3 - The message format has to include information about number of fds
4 in the message so we can skip a message correctly. Or we should
5 just give up on trying to recover from unknown messages.
7 - generate pointer_focus (and drag focus) on raise/lower, move
8 windows, all kinds of changes in surface stacking.
14 Root window must send NULL type (to decline drop) or
15 x-wayland/root-something type if the source offers that. But the
16 target deletes the drag_offer object when drag.pointer_focus leaves
19 How do we animate the drag icon back to the drag origin in case of
22 How to handle surfaces from clients that don't know about dnd or
23 don't care? Maybe the dnd object should have a
24 dnd.register_surface() method so clients can opt-in the surfaces
25 that will participate in dnd. Or just assume client is not
26 participating until we receive an accept request.
28 - Pointer image issue:
30 - A touch input device doesn't have a pointer; indicate that
33 - Cursor themes, tie in with glyph/image cache.
35 - copy-n-paste, store data in server (only one mime-type available)
36 or do X style (content mime-type negotiation, but data goes away
39 - Discard buffer, as in "wayland discarded your buffer, it's no
40 longer visible, you can stop updating it now.", reattach, as in "oh
41 hey, I'm about to show your buffer that I threw away, what was it
42 again?". for wayland system compositor vt switcing, for example,
43 to be able to throw away the surfaces in the session we're
44 switching away from. for minimized windows that we don't want live
47 - Initial placement of surfaces. Guess we can do, 1)
48 surface-relative (menus), 2) pointer-relative (tooltips and
49 right-click menus) or 3) server-decides (all other top-levels).
51 - When a surface is the size of the screen and on top, we can set the
52 scanout buffer to that surface directly. Like compiz unredirect
53 top-level window feature. Except it won't have any protocol state
54 side-effects and the client that owns the surface won't know. We
55 lose control of updates. Should work well for X server root window
56 under wayland. Should be possible for yuv overlays as well.
58 - what about cursors then? maybe use hw cursors if the cursor
59 satisfies hw limitations (64x64, only one cursor), switch to
60 composited cursors if not.
62 - clients needs to allocate the surface to be suitable for
63 scanout, which they can do whenever they go fullscreen.
65 - multihead, screen geometry and crtc layout protocol, hotplug
67 - input device discovery, hotplug
69 - Advertise axes as part of the discovery, use something like
70 "org.wayland.input.x" to identify the axes.
72 - keyboard state, layout events at connect time and when it
73 changes, keyboard leds
79 - synaptics, 3-button emulation, scim
81 - auth; We need to generate a random socket name and advertise that
82 on dbus along with a connection cookie. Something like a method
83 that returns the socket name and a connection cookie. The
84 connection cookie is just another random string that the client
85 must pass to the wayland server to become authenticated. The
86 Wayland server generates the cookie on demand when the dbus method
87 is called and expires it after 5s or so.
89 - or just pass the fd over dbus
91 - drm bo access control, authentication, flink_to
93 - Range protocol may not be sufficient... if a server cycles through
94 2^32 object IDs we don't have a way to handle wrapping. And since
95 we hand out a range of 256 IDs to each new clients, we're just
96 talking about 2^24 clients. That's 31 years with a new client
97 every minute... Maybe just use bigger ranges, then it's feasible
98 to track and garbage collect them when a client dies.
100 - Add protocol to let applications specify the effective/logical
101 surface rectangle, that is, the edge of the window, ignoring drop
102 shadows and other padding. The compositor needs this for snapping
103 and constraining window motion. Also, maybe communicate the opaque
104 region of the window (or just a conservative, simple estimate), to
105 let the compositor reduce overdraw.
107 - multi gpu, needs queue and seqno to wait on in requests
113 - draw window decorations in gtkwindow.c
115 - Details about pointer grabs. wayland doesn't have active grabs,
116 menus will behave subtly different. Under X, clicking a menu
117 open grabs the pointer and clicking outside the window pops down
118 the menu and swallows the click. without active grabs we can't
119 swallow the click. I'm sure there much more...
121 - Port Qt? There's already talk about this on the list.
125 - move most of the code from xf86-video-intel into a Xorg wayland
128 - don't ask KMS for available output and modes, use the info from
129 the wayland server. then stop mooching off of drmmode.c.
131 - map multiple wayland input devices to MPX in Xorg.
133 - rootless; avoid allocating and setting the front buffer, draw
134 window decorations in the X server (!), how to map input?
136 - gnome-shell as a wayland session compositor
138 - runs as a client of the wayland session compositor, uses
139 clutter+egl on wayland
141 - talks to an Xorg server as the compositing and window manager
142 for that server and renders the output to a wayland surface.
143 the Xorg server should be modified to take input from the system
144 compositor through gnome-shell, but not allocate a front buffer.
146 - make gnome-shell itself a nested wayland server and allow native
147 wayland clients to connect and can native wayland windows with
148 the windows from the X server.
150 - qemu as a wayland client; session surface as X case
152 - qemu has too simple acceleration, so a Wayland backend like the
153 SDL/VNC ones it has now is trivial.
155 - paravirt: forward wayland screen info as mmio, expose gem ioctls as mmio
157 - mapping vmem is tricky, should try to only use ioctl (pwrite+pread)
159 - not useful for Windows without a windows paravirt driver.
161 - two approaches: 1) do a toplevel qemu window, or 2) expose a
162 wayland server in the guest that forwards to the host wayland
163 server, ie a "remote" compositor, but with the gem buffers
164 shared. could do a wl_connection directly on mmio memory, with
165 head and tail pointers. use an alloc_head register to indicate
166 desired data to write, if it overwrites tail, block guest. just
167 a socket would be easier.
169 - moblin as a wayland compositor
171 - clutter as a wayland compositors