4 - the spec should say "the protocol will not be changed incompatibly after Month DD, YYYY"
6 Important for 1.0 GLib Bindings
9 - Test point-to-point mode
11 - Add support for getting sender
13 - format_version in the object info doesn't look like it's handled correctly. The creator
14 of the object info should specify some fixed number per struct version; the library
15 should handle only specific numbers it knows about. There's no assumption that all
16 numbers >= the given one are compatible. The idea is that new versions of the lib
17 can offer totally different object info structs, but old versions
20 Important for 1.0 Python bindings
25 - Fix removing of signals from the match tree
27 - Fix refcounting and userdata lifecycles
29 - Write a generic mainloop
34 - protocol version in each message is pretty silly
39 - revamp dbus-launch a bit,
40 see http://lists.freedesktop.org/archives/dbus/2006-October/005906.html
43 - clean up the creds issue on *BSD's in dbus/dbus-sysdeps-unix.c.
44 They should work as is but we need to rearange it to make it
45 clearer which method is being used. configure.in should
46 be fixed up to make that decition.
48 - _dbus_connection_unref_unlocked() is essentially always broken because
49 the connection finalizer calls non-unlocked functions. One fix is to make
50 the finalizer run with the lock held, but since it calls out to the app that may
51 be pretty broken. More likely all the uses of unref_unlocked are just wrong.
53 - if the GUID is obtained only during authentication, not in the address,
54 we could still share the connection
56 - Allow a dbus_g_proxy_to_string()/g_object_to_string() that
57 would convert the proxy to an "IOR" and dbus_g_proxy_from_string()
58 that would decode; using these, dbus-glib users could avoid
59 DBusConnection entirely. Of course the same applies to other kinds
60 of binding. This would use dbus_connection_open()'s connection-sharing
61 feature to avoid massive proliferation of connections.
63 - DBusWatchList/TimeoutList duplicate a lot of code, as do
64 protected_change_watch/protected_change_timeout in dbus-connection.c
65 and dbus-server.c. This could all be mopped up, cut-and-paste
66 fixed, code size reduced.
68 - change .service files to allow Names=list in addition to Name=string
70 - The message bus internal code still says "service" for
71 "name", "base service" for "unique name", "activate" for
72 "start"; would be nice to clean up.
74 - Property list feature on message bus (list of properties associated
75 with a connection). May also include message matching rules
76 that involve the properties of the source or destination
79 - Disconnecting the remote end on invalid UTF-8 is probably not a good
80 idea. The definition of "valid" is slightly fuzzy. I think it might
81 be better to just silently "fix" the UTF-8, or perhaps return an error.
83 - build and install the Doxygen manual in Makefile when --enable-docs
85 - if you send the same message to multiple connections, the serial number
86 will only be right for one of them. Probably need to just write() the serial
87 number, rather than putting it in the DBusMessage, or something.
89 - perhaps the bus driver should have properties that reflect attributes
90 of the session, such as hostname, architecture, operating system,
91 etc. Could be useful for code that wants to special-case behavior
92 for a particular host or class of hosts, for example.
94 - currently the security policy stuff for messages to/from
95 the bus driver is kind of strange; basically it's hardcoded that
96 you can always talk to the driver, but the default config file
97 has rules for it anyway, or something. it's conceptually
100 - when making a method call, if the call serial were globally unique,
101 we could forward the call serial along with any method calls made
102 as a result of the first method call, and allow reentrancy that was
103 strictly part of the call stack of said method call. But I don't
104 really see how to do this without making the user pass around the
105 call serial to all method calls all the time, or disallowing
108 If done post 1.0 will probably be an optional/ugly-API type
111 - I don't want to introduce DBusObject, but refcounting and object
112 data could still be factored out into an internal "base class"
115 - Keep convenience wrappers in sync with bus methods
117 - document the auth protocol as a set of states and transitions, and
118 then reimplement it in those terms
120 - recursive dispatch, see dbus_connection_dispatch()
122 - do we need per-display activation; if so I'd like to do this by setting a
123 "display ID" property on screen 0, with a GUID, and keying activation by
124 said GUID. Otherwise you get all kinds of unrobust
125 string/hostname-based mess. per-screen is then done by appending screen number
126 to the display. If displays have a deterministic ID like this, you can
127 do per-display by simply including GUID in the service name.
129 - optimization and profiling!
131 - Match rules aren't in the spec (probably a lot of methods on the bus
134 - the "break loader" and valid/invalid message tests are all disabled;
135 they need to be fixed and re-enabled with the new message args stuff.
136 I think I want to drop the .message files thing and just have code
137 that generates messages, more like the tests for
138 dbus-marshal-recursive.c (this is mostly done now, just needs some
141 - just before 1.0, try a HAVE_INT64=0 build and be sure it runs
143 - Windows port needs recursive mutexes
148 - look into supporting the concept of a "connection" generically
149 (what does this TODO item mean?)
151 - test/name-test should be named test/with-bus or something like that