4 - change .service files to have Names=list rather than Name=string
6 - How we will handle DCOP needs sorting out. Among other things, we
7 need to check that service and service-ownership semantics map to DCOP
10 - Activation needs some careful additional thinking-through.
12 - Audit @todo and FIXME for security issues
14 - The convenience functions in dbus-bus.h should perhaps have
15 the signatures that they would have if they were autogenerated
16 stubs. e.g. the acquire service function. We should also evaluate
17 which of these functions to include, in light of the fact that
18 GLib/Qt native stubs will probably also exist.
20 - the "break loader" and valid/invalid message tests are all disabled;
21 they need to be fixed and re-enabled with the new message args stuff.
22 I think I want to drop the .message files thing and just have code
23 that generates messages, more like the tests for
24 dbus-marshal-recursive.c
26 - modify the auth protocol to also support other initial-handshake
29 Perhaps the auth protocol should be able to negotiate a protocol
30 version to the least-common-denominator between client and server?
31 Though in practice ever using this feature would be pretty tough,
32 since protocol probably modifies the API. But we could have it there
35 - need to define bus behavior if you send a message to
36 yourself; is it an error, or allowed? If allowed,
37 we need to have a test for it in the test suite.
39 - add string array support back to append_args()
41 - validate dict entry number of fields
43 - just before 1.0, try a HAVE_INT64=0 build and be sure it runs
45 - the spec and implementation should probably require dict keys
48 Important for 1.0 GLib Bindings
51 - finish dbus-glib-tool support for adding introspection
52 data to GObject and autoexporting GObject using same
54 - the GLib bindings varargs take DBUS_TYPE_WHATEVER and
55 return stuff allocated with dbus_malloc(); should this
56 be made more "G" at some expense in code duplication?
57 You also still have to use some D-BUS functions such as
58 dbus_message_get_args() which takes a DBusError.
59 Probably we need to either fully encapsulate and hide
60 dbus/dbus.h, or encapsulate it slightly less e.g. no
61 GError. Or maybe it's as simple as "never return dbus_malloc()
62 memory" and just fully encapsulate the get_args() type of
65 - dbus_gproxy_connect_signal() has to take a signature for the signal
66 so it can figure out how to invoke the callback, or we have to rely
67 on having introspection data.
69 - DBusGProxy doesn't emit "destroy" when it should
74 - add dbus_message_has_path(), maybe has_member/interface
76 - connection_open/connection_disconnect lacks symmetry, open/close
79 - protocol version in each message is pretty silly
84 - The message bus internal code still says "service" for
85 "name", "base service" for "unique name", "activate" for
86 "start"; would be nice to clean up.
88 - Property list feature on message bus (list of properties associated
89 with a connection). May also include message matching rules
90 that involve the properties of the source or destination
93 - Disconnecting the remote end on invalid UTF-8 is probably not a good
94 idea. The definition of "valid" is slightly fuzzy. I think it might
95 be better to just silently "fix" the UTF-8, or perhaps return an error.
97 Owen says we should only validate the UTF-8 on dbus_message_get_string()
98 (changing get_string to have an error return, and allowing a type error
101 - assorted _-prefixed symbols in libdbus aren't actually used by
102 libdbus, only by the message bus. These bloat up the library
103 size. Not sure how to fix, really.
105 - build and install the Doxygen manual in Makefile when --enable-docs
107 - if you send the same message to multiple connections, the serial number
108 will only be right for one of them. Probably need to just write() the serial
109 number, rather than putting it in the DBusMessage, or something.
111 - perhaps the bus driver should have properties that reflect attributes
112 of the session, such as hostname, architecture, operating system,
113 etc. Could be useful for code that wants to special-case behavior
114 for a particular host or class of hosts, for example.
116 - currently the security policy stuff for messages to/from
117 the bus driver is kind of strange; basically it's hardcoded that
118 you can always talk to the driver, but the default config file
119 has rules for it anyway, or something. it's conceptually
120 screwy at the moment.
122 - when making a method call, if the call serial were globally unique,
123 we could forward the call serial along with any method calls made
124 as a result of the first method call, and allow reentrancy that was
125 strictly part of the call stack of said method call. But I don't
126 really see how to do this without making the user pass around the
127 call serial to all method calls all the time, or disallowing
130 If done post 1.0 will probably be an optional/ugly-API type
133 - I don't want to introduce DBusObject, but refcounting and object
134 data could still be factored out into an internal "base class"
137 - document the auth protocol as a set of states and transitions, and
138 then reimplement it in those terms
140 - recursive dispatch, see dbus_connection_dispatch()
142 - do we need per-display activation; if so I'd like to do this by setting a
143 "display ID" property on screen 0, with a GUID, and keying activation by
144 said GUID. Otherwise you get all kinds of unrobust
145 string/hostname-based mess. per-screen is then done by appending screen number
146 to the display. If displays have a deterministic ID like this, you can
147 do per-display by simply including GUID in the service name.
149 - optimization and profiling!