4 - add a new return code from dbus_connection_dispatch() called
5 IN_PROGRESS or RECURSED or something, indicating that DATA_REMAINS
6 but another dispatch is in progress, so we can't dispatch at
7 this time. OR maybe just switch to recursive locks for the dispatch
8 locks. Fixes the recursive deadlock. See the @todo for more
9 and this thread: http://lists.freedesktop.org/archives/dbus/2006-February/004128.html
11 - Take a look at the issues marked @todo 1.0 or FIXME 1.0. Ones with
12 Question marks at the ends either need clarification or are not
13 really needed for 1.0 but would be nice.
15 - the "break loader" and valid/invalid message tests are all disabled;
16 they need to be fixed and re-enabled with the new message args stuff.
17 I think I want to drop the .message files thing and just have code
18 that generates messages, more like the tests for
19 dbus-marshal-recursive.c (this is mostly done now, just needs some
22 - just before 1.0, try a HAVE_INT64=0 build and be sure it runs
24 Important for 1.0 GLib Bindings
27 - Test point-to-point mode
29 - Add support for getting sender
31 - format_version in the object info doesn't look like it's handled correctly. The creator
32 of the object info should specify some fixed number per struct version; the library
33 should handle only specific numbers it knows about. There's no assumption that all
34 numbers >= the given one are compatible. The idea is that new versions of the lib
35 can offer totally different object info structs, but old versions
38 Important for 1.0 Python bindings
43 - Fix removing of signals from the match tree
45 - Fix refcounting and userdata lifecycles
47 - Write a generic mainloop
52 - protocol version in each message is pretty silly
57 - if the GUID is obtained only during authentication, not in the address,
58 we could still share the connection
60 - Allow a dbus_g_proxy_to_string()/g_object_to_string() that
61 would convert the proxy to an "IOR" and dbus_g_proxy_from_string()
62 that would decode; using these, dbus-glib users could avoid
63 DBusConnection entirely. Of course the same applies to other kinds
64 of binding. This would use dbus_connection_open()'s connection-sharing
65 feature to avoid massive proliferation of connections.
67 - DBusWatchList/TimeoutList duplicate a lot of code, as do
68 protected_change_watch/protected_change_timeout in dbus-connection.c
69 and dbus-server.c. This could all be mopped up, cut-and-paste
70 fixed, code size reduced.
72 - change .service files to allow Names=list in addition to Name=string
74 - The message bus internal code still says "service" for
75 "name", "base service" for "unique name", "activate" for
76 "start"; would be nice to clean up.
78 - Property list feature on message bus (list of properties associated
79 with a connection). May also include message matching rules
80 that involve the properties of the source or destination
83 - Disconnecting the remote end on invalid UTF-8 is probably not a good
84 idea. The definition of "valid" is slightly fuzzy. I think it might
85 be better to just silently "fix" the UTF-8, or perhaps return an error.
87 - build and install the Doxygen manual in Makefile when --enable-docs
89 - if you send the same message to multiple connections, the serial number
90 will only be right for one of them. Probably need to just write() the serial
91 number, rather than putting it in the DBusMessage, or something.
93 - perhaps the bus driver should have properties that reflect attributes
94 of the session, such as hostname, architecture, operating system,
95 etc. Could be useful for code that wants to special-case behavior
96 for a particular host or class of hosts, for example.
98 - currently the security policy stuff for messages to/from
99 the bus driver is kind of strange; basically it's hardcoded that
100 you can always talk to the driver, but the default config file
101 has rules for it anyway, or something. it's conceptually
102 screwy at the moment.
104 - when making a method call, if the call serial were globally unique,
105 we could forward the call serial along with any method calls made
106 as a result of the first method call, and allow reentrancy that was
107 strictly part of the call stack of said method call. But I don't
108 really see how to do this without making the user pass around the
109 call serial to all method calls all the time, or disallowing
112 If done post 1.0 will probably be an optional/ugly-API type
115 - I don't want to introduce DBusObject, but refcounting and object
116 data could still be factored out into an internal "base class"
119 - Keep convenience wrappers in sync with bus methods
121 - document the auth protocol as a set of states and transitions, and
122 then reimplement it in those terms
124 - recursive dispatch, see dbus_connection_dispatch()
126 - do we need per-display activation; if so I'd like to do this by setting a
127 "display ID" property on screen 0, with a GUID, and keying activation by
128 said GUID. Otherwise you get all kinds of unrobust
129 string/hostname-based mess. per-screen is then done by appending screen number
130 to the display. If displays have a deterministic ID like this, you can
131 do per-display by simply including GUID in the service name.
133 - optimization and profiling!
135 - Match rules aren't in the spec (probably a lot of methods on the bus
141 - look into supporting the concept of a "connection" generically