1 Important for 0.90 freeze
4 - Audit @todo and FIXME for security issues that require API/ABI changes
6 - dbus-pending-call.c has some API and thread safety issues to review.
7 DBusPendingCall is used from multiple threads with no locks.
8 Either DBusConnection's lock has to protect all associated pending
9 call (means pending->connection can't ever be set to null) or
10 or DBusPendingCall needs its own lock
11 http://lists.freedesktop.org/archives/dbus/2006-June/004945.html
16 - add a new return code from dbus_connection_dispatch() called
17 IN_PROGRESS or RECURSED or something, indicating that DATA_REMAINS
18 but another dispatch is in progress, so we can't dispatch at
19 this time. OR maybe just switch to recursive locks for the dispatch
20 locks. Fixes the recursive deadlock. See the @todo for more
21 and this thread: http://lists.freedesktop.org/archives/dbus/2006-February/004128.html
23 - Audit @todo and FIXME for security issues
25 - the "break loader" and valid/invalid message tests are all disabled;
26 they need to be fixed and re-enabled with the new message args stuff.
27 I think I want to drop the .message files thing and just have code
28 that generates messages, more like the tests for
29 dbus-marshal-recursive.c (this is mostly done now, just needs some
32 - just before 1.0, try a HAVE_INT64=0 build and be sure it runs
34 - publish the introspection dtd at its URL
36 Important for 1.0 GLib Bindings
39 - Test point-to-point mode
41 - Add support for getting sender
43 - format_version in the object info doesn't look like it's handled correctly. The creator
44 of the object info should specify some fixed number per struct version; the library
45 should handle only specific numbers it knows about. There's no assumption that all
46 numbers >= the given one are compatible. The idea is that new versions of the lib
47 can offer totally different object info structs, but old versions
50 Important for 1.0 Python bindings
55 - Fix removing of signals from the match tree
57 - Fix refcounting and userdata lifecycles
59 - Write a generic mainloop
64 - protocol version in each message is pretty silly
69 - if the GUID is obtained only during authentication, not in the address,
70 we could still share the connection
72 - Allow a dbus_g_proxy_to_string()/g_object_to_string() that
73 would convert the proxy to an "IOR" and dbus_g_proxy_from_string()
74 that would decode; using these, dbus-glib users could avoid
75 DBusConnection entirely. Of course the same applies to other kinds
76 of binding. This would use dbus_connection_open()'s connection-sharing
77 feature to avoid massive proliferation of connections.
79 - DBusWatchList/TimeoutList duplicate a lot of code, as do
80 protected_change_watch/protected_change_timeout in dbus-connection.c
81 and dbus-server.c. This could all be mopped up, cut-and-paste
82 fixed, code size reduced.
84 - change .service files to allow Names=list in addition to Name=string
86 - The message bus internal code still says "service" for
87 "name", "base service" for "unique name", "activate" for
88 "start"; would be nice to clean up.
90 - Property list feature on message bus (list of properties associated
91 with a connection). May also include message matching rules
92 that involve the properties of the source or destination
95 - Disconnecting the remote end on invalid UTF-8 is probably not a good
96 idea. The definition of "valid" is slightly fuzzy. I think it might
97 be better to just silently "fix" the UTF-8, or perhaps return an error.
99 - build and install the Doxygen manual in Makefile when --enable-docs
101 - if you send the same message to multiple connections, the serial number
102 will only be right for one of them. Probably need to just write() the serial
103 number, rather than putting it in the DBusMessage, or something.
105 - perhaps the bus driver should have properties that reflect attributes
106 of the session, such as hostname, architecture, operating system,
107 etc. Could be useful for code that wants to special-case behavior
108 for a particular host or class of hosts, for example.
110 - currently the security policy stuff for messages to/from
111 the bus driver is kind of strange; basically it's hardcoded that
112 you can always talk to the driver, but the default config file
113 has rules for it anyway, or something. it's conceptually
114 screwy at the moment.
116 - when making a method call, if the call serial were globally unique,
117 we could forward the call serial along with any method calls made
118 as a result of the first method call, and allow reentrancy that was
119 strictly part of the call stack of said method call. But I don't
120 really see how to do this without making the user pass around the
121 call serial to all method calls all the time, or disallowing
124 If done post 1.0 will probably be an optional/ugly-API type
127 - I don't want to introduce DBusObject, but refcounting and object
128 data could still be factored out into an internal "base class"
131 - Keep convenience wrappers in sync with bus methods
133 - document the auth protocol as a set of states and transitions, and
134 then reimplement it in those terms
136 - recursive dispatch, see dbus_connection_dispatch()
138 - do we need per-display activation; if so I'd like to do this by setting a
139 "display ID" property on screen 0, with a GUID, and keying activation by
140 said GUID. Otherwise you get all kinds of unrobust
141 string/hostname-based mess. per-screen is then done by appending screen number
142 to the display. If displays have a deterministic ID like this, you can
143 do per-display by simply including GUID in the service name.
145 - optimization and profiling!
147 - Match rules aren't in the spec (probably a lot of methods on the bus
153 - look into supporting the concept of a "connection" generically