protocol is a SASL profile, and maps fairly
1212 directly from the SASL specification. The message encoding is
1213 NOT used here, only plain text messages.
1216 In examples, "C:" and "S:" indicate lines sent by the client and
1217 server respectively.
1219 <sect2 id="auth-protocol-overview">
1220 <title>Protocol Overview</title>
1222 The protocol is a line-based protocol, where each line ends with
1223 \r\n. Each line begins with an all-caps ASCII command name containing
1224 only the character range [A-Z], a space, then any arguments for the
1225 command, then the \r\n ending the line. The protocol is
1226 case-sensitive. All bytes must be in the ASCII character set.
1228 Commands from the client to the server are as follows:
1231 <listitem><para>AUTH [mechanism] [initial-response]</para></listitem>
1232 <listitem><para>CANCEL</para></listitem>
1233 <listitem><para>BEGIN</para></listitem>
1234 <listitem><para>DATA <data in hex encoding></para></listitem>
1235 <listitem><para>ERROR [human-readable error explanation]</para></listitem>
1238 From server to client are as follows:
1241 <listitem><para>REJECTED <space-separated list of mechanism names></para></listitem>
1242 <listitem><para>OK</para></listitem>
1243 <listitem><para>DATA <data in hex encoding></para></listitem>
1244 <listitem><para>ERROR</para></listitem>
1248 <sect2 id="auth-nul-byte">
1249 <title>Special credentials-passing nul byte</title>
1251 Immediately after connecting to the server, the client must send a
1252 single nul byte. This byte may be accompanied by credentials
1253 information on some operating systems that use sendmsg() with
1254 SCM_CREDS or SCM_CREDENTIALS to pass credentials over UNIX domain
1255 sockets. However, the nul byte MUST be sent even on other kinds of
1256 socket, and even on operating systems that do not require a byte to be
1257 sent in order to transmit credentials. The text protocol described in
1258 this document begins after the single nul byte. If the first byte
1259 received from the client is not a nul byte, the server may disconnect
1263 A nul byte in any context other than the initial byte is an error;
1264 the protocol is ASCII-only.
1267 The credentials sent along with the nul byte may be used with the
1268 SASL mechanism EXTERNAL.
1271 <sect2 id="auth-command-auth">
1272 <title>AUTH command</title>
1274 If an AUTH command has no arguments, it is a request to list
1275 available mechanisms. The server SHOULD respond with a REJECTED
1276 command listing the mechanisms it understands.
1279 If an AUTH command specifies a mechanism, and the server supports
1280 said mechanism, the server SHOULD begin exchanging SASL
1281 challenge-response data with the client using DATA commands.
1284 If the server does not support the mechanism given in the AUTH
1285 command, it SHOULD send a REJECTED command listing the mechanisms
1289 If the [initial-response] argument is provided, it is intended for
1290 use with mechanisms that have no initial challenge (or an empty
1291 initial challenge), as if it were the argument to an initial DATA
1292 command. If the selected mechanism has an initial challenge, the
1293 server should reject authentication by sending REJECTED.
1296 If authentication succeeds after exchanging DATA commands,
1297 an OK command should be sent to the client.
1300 The first octet received by the client after the \r\n of the OK
1301 command MUST be the first octet of the authenticated/encrypted
1302 stream of D-BUS messages.
1305 The first octet received by the server after the \r\n of the BEGIN
1306 command from the client MUST be the first octet of the
1307 authenticated/encrypted stream of D-BUS messages.
1310 <sect2 id="auth-command-cancel">
1311 <title>CANCEL Command</title>
1313 At any time up to sending the BEGIN command, the client may send a
1314 CANCEL command. On receiving the CANCEL command, the server MUST
1315 send a REJECTED command and abort the current authentication
1319 <sect2 id="auth-command-data">
1320 <title>DATA Command</title>
1322 The DATA command may come from either client or server, and simply
1323 contains a hex-encoded block of data to be interpreted
1324 according to the SASL mechanism in use.
1327 Some SASL mechanisms support sending an "empty string";
1328 FIXME we need some way to do this.
1331 <sect2 id="auth-command-begin">
1332 <title>BEGIN Command</title>
1334 The BEGIN command acknowledges that the client has received an
1335 OK command from the server, and that the stream of messages
1339 The first octet received by the server after the \r\n of the BEGIN
1340 command from the client MUST be the first octet of the
1341 authenticated/encrypted stream of D-BUS messages.
1344 <sect2 id="auth-command-rejected">
1345 <title>REJECTED Command</title>
1347 The REJECTED command indicates that the current authentication
1348 exchange has failed, and further exchange of DATA is inappropriate.
1349 The client would normally try another mechanism, or try providing
1350 different responses to challenges.
1352 Optionally, the REJECTED command has a space-separated list of
1353 available auth mechanisms as arguments. If a server ever provides
1354 a list of supported mechanisms, it MUST provide the same list
1355 each time it sends a REJECTED message. Clients are free to
1356 ignore all lists received after the first.
1359 <sect2 id="auth-command-ok">
1360 <title>OK Command</title>
1362 The OK command indicates that the client has been authenticated,
1363 and that further communication will be a stream of D-BUS messages
1364 (optionally encrypted, as negotiated) rather than this protocol.
1367 The first octet received by the client after the \r\n of the OK
1368 command MUST be the first octet of the authenticated/encrypted
1369 stream of D-BUS messages.
1372 The client MUST respond to the OK command by sending a BEGIN
1373 command, followed by its stream of messages, or by disconnecting.
1374 The server MUST NOT accept additional commands using this protocol
1375 after the OK command has been sent.
1378 <sect2 id="auth-command-error">
1379 <title>ERROR Command</title>
1381 The ERROR command indicates that either server or client did not
1382 know a command, does not accept the given command in the current
1383 context, or did not understand the arguments to the command. This
1384 allows the protocol to be extended; a client or server can send a
1385 command present or permitted only in new protocol versions, and if
1386 an ERROR is received instead of an appropriate response, fall back
1387 to using some other technique.
1390 If an ERROR is sent, the server or client that sent the
1391 error MUST continue as if the command causing the ERROR had never been
1392 received. However, the the server or client receiving the error
1393 should try something other than whatever caused the error;
1394 if only canceling/rejecting the authentication.
1397 <sect2 id="auth-examples">
1398 <title>Authentication examples</title>
1402 <title>Example of successful magic cookie authentication</title>
1404 (MAGIC_COOKIE is a made up mechanism)
1406 C: AUTH MAGIC_COOKIE 3138363935333137393635383634
1412 <title>Example of finding out mechanisms then picking one</title>
1415 S: REJECTED KERBEROS_V4 SKEY
1416 C: AUTH SKEY 7ab83f32ee
1417 S: DATA 8799cabb2ea93e
1418 C: DATA 8ac876e8f68ee9809bfa876e6f9876g8fa8e76e98f
1424 <title>Example of client sends unknown command then falls back to regular auth</title>
1428 C: AUTH MAGIC_COOKIE 3736343435313230333039
1434 <title>Example of server doesn't support initial auth mechanism</title>
1436 C: AUTH MAGIC_COOKIE 3736343435313230333039
1437 S: REJECTED KERBEROS_V4 SKEY
1438 C: AUTH SKEY 7ab83f32ee
1439 S: DATA 8799cabb2ea93e
1440 C: DATA 8ac876e8f68ee9809bfa876e6f9876g8fa8e76e98f
1446 <title>Example of wrong password or the like followed by successful retry</title>
1448 C: AUTH MAGIC_COOKIE 3736343435313230333039
1449 S: REJECTED KERBEROS_V4 SKEY
1450 C: AUTH SKEY 7ab83f32ee
1451 S: DATA 8799cabb2ea93e
1452 C: DATA 8ac876e8f68ee9809bfa876e6f9876g8fa8e76e98f
1454 C: AUTH SKEY 7ab83f32ee
1455 S: DATA 8799cabb2ea93e
1456 C: DATA 8ac876e8f68ee9809bfa876e6f9876g8fa8e76e98f
1462 <title>Example of skey cancelled and restarted</title>
1464 C: AUTH MAGIC_COOKIE 3736343435313230333039
1465 S: REJECTED KERBEROS_V4 SKEY
1466 C: AUTH SKEY 7ab83f32ee
1467 S: DATA 8799cabb2ea93e
1470 C: AUTH SKEY 7ab83f32ee
1471 S: DATA 8799cabb2ea93e
1472 C: DATA 8ac876e8f68ee9809bfa876e6f9876g8fa8e76e98f
1479 <sect2 id="auth-states">
1480 <title>Authentication state diagrams</title>
1483 This section documents the auth protocol in terms of
1484 a state machine for the client and the server. This is
1485 probably the most robust way to implement the protocol.
1488 <sect3 id="auth-states-client">
1489 <title>Client states</title>
1492 To more precisely describe the interaction between the
1493 protocol state machine and the authentication mechanisms the
1494 following notation is used: MECH(CHALL) means that the
1495 server challenge CHALL was fed to the mechanism MECH, which
1501 CONTINUE(RESP) means continue the auth conversation
1502 and send RESP as the response to the server;
1508 OK(RESP) means that after sending RESP to the server
1509 the client side of the auth conversation is finished
1510 and the server should return "OK";
1516 ERROR means that CHALL was invalid and could not be
1522 Both RESP and CHALL may be empty.
1526 The Client starts by getting an initial response from the
1527 default mechanism and sends AUTH MECH RESP, or AUTH MECH if
1528 the mechanism did not provide an initial response. If the
1529 mechanism returns CONTINUE, the client starts in state
1530 <emphasis>WaitingForData</emphasis>, if the mechanism
1531 returns OK the client starts in state
1532 <emphasis>WaitingForOK</emphasis>.
1536 The client should keep track of available mechanisms and
1537 which it mechanisms it has already attempted. This list is
1538 used to decide which AUTH command to send. When the list is
1539 exhausted, the client should give up and close the
1544 <title><emphasis>WaitingForData</emphasis></title>
1552 MECH(CHALL) returns CONTINUE(RESP) → send
1554 <emphasis>WaitingForData</emphasis>
1558 MECH(CHALL) returns OK(RESP) → send DATA
1559 RESP, goto <emphasis>WaitingForOK</emphasis>
1563 MECH(CHALL) returns ERROR → send ERROR
1564 [msg], goto <emphasis>WaitingForData</emphasis>
1572 Receive REJECTED [mechs] →
1573 send AUTH [next mech], goto
1574 WaitingForData or <emphasis>WaitingForOK</emphasis>
1579 Receive ERROR → send
1581 <emphasis>WaitingForReject</emphasis>
1586 Receive OK → send
1587 BEGIN, terminate auth
1588 conversation, authenticated
1593 Receive anything else → send
1595 <emphasis>WaitingForData</emphasis>
1603 <title><emphasis>WaitingForOK</emphasis></title>
1608 Receive OK → send BEGIN, terminate auth
1609 conversation, <emphasis>authenticated</emphasis>
1614 Receive REJECT [mechs] → send AUTH [next mech],
1615 goto <emphasis>WaitingForData</emphasis> or
1616 <emphasis>WaitingForOK</emphasis>
1622 Receive DATA → send CANCEL, goto
1623 <emphasis>WaitingForReject</emphasis>
1629 Receive ERROR → send CANCEL, goto
1630 <emphasis>WaitingForReject</emphasis>
1636 Receive anything else → send ERROR, goto
1637 <emphasis>WaitingForOK</emphasis>
1645 <title><emphasis>WaitingForReject</emphasis></title>
1650 Receive REJECT [mechs] → send AUTH [next mech],
1651 goto <emphasis>WaitingForData</emphasis> or
1652 <emphasis>WaitingForOK</emphasis>
1658 Receive anything else → terminate auth
1659 conversation, disconnect
1668 <sect3 id="auth-states-server">
1669 <title>Server states</title>
1672 For the server MECH(RESP) means that the client response
1673 RESP was fed to the the mechanism MECH, which returns one of
1678 CONTINUE(CHALL) means continue the auth conversation and
1679 send CHALL as the challenge to the client;
1685 OK means that the client has been successfully
1692 REJECT means that the client failed to authenticate or
1693 there was an error in RESP.
1698 The server starts out in state
1699 <emphasis>WaitingForAuth</emphasis>. If the client is
1700 rejected too many times the server must disconnect the
1705 <title><emphasis>WaitingForAuth</emphasis></title>
1711 Receive AUTH → send REJECTED [mechs], goto
1712 <emphasis>WaitingForAuth</emphasis>
1718 Receive AUTH MECH RESP
1722 MECH not valid mechanism → send REJECTED
1724 <emphasis>WaitingForAuth</emphasis>
1728 MECH(RESP) returns CONTINUE(CHALL) → send
1730 <emphasis>WaitingForData</emphasis>
1734 MECH(RESP) returns OK → send OK, goto
1735 <emphasis>WaitingForBegin</emphasis>
1739 MECH(RESP) returns REJECT → send REJECTED
1741 <emphasis>WaitingForAuth</emphasis>
1749 Receive BEGIN → terminate
1750 auth conversation, disconnect
1756 Receive ERROR → send REJECTED [mechs], goto
1757 <emphasis>WaitingForAuth</emphasis>
1763 Receive anything else → send
1765 <emphasis>WaitingForAuth</emphasis>
1774 <title><emphasis>WaitingForData</emphasis></title>
1782 MECH(RESP) returns CONTINUE(CHALL) → send
1784 <emphasis>WaitingForData</emphasis>
1788 MECH(RESP) returns OK → send OK, goto
1789 <emphasis>WaitingForBegin</emphasis>
1793 MECH(RESP) returns REJECT → send REJECTED
1795 <emphasis>WaitingForAuth</emphasis>
1803 Receive BEGIN → terminate auth conversation,
1810 Receive CANCEL → send REJECTED [mechs], goto
1811 <emphasis>WaitingForAuth</emphasis>
1817 Receive ERROR → send REJECTED [mechs], goto
1818 <emphasis>WaitingForAuth</emphasis>
1824 Receive anything else → send ERROR, goto
1825 <emphasis>WaitingForData</emphasis>
1833 <title><emphasis>WaitingForBegin</emphasis></title>
1838 Receive BEGIN → terminate auth conversation,
1839 client authenticated
1845 Receive CANCEL → send REJECTED [mechs], goto
1846 <emphasis>WaitingForAuth</emphasis>
1852 Receive ERROR → send REJECTED [mechs], goto
1853 <emphasis>WaitingForAuth</emphasis>
1859 Receive anything else → send ERROR, goto
1860 <emphasis>WaitingForBegin</emphasis>
1870 <sect2 id="auth-mechanisms">
1871 <title>Authentication mechanisms</title>
1873 This section describes some new authentication mechanisms.
1874 D-BUS also allows any standard SASL mechanism of course.
1876 <sect3 id="auth-mechanisms-sha">
1877 <title>DBUS_COOKIE_SHA1</title>
1879 The DBUS_COOKIE_SHA1 mechanism is designed to establish that a client
1880 has the ability to read a private file owned by the user being
1881 authenticated. If the client can prove that it has access to a secret
1882 cookie stored in this file, then the client is authenticated.
1883 Thus the security of DBUS_COOKIE_SHA1 depends on a secure home
1887 Authentication proceeds as follows:
1891 The client sends the username it would like to authenticate
1897 The server sends the name of its "cookie context" (see below); a
1898 space character; the integer ID of the secret cookie the client
1899 must demonstrate knowledge of; a space character; then a
1900 hex-encoded randomly-generated challenge string.
1905 The client locates the cookie, and generates its own hex-encoded
1906 randomly-generated challenge string. The client then
1907 concatentates the server's hex-encoded challenge, a ":"
1908 character, its own hex-encoded challenge, another ":" character,
1909 and the hex-encoded cookie. It computes the SHA-1 hash of this
1910 composite string. It sends back to the server the client's
1911 hex-encoded challenge string, a space character, and the SHA-1
1917 The server generates the same concatenated string used by the
1918 client and computes its SHA-1 hash. It compares the hash with
1919 the hash received from the client; if the two hashes match, the
1920 client is authenticated.
1926 Each server has a "cookie context," which is a name that identifies a
1927 set of cookies that apply to that server. A sample context might be
1928 "org_freedesktop_session_bus". Context names must be valid ASCII,
1929 nonzero length, and may not contain the characters slash ("/"),
1930 backslash ("\"), space (" "), newline ("\n"), carriage return ("\r"),
1931 tab ("\t"), or period ("."). There is a default context,
1932 "org_freedesktop_global" that's used by servers that do not specify
1936 Cookies are stored in a user's home directory, in the directory
1937 <filename>~/.dbus-keyrings/</filename>. This directory must
1938 not be readable or writable by other users. If it is,
1939 clients and servers must ignore it. The directory
1940 contains cookie files named after the cookie context.
1943 A cookie file contains one cookie per line. Each line
1944 has three space-separated fields:
1948 The cookie ID number, which must be a non-negative integer and
1949 may not be used twice in the same file.
1954 The cookie's creation time, in UNIX seconds-since-the-epoch
1960 The cookie itself, a hex-encoded random block of bytes.
1966 Only server processes modify the cookie file.
1967 They must do so with this procedure:
1971 Create a lockfile name by appending ".lock" to the name of the
1972 cookie file. The server should attempt to create this file
1973 using <literal>O_CREAT | O_EXCL</literal>. If file creation
1974 fails, the lock fails. Servers should retry for a reasonable
1975 period of time, then they may choose to delete an existing lock
1976 to keep users from having to manually delete a stale
1977 lock. <footnote><para>Lockfiles are used instead of real file
1978 locking <literal>fcntl()</literal> because real locking
1979 implementations are still flaky on network
1980 filesystems.</para></footnote>
1985 Once the lockfile has been created, the server loads the cookie
1986 file. It should then delete any cookies that are old (the
1987 timeout can be fairly short), or more than a reasonable
1988 time in the future (so that cookies never accidentally
1989 become permanent, if the clock was set far into the future
1990 at some point). If no recent keys remain, the
1991 server may generate a new key.
1996 The pruned and possibly added-to cookie file
1997 must be resaved atomically (using a temporary
1998 file which is rename()'d).
2003 The lock must be dropped by deleting the lockfile.
2009 Clients need not lock the file in order to load it,
2010 because servers are required to save the file atomically.
2015 <sect1 id="addresses">
2016 <title>Server Addresses</title>
2018 Server addresses consist of a transport name followed by a colon, and
2019 then an optional, comma-separated list of keys and values in the form key=value.
2020 [FIXME how do you escape colon, comma, and semicolon in the values of the key=value pairs?]
2024 <programlisting>unix:path=/tmp/dbus-test</programlisting>
2025 Which is the address to a unix socket with the path /tmp/dbus-test.
2028 [FIXME clarify if attempting to connect to each is a requirement
2029 or just a suggestion]
2030 When connecting to a server, multiple server addresses can be
2031 separated by a semi-colon. The library will then try to connect
2032 to the first address and if that fails, it'll try to connect to
2033 the next one specified, and so forth. For example
2034 <programlisting>unix:path=/tmp/dbus-test;unix:path=/tmp/dbus-test2</programlisting>
2037 [FIXME we need to specify in detail each transport and its possible arguments]
2038 Current transports include: unix domain sockets (including
2039 abstract namespace on linux), TCP/IP, and a debug/testing transport using
2040 in-process pipes. Future possible transports include one that
2041 tunnels over X11 protocol.
2045 <sect1 id="standard-messages">
2046 <title>Standard One-to-One Messages</title>
2048 See <xref linkend="message-protocol-types-notation"/> for details on
2049 the notation used in this section.
2051 <sect2 id="standard-messages-ping">
2052 <title><literal>org.freedesktop.Peer.Ping</literal></title>
2055 org.freedesktop.Peer.Ping ()
2059 On receipt of the <literal>METHOD_CALL</literal>
2060 message <literal>org.freedesktop.Peer.Ping</literal>, an application
2061 should do nothing other than reply with a <literal>METHOD_RETURN</literal> as usual.
2067 <sect1 id="message-bus">
2068 <title>Message Bus Specification</title>
2069 <sect2 id="message-bus-overview">
2070 <title>Message Bus Overview</title>
2072 The message bus accepts connections from one or more applications.
2073 Once connected, applications can exchange messages with other
2074 applications that are also connected to the bus.
2077 In order to route messages among connections, the message bus keeps a
2078 mapping from names to connections. Each connection has one
2079 unique-for-the-lifetime-of-the-bus name automatically assigned.
2080 Applications may request additional names for a connection. Additional
2081 names are usually "well-known names" such as
2082 "org.freedesktop.TextEditor". When a name is bound to a connection,
2083 that connection is said to <firstterm>own</firstterm> the name.
2086 The bus itself owns a special name, <literal>org.freedesktop.DBus</literal>.
2087 This name routes messages to the bus, allowing applications to make
2088 administrative requests. For example, applications can ask the bus
2089 to assign a name to a connection.
2092 Each name may have <firstterm>queued owners</firstterm>. When an
2093 application requests a name for a connection and the name is already in
2094 use, the bus will optionally add the connection to a queue waiting for
2095 the name. If the current owner of the name disconnects or releases
2096 the name, the next connection in the queue will become the new owner.
2100 This feature causes the right thing to happen if you start two text
2101 editors for example; the first one may request "org.freedesktop.TextEditor",
2102 and the second will be queued as a possible owner of that name. When
2103 the first exits, the second will take over.
2107 Messages may have a <literal>DESTINATION</literal> field (see <xref
2108 linkend="message-protocol-header-fields"/>). If the
2109 <literal>DESTINATION</literal> field is present, it specifies a message
2110 recipient by name. Method calls and replies normally specify this field.
2114 Signals normally do not specify a destination; they are sent to all
2115 applications with <firstterm>message matching rules</firstterm> that
2120 When the message bus receives a method call, if the
2121 <literal>DESTINATION</literal> field is absent, the call is taken to be
2122 a standard one-to-one message and interpreted by the message bus
2123 itself. For example, sending an
2124 <literal>org.freedesktop.Peer.Ping</literal> message with no
2125 <literal>DESTINATION</literal> will cause the message bus itself to
2126 reply to the ping immediately; the message bus will not make this
2127 message visible to other applications.
2131 Continuing the <literal>org.freedesktop.Peer.Ping</literal> example, if
2132 the ping message were sent with a <literal>DESTINATION</literal> name of
2133 <literal>com.yoyodyne.Screensaver</literal>, then the ping would be
2134 forwarded, and the Yoyodyne Corporation screensaver application would be
2135 expected to reply to the ping.
2139 <sect2 id="message-bus-names">
2140 <title>Message Bus Names</title>
2142 Each connection has at least one name, assigned at connection time and
2143 returned in response to the
2144 <literal>org.freedesktop.DBus.Hello</literal> method call. This
2145 automatically-assigned name is called the connection's <firstterm>unique
2146 name</firstterm>. Unique names are never reused for two different
2147 connections to the same bus.
2150 Ownership of a unique name is a prerequisite for interaction with
2151 the message bus. It logically follows that the unique name is always
2152 the first name that an application comes to own, and the last
2153 one that it loses ownership of.
2156 Unique connection names must begin with the character ':' (ASCII colon
2157 character); bus names that are not unique names must not begin
2158 with this character. (The bus must reject any attempt by an application
2159 to manually request a name beginning with ':'.) This restriction
2160 categorically prevents "spoofing"; messages sent to a unique name
2161 will always go to the expected connection.
2164 When a connection is closed, all the names that it owns are deleted (or
2165 transferred to the next connection in the queue if any).
2168 A connection can request additional names to be associated with it using
2169 the <literal>org.freedesktop.DBus.RequestName</literal> message. <xref
2170 linkend="message-protocol-names-bus"/> describes the format of a valid
2174 <sect3 id="bus-messages-request-name">
2175 <title><literal>org.freedesktop.DBus.RequestName</literal></title>
2179 UINT32 RequestName (in STRING name, in UINT32 flags)
2186 <entry>Argument</entry>
2188 <entry>Description</entry>
2194 <entry>STRING</entry>
2195 <entry>Name to request</entry>
2199 <entry>UINT32</entry>
2200 <entry>Flags</entry>
2210 <entry>Argument</entry>
2212 <entry>Description</entry>
2218 <entry>UINT32</entry>
2219 <entry>Return value</entry>
2226 This method call should be sent to
2227 <literal>org.freedesktop.DBus</literal> and asks the message bus to
2228 assign the given name to the method caller. The flags argument
2229 contains any of the following values logically ORed together:
2235 <entry>Conventional Name</entry>
2236 <entry>Value</entry>
2237 <entry>Description</entry>
2242 <entry>DBUS_NAME_FLAG_PROHIBIT_REPLACEMENT</entry>
2245 If the application succeeds in becoming the owner of the specified name,
2246 then ownership of the name can't be transferred until the application
2247 disconnects. If this flag is not set, then any application trying to become
2248 the owner of the name will succeed and the previous owner will be
2249 sent a <literal>org.freedesktop.DBus.NameOwnerChanged</literal> signal.
2253 <entry>DBUS_NAME_FLAG_REPLACE_EXISTING</entry>
2256 Try to replace the current owner if there is one. If this
2257 flag is not set the application will only become the owner of
2258 the name if there is no current owner.
2265 The return code can be one of the following values:
2271 <entry>Conventional Name</entry>
2272 <entry>Value</entry>
2273 <entry>Description</entry>
2278 <entry>DBUS_REQUEST_NAME_REPLY_PRIMARY_OWNER</entry>
2279 <entry>1</entry> <entry>The caller is now the primary owner of
2280 the name, replacing any previous owner. Either the name had no
2281 owner before, or the caller specified
2282 DBUS_NAME_FLAG_REPLACE_EXISTING and the current owner did not
2283 specify DBUS_NAME_FLAG_PROHIBIT_REPLACEMENT.</entry>
2286 <entry>DBUS_REQUEST_NAME_REPLY_IN_QUEUE</entry>
2288 <entry>The name already had an owner, DBUS_NAME_FLAG_REPLACE_EXISTING was not specified, and the current owner specified DBUS_NAME_FLAG_PROHIBIT_REPLACEMENT.</entry>
2291 <entry>DBUS_REQUEST_NAME_REPLY_EXISTS</entry>
2293 <entry>The name already has an owner, and DBUS_NAME_FLAG_REPLACE_EXISTING was not specified.</entry>
2296 <entry>DBUS_REQUEST_NAME_REPLY_ALREADY_OWNER</entry>
2298 <entry>The application trying to request ownership of a name is already the owner of it.</entry>
2307 <sect2 id="message-bus-routing">
2308 <title>Message Bus Message Routing</title>
2313 <sect2 id="message-bus-starting-services">
2314 <title>Message Bus Starting Services</title>
2316 The message bus can start applications on behalf of other applications.
2317 In CORBA terms, this would be called <firstterm>activation</firstterm>.
2318 An application that can be started in this way is called a
2319 <firstterm>service</firstterm>.
2322 With D-BUS, starting a service is normally done by name. That is,
2323 applications ask the message bus to start some program that will own a
2324 well-known name, such as <literal>org.freedesktop.TextEditor</literal>.
2325 This implies a contract documented along with the name
2326 <literal>org.freedesktop.TextEditor</literal> for which objects
2327 the owner of that name will provide, and what interfaces those
2331 To find an executable corresponding to a particular name, the bus daemon
2332 looks for <firstterm>service description files</firstterm>. Service
2333 description files define a mapping from names to executables. Different
2334 kinds of message bus will look for these files in different places, see
2335 <xref linkend="message-bus-types"/>.
2338 [FIXME the file format should be much better specified than "similar to
2339 .desktop entries" esp. since desktop entries are already
2340 badly-specified. ;-)] Service description files have the ".service" file
2341 extension. The message bus will only load service description files
2342 ending with .service; all other files will be ignored. The file format
2343 is similar to that of <ulink
2344 url="http://www.freedesktop.org/standards/desktop-entry-spec/desktop-entry-spec.html">desktop
2345 entries</ulink>. All service description files must be in UTF-8
2346 encoding. To ensure that there will be no name collisions, service files
2347 must be namespaced using the same mechanism as messages and service
2351 <title>Example service description file</title>
2353 # Sample service description file
2355 Names=org.freedesktop.ConfigurationDatabase;org.gnome.GConf;
2356 Exec=/usr/libexec/gconfd-2
2361 When an application asks to start a service by name, the bus daemon tries to
2362 find a service that will own that name. It then tries to spawn the
2363 executable associated with it. If this fails, it will report an
2364 error. [FIXME what happens if two .service files offer the same service;
2365 what kind of error is reported, should we have a way for the client to
2369 The executable launched will have the environment variable
2370 <literal>DBUS_STARTER_ADDRESS</literal> set to the address of the
2371 message bus so it can connect and request the appropriate names.
2374 The executable being launched may want to know whether the message bus
2375 starting it is one of the well-known message buses (see <xref
2376 linkend="message-bus-types"/>). To facilitate this, the bus MUST also set
2377 the <literal>DBUS_STARTER_BUS_TYPE</literal> environment variable if it is one
2378 of the well-known buses. The currently-defined values for this variable
2379 are <literal>system</literal> for the systemwide message bus,
2380 and <literal>session</literal> for the per-login-session message
2381 bus. The new executable must still connect to the address given
2382 in <literal>DBUS_STARTER_ADDRESS</literal>, but may assume that the
2383 resulting connection is to the well-known bus.
2386 [FIXME there should be a timeout somewhere, either specified
2387 in the .service file, by the client, or just a global value
2388 and if the client being activated fails to connect within that
2389 timeout, an error should be sent back.]
2393 <sect2 id="message-bus-types">
2394 <title>Well-known Message Bus Instances</title>
2396 Two standard message bus instances are defined here, along with how
2397 to locate them and where their service files live.
2399 <sect3 id="message-bus-types-login">
2400 <title>Login session message bus</title>
2402 Each time a user logs in, a <firstterm>login session message
2403 bus</firstterm> may be started. All applications in the user's login
2404 session may interact with one another using this message bus.
2407 The address of the login session message bus is given
2408 in the <literal>DBUS_SESSION_BUS_ADDRESS</literal> environment
2409 variable. If that variable is not set, applications may
2410 also try to read the address from the X Window System root
2411 window property <literal>_DBUS_SESSION_BUS_ADDRESS</literal>.
2412 The root window property must have type <literal>STRING</literal>.
2413 The environment variable should have precedence over the
2414 root window property.
2417 [FIXME specify location of .service files, probably using
2418 DESKTOP_DIRS etc. from basedir specification, though login session
2419 bus is not really desktop-specific]
2422 <sect3 id="message-bus-types-system">
2423 <title>System message bus</title>
2425 A computer may have a <firstterm>system message bus</firstterm>,
2426 accessible to all applications on the system. This message bus may be
2427 used to broadcast system events, such as adding new hardware devices,
2428 changes in the printer queue, and so forth.
2431 The address of the system message bus is given
2432 in the <literal>DBUS_SYSTEM_BUS_ADDRESS</literal> environment
2433 variable. If that variable is not set, applications should try
2434 to connect to the well-known address
2435 <literal>unix:path=/var/run/dbus/system_bus_socket</literal>.
2438 The D-BUS reference implementation actually honors the
2439 <literal>$(localstatedir)</literal> configure option
2440 for this address, on both client and server side.
2445 [FIXME specify location of system bus .service files]
2450 <sect2 id="message-bus-messages">
2451 <title>Message Bus Messages</title>
2453 The special message bus name <literal>org.freedesktop.DBus</literal>
2454 responds to a number of additional messages.
2457 <sect3 id="bus-messages-hello">
2458 <title><literal>org.freedesktop.DBus.Hello</literal></title>
2469 <entry>Argument</entry>
2471 <entry>Description</entry>
2477 <entry>STRING</entry>
2478 <entry>Unique name assigned to the connection</entry>
2485 Before an application is able to send messages to other applications
2486 it must send the <literal>org.freedesktop.DBus.Hello</literal> message
2487 to the message bus to obtain a unique name. If an application without
2488 a unique name tries to send a message to another application, or a
2489 message to the message bus itself that isn't the
2490 <literal>org.freedesktop.DBus.Hello</literal> message, it will be
2491 disconnected from the bus.
2494 There is no corresponding "disconnect" request; if a client wishes to
2495 disconnect from the bus, it simply closes the socket (or other
2496 communication channel).
2499 <sect3 id="bus-messages-list-names">
2500 <title><literal>org.freedesktop.DBus.ListNames</literal></title>
2504 ARRAY of STRING ListNames ()
2511 <entry>Argument</entry>
2513 <entry>Description</entry>
2519 <entry>ARRAY of STRING</entry>
2520 <entry>Array of strings where each string is a bus name</entry>
2527 Returns a list of all currently-owned names on the bus.
2530 <sect3 id="bus-messages-name-exists">
2531 <title><literal>org.freedesktop.DBus.NameHasOwner</literal></title>
2535 BOOLEAN NameHasOwner (in STRING name)
2542 <entry>Argument</entry>
2544 <entry>Description</entry>
2550 <entry>STRING</entry>
2551 <entry>Name to check</entry>
2561 <entry>Argument</entry>
2563 <entry>Description</entry>
2569 <entry>BOOLEAN</entry>
2570 <entry>Return value, true if the name exists</entry>
2577 Checks if the specified name exists (currently has an owner).
2581 <sect3 id="bus-messages-name-owner-changed">
2582 <title><literal>org.freedesktop.DBus.NameOwnerChanged</literal></title>
2586 NameOwnerChanged (STRING name, STRING old_owner, STRING new_owner)
2593 <entry>Argument</entry>
2595 <entry>Description</entry>
2601 <entry>STRING</entry>
2602 <entry>Name with a new owner</entry>
2606 <entry>STRING</entry>
2607 <entry>Old owner or empty string if none</entry>
2611 <entry>STRING</entry>
2612 <entry>New owner or empty string if none</entry>
2619 This signal indicates that the owner of a name has changed.
2620 It's also the signal to use to detect the appearance of
2621 new names on the bus.
2624 <sect3 id="bus-messages-name-lost">
2625 <title><literal>org.freedesktop.DBus.NameLost</literal></title>
2629 NameLost (STRING name)
2636 <entry>Argument</entry>
2638 <entry>Description</entry>
2644 <entry>STRING</entry>
2645 <entry>Name which was lost</entry>
2652 This signal is sent to a specific application when it loses
2653 ownership of a name.
2657 <sect3 id="bus-messages-name-acquired">
2658 <title><literal>org.freedesktop.DBus.NameAcquired</literal></title>
2662 NameAcquired (STRING name)
2669 <entry>Argument</entry>
2671 <entry>Description</entry>
2677 <entry>STRING</entry>
2678 <entry>Name which was acquired</entry>
2685 This signal is sent to a specific application when it gains
2686 ownership of a name.
2690 <sect3 id="bus-messages-start-service-by-name">
2691 <title><literal>org.freedesktop.DBus.StartServiceByName</literal></title>
2695 UINT32 StartServiceByName (in STRING name, in UINT32 flags)
2702 <entry>Argument</entry>
2704 <entry>Description</entry>
2710 <entry>STRING</entry>
2711 <entry>Name of the service to start</entry>
2715 <entry>UINT32</entry>
2716 <entry>Flags (currently not used)</entry>
2726 <entry>Argument</entry>
2728 <entry>Description</entry>
2734 <entry>UINT32</entry>
2735 <entry>Return value</entry>
2740 Tries to launch the executable associated with a name. For more information, see <xref linkend="message-bus-starting-services"/>.
2744 The return value can be one of the following values:
2749 <entry>Identifier</entry>
2750 <entry>Value</entry>
2751 <entry>Description</entry>
2756 <entry>DBUS_START_REPLY_SUCCESS</entry>
2758 <entry>The service was successfully started.</entry>
2761 <entry>DBUS_START_REPLY_ALREADY_RUNNING</entry>
2763 <entry>A connection already owns the given name.</entry>
2772 <sect3 id="bus-messages-get-name-owner">
2773 <title><literal>org.freedesktop.DBus.GetNameOwner</literal></title>
2777 STRING GetNameOwner (in STRING name)
2784 <entry>Argument</entry>
2786 <entry>Description</entry>
2792 <entry>STRING</entry>
2793 <entry>Name to get the owner of</entry>
2803 <entry>Argument</entry>
2805 <entry>Description</entry>
2811 <entry>STRING</entry>
2812 <entry>Return value, a unique connection name</entry>
2817 Returns the unique connection name of the primary owner of the name
2818 given. If the requested name doesn't have an owner, returns a
2819 <literal>org.freedesktop.DBus.Error.NameHasNoOwner</literal> error.
2823 <sect3 id="bus-messages-get-connection-unix-user">
2824 <title><literal>org.freedesktop.DBus.GetConnectionUnixUser</literal></title>
2828 UINT32 GetConnectionUnixUser (in STRING connection_name)
2835 <entry>Argument</entry>
2837 <entry>Description</entry>
2843 <entry>STRING</entry>
2844 <entry>Name of the connection to query</entry>
2854 <entry>Argument</entry>
2856 <entry>Description</entry>
2862 <entry>UINT32</entry>
2863 <entry>unix user id</entry>
2868 Returns the unix uid of the process connected to the server. If unable to
2869 determine it, a <literal>org.freedesktop.DBus.Error.Failed</literal>
2878 <appendix id="implementation-notes">
2879 <title>Implementation notes</title>
2880 <sect1 id="implementation-notes-subsection">
2888 <glossary><title>Glossary</title>
2890 This glossary defines some of the terms used in this specification.
2893 <glossentry id="term-bus-name"><glossterm>Bus Name</glossterm>
2896 The message bus maintains an association between names and
2897 connections. (Normally, there's one connection per application.) A
2898 bus name is simply an identifier used to locate connections. For
2899 example, the hypothetical <literal>com.yoyodyne.Screensaver</literal>
2900 name might be used to send a message to a screensaver from Yoyodyne
2901 Corporation. An application is said to <firstterm>own</firstterm> a
2902 name if the message bus has associated the application's connection
2903 with the name. Names may also have <firstterm>queued
2904 owners</firstterm> (see <xref linkend="term-queued-owner"/>).
2905 The bus assigns a unique name to each connection,
2906 see <xref linkend="term-unique-name"/>. Other names
2907 can be thought of as "well-known names" and are
2908 used to find applications that offer specific functionality.
2913 <glossentry id="term-message"><glossterm>Message</glossterm>
2916 A message is the atomic unit of communication via the D-BUS
2917 protocol. It consists of a <firstterm>header</firstterm> and a
2918 <firstterm>body</firstterm>; the body is made up of
2919 <firstterm>arguments</firstterm>.
2924 <glossentry id="term-message-bus"><glossterm>Message Bus</glossterm>
2927 The message bus is a special application that forwards
2928 or routes messages between a group of applications
2929 connected to the message bus. It also manages
2930 <firstterm>names</firstterm> used for routing
2936 <glossentry id="term-name"><glossterm>Name</glossterm>
2939 See <xref linkend="term-bus-name"/>. "Name" may
2940 also be used to refer to some of the other names
2941 in D-BUS, such as interface names.
2946 <glossentry id="namespace"><glossterm>Namespace</glossterm>
2949 Used to prevent collisions when defining new interfaces or bus
2950 names. The convention used is the same one Java uses for defining
2951 classes: a reversed domain name.
2956 <glossentry id="term-object"><glossterm>Object</glossterm>
2959 Each application contains <firstterm>objects</firstterm>, which have
2960 <firstterm>interfaces</firstterm> and
2961 <firstterm>methods</firstterm>. Objects are referred to by a name,
2962 called a <firstterm>path</firstterm>.
2967 <glossentry id="one-to-one"><glossterm>One-to-One</glossterm>
2970 An application talking directly to another application, without going
2971 through a message bus. One-to-one connections may be "peer to peer" or
2972 "client to server." The D-BUS protocol has no concept of client
2973 vs. server after a connection has authenticated; the flow of messages
2974 is symmetrical (full duplex).
2979 <glossentry id="term-path"><glossterm>Path</glossterm>
2982 Object references (object names) in D-BUS are organized into a
2983 filesystem-style hierarchy, so each object is named by a path. As in
2984 LDAP, there's no difference between "files" and "directories"; a path
2985 can refer to an object, while still having child objects below it.
2990 <glossentry id="term-queued-owner"><glossterm>Queued Name Owner</glossterm>
2993 Each bus name has a primary owner; messages sent to the name go to the
2994 primary owner. However, certain names also maintain a queue of
2995 secondary owners "waiting in the wings." If the primary owner releases
2996 the name, then the first secondary owner in the queue automatically
2997 becomes the new owner of the name.
3002 <glossentry id="term-service"><glossterm>Service</glossterm>
3005 A service is an executable that can be launched by the bus daemon.
3006 Services normally guarantee some particular features, for example they
3007 may guarantee that they will request a specific name such as
3008 "org.freedesktop.Screensaver", have a singleton object
3009 "/org/freedesktop/Application", and that object will implement the
3010 interface "org.freedesktop.ScreensaverControl".
3015 <glossentry id="term-service-description-files"><glossterm>Service Description Files</glossterm>
3018 ".service files" tell the bus about service applications that can be
3019 launched (see <xref linkend="term-service"/>). Most importantly they
3020 provide a mapping from bus names to services that will request those
3021 names when they start up.
3026 <glossentry id="term-unique-name"><glossterm>Unique Connection Name</glossterm>
3029 The special name automatically assigned to each connection by the
3030 message bus. This name will never change owner, and will be unique
3031 (never reused during the lifetime of the message bus).
3032 It will begin with a ':' character.