4 - Remove all deprecated functions
6 - Audit @todo and FIXME for security issues
8 - the "break loader" and valid/invalid message tests are all disabled;
9 they need to be fixed and re-enabled with the new message args stuff.
10 I think I want to drop the .message files thing and just have code
11 that generates messages, more like the tests for
12 dbus-marshal-recursive.c (this is mostly done now, just needs some
15 - need to define bus behavior if you send a message to
16 yourself; is it an error, or allowed? If allowed,
17 we need to have a test for it in the test suite.
19 - just before 1.0, try a HAVE_INT64=0 build and be sure it runs
21 - dbus-pending-call.c has some API and thread safety issues to review
23 - Add test harness for selinux allow/deny cf. this message
24 http://lists.freedesktop.org/archives/dbus/2005-April/002506.html
26 - Add a test case for handling the Ping message
28 Important for 1.0 GLib Bindings
31 - Test point-to-point mode
33 - Add support for getting sender
38 - protocol version in each message is pretty silly
43 - if the GUID is obtained only during authentication, not in the address,
44 we could still share the connection
46 - Allow a dbus_g_proxy_to_string()/g_object_to_string() that
47 would convert the proxy to an "IOR" and dbus_g_proxy_from_string()
48 that would decode; using these, dbus-glib users could avoid
49 DBusConnection entirely. Of course the same applies to other kinds
50 of binding. This would use dbus_connection_open()'s connection-sharing
51 feature to avoid massive proliferation of connections.
53 - DBusWatchList/TimeoutList duplicate a lot of code, as do
54 protected_change_watch/protected_change_timeout in dbus-connection.c
55 and dbus-server.c. This could all be mopped up, cut-and-paste
56 fixed, code size reduced.
58 - change .service files to allow Names=list in addition to Name=string
60 - The message bus internal code still says "service" for
61 "name", "base service" for "unique name", "activate" for
62 "start"; would be nice to clean up.
64 - Property list feature on message bus (list of properties associated
65 with a connection). May also include message matching rules
66 that involve the properties of the source or destination
69 - Disconnecting the remote end on invalid UTF-8 is probably not a good
70 idea. The definition of "valid" is slightly fuzzy. I think it might
71 be better to just silently "fix" the UTF-8, or perhaps return an error.
73 - build and install the Doxygen manual in Makefile when --enable-docs
75 - if you send the same message to multiple connections, the serial number
76 will only be right for one of them. Probably need to just write() the serial
77 number, rather than putting it in the DBusMessage, or something.
79 - perhaps the bus driver should have properties that reflect attributes
80 of the session, such as hostname, architecture, operating system,
81 etc. Could be useful for code that wants to special-case behavior
82 for a particular host or class of hosts, for example.
84 - currently the security policy stuff for messages to/from
85 the bus driver is kind of strange; basically it's hardcoded that
86 you can always talk to the driver, but the default config file
87 has rules for it anyway, or something. it's conceptually
90 - when making a method call, if the call serial were globally unique,
91 we could forward the call serial along with any method calls made
92 as a result of the first method call, and allow reentrancy that was
93 strictly part of the call stack of said method call. But I don't
94 really see how to do this without making the user pass around the
95 call serial to all method calls all the time, or disallowing
98 If done post 1.0 will probably be an optional/ugly-API type
101 - I don't want to introduce DBusObject, but refcounting and object
102 data could still be factored out into an internal "base class"
105 - document the auth protocol as a set of states and transitions, and
106 then reimplement it in those terms
108 - recursive dispatch, see dbus_connection_dispatch()
110 - do we need per-display activation; if so I'd like to do this by setting a
111 "display ID" property on screen 0, with a GUID, and keying activation by
112 said GUID. Otherwise you get all kinds of unrobust
113 string/hostname-based mess. per-screen is then done by appending screen number
114 to the display. If displays have a deterministic ID like this, you can
115 do per-display by simply including GUID in the service name.
117 - optimization and profiling!
122 - look into supporting the concept of a "connection" generically