change how focus falls back for windows being unmanaged, don't fall back immediately, instead wait for the focus out event and use it (break stuff maybe! yay)
1) don't count non-normal windows as parents when placing/stacking transients
2) in stacking.c, when a window is transient for the group but has no parents, then don't loop forever looking for its parents
add ObClientTimeHeap. This is a max-heap of the clients based on their user times. this only includes the clients whose user time is not CurrentTime. the maximum from this heap replaces the client_last_user_time variable, so that you always have the latest time, not the last time that was changed.
hoefully it works, so far it seems to.
Send ConfigureNotify events when a window is mapped at the position it has specified. When we add decorations, the window gets moved in reference to the root window, so it needs to be told.
probably it was a dumb idea all along. you're not picking the numbers yourself and you're probably not remembering which window they apply to and the focus indicator thinger is much more clever than numbers
1) THIS ONE IS IMPORTANT: don't set client->iconfied if the window is not actually going to be allowed to iconify. heh!
2) changes to focus fallback to avoid crashing and to avoid losing focus all at the same time.
b) adding a function to let you see what size/position a window will be given when you actually try move/resize it to some given values. (client_try_configure)
as well, this turned out to be a nice cleanup of the wmstate code.
regards to the random dude in #openbox complaining about the ICCCM to make me think to look this up and see if openbox was compliant.
much cleanup on how per-app settings are applied to new mapping windows. maybe i broke it? but it'll be much quicker now, and it's consistant with the rest of the code, so hooray for that. pls let me know if i broke any of it :( hee..
this was a pretty invasive change in client.c though, so it may break things?
it did expose some bugginess in client_calc_layer, which is now better than ever, hopefully there isn't more to be found.
all related to _NET_WM_USER_TIME and focus stealing prevention
a) add launcher startup notification. this means when you run something from
the openbox menu or a key/mouse binding, that startup notification will go
on in openbox and other applications like your panel or something
b) add the _NET_WM_USER_TIME property for windows
c) use the _NET_WM_USER_TIME data and startup notification to prevent focus
stealing.
d) cookie party !! ! all are invited.
e) oh yeah, and pass around timestamps for a lot more things. like, when you
run an action, send the timestamp for the event that is running the action.
this is important for startup notification. this also affects menus.
f) yes.. cookies..
would it be a good idea to disable focus stealing prevention if a window takes
too long to load? i mean.. maybe after a certain length of time, a user can't be
expected to not do anything in any other windows, but would they still want the
new application to focus then? HMM. open question i guess..
make sure its 10% on the screen in some direction always..
but if the application is placing itself, make sure its on the screen entirely,
and also put it entirely on one monitor if you have xinerama.
if it's bigger than the monitor's space though, it won't do anything with it..
1. when another wm requests to replace openbox, openbox exits. but the SM will just restart openbox unless we tell it not to. so now ob_exit_replace() will change the session manager's view of openbox to not restart it. that way the new WM will be able to run.
2. allow windows to move themselves off of the screen 90% of the way, if they really want to. but only 90% to the left, right, and bottom of the screen. it won't let the app move off the top of the screen on its own at all now, since hiding the titlebar on you without you being a part of the process is pretty darn evil!
this is really to address bug # 2982 - for the tilda application. but i guess if windows really want to move off the screen, who's to say no? also, every other window manager will let them - except metacity won't let them on the left/top side of the screen.
this requires that when a client is using a 32-bit visual, we have to
make the frame windows that sit underneath it to use the same visual (and a
colormap which matches it)
1. some random compiling/style cleanups
2. some bigfixes
- mislogic in per-window-settings and focusing new windows
- use client_can_focus rather than checking variables for directional focus
- MAYBE fix all those lock-ups forever. using event_curtime (a new variable) now instead of event_lasttime. event_lasttime is still used however when the event being processed did not have a time associated with it. this may or may not be a problem, and will be seen.
3. um.. i forget
4. oh yeah, 3rd party docks are now treated like the internal ob dock irt focus. that is, clicking on them won't pass them focus. this is going to be ratified as expected behavior in the wm-spec just now. if docks/panels want focus they can request it with _net_active_window, and then they can have all the focus they want! one day alt-tabbing around dock windows might be nice. but not until the ob dock is moved out into a separate application. going to have to add a wmapp selection and stuff for that though... ugly. who uses wmdockapps anymore !? someone must.. *sigh*
