nilix/openbox
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
6457 commits 2 branches 0 tags 11 MiB
Commit graph

741 commits

Author SHA1 Message Date
Dana Jansens
a16f1374ee comment out the applications example 2007-05-22 20:52:42 +00:00
Dana Jansens
e97a443590 don't compare variables to themselves 2007-05-22 02:28:03 +00:00
Dana Jansens
4c25342316 dont use the client's area rect when figuring out where to place it because then when you want to move ti there, it thinks that it is already there! i actually fixed another bug recently that would have affectd this before but now it should work perfect.. 2007-05-22 01:36:17 +00:00
Dana Jansens
e8bac610ad adjust the frame's size before placing the window so we know how much decor it has 2007-05-22 01:32:28 +00:00
Dana Jansens
a0941554f6 fix showing the resize cursors at the right times. 
fix losing the pre-size/position for windows mapped maximized or fullscreen states (and on restart)
 2007-05-22 01:03:13 +00:00
Dana Jansens
c5508ec16f we put desktop windows on all desktops so dont let them move. 
make sure the window can move/resize when shutting down and restoring its geometry
 2007-05-22 00:24:06 +00:00
Dana Jansens
f0940aba64 missing another break; ! wow 2007-05-22 00:19:04 +00:00
Dana Jansens
798ed11be8 missing a break; heh 2007-05-22 00:18:45 +00:00
Dana Jansens
62c311ba92 better ordering so you can unmaximize maximized windows.. but you cant resize them 2007-05-22 00:11:14 +00:00
Dana Jansens
119c48fc6c you can't resize fully maxed windows 2007-05-22 00:09:08 +00:00
Dana Jansens
b8f207892a some repositioning of frame elements for maximized windows. 
fix clicking on borders not counting as the button in that corner for full max windows
disable the resize cursors for full max windows
 2007-05-21 23:56:07 +00:00
Dana Jansens
bca4fe6c8b maybe the user time stuff should be more like other wm implementtions and less in the way 2007-05-21 21:48:27 +00:00
Dana Jansens
fd8ce9414a use simple pattern matching for per-app settings. all rules that match are applied to a window rather than just the first 2007-05-21 03:42:25 +00:00
Dana Jansens
64d812c6d7 fix windows moving to monitors other than their own. client_find_onscreen uses their desired coordiantes instead of their old ones 2007-05-20 15:16:57 +00:00
Dana Jansens
9752670920 remove a function that is no longer used. and backwards boolean algebra was causing us to not ignore the right enter events 2007-05-19 00:38:06 +00:00
Dana Jansens
6d26502631 super correct enter event ignoring that will only ignore what it has to, yay? 2007-05-18 00:11:01 +00:00
Dana Jansens
382d37553d add an atom ob_wm_action_undecorate to match with ob_wm_state_undecorated, so stuff could know if the window can be undecorated 2007-05-17 04:15:34 +00:00
Dana Jansens
c6dac85397 don't let the user toggle decorations on widnows that don't have any decorations to toggle 2007-05-17 02:29:13 +00:00
Dana Jansens
281c1edb42 kill enter events without using pointer grabs 2007-05-17 02:21:22 +00:00
Dana Jansens
b1076c9333 rename client_configure_full to client_configure 2007-05-17 01:45:00 +00:00
Dana Jansens
2ca83f6713 grab the pointer when windows move them selves so no enter events happen. i wonder if this will cause but reports. it still does for restacking though.. 
little cleanup in client.c, and remove client_configure macro which was now
redundant
 2007-05-17 01:38:01 +00:00
Dana Jansens
619eeaa962 grab the pointer accross actions run by the applications, the same as for 
actions run by the user
 2007-05-17 00:56:29 +00:00
Dana Jansens
4fdf8a4315 "fullscreen" windows that are undecorated by the user do not count as legacy fullscreen windows 2007-05-17 00:50:09 +00:00
Dana Jansens
b69bf80982 fullscreen only if no decorations 2007-05-16 17:01:47 +00:00
Dana Jansens
d30809e5fa always ignore errors for client_focus like we used to. 2007-05-15 17:38:52 +00:00
Dana Jansens
361bf6af5c simplify focus handling a bit. no need to listen to focus going to the frame because if it ever does that, the window is unmapping and the frame will just unmap and send it to root. 
make focus fallback a lil more robust, it checks for errors when it sets focus on the window, and if an error occured it will focus the next option.
 2007-05-15 04:32:59 +00:00
Dana Jansens
5d85502e3b also if you hide the focused window, then kill the interactive action so focus isnt all confused for the app 2007-05-14 23:34:05 +00:00
Dana Jansens
47c8d40729 don't assert, users can bind stuff however they like. but if they do something that moves focus during an interactive action, kill the grab 2007-05-14 23:32:50 +00:00
Dana Jansens
8f2291a7b6 excep the assert was backwards. heh 2007-05-14 23:22:06 +00:00
Dana Jansens
f5d7952a94 big scary g_assert(). interactive actions should never move focus until they're done. 2007-05-14 23:21:53 +00:00
Dana Jansens
f9fe78d970 get rid of the fullscreen below layer. its the same as maximized windows.. 2007-05-14 23:00:36 +00:00
Dana Jansens
ec21f8c9ec yay for simplifying code. if we check for errors we dont need the "focus_tried" stuff 2007-05-14 22:58:08 +00:00
Dana Jansens
415ce6e5de watch for errors while setting focus on windows, return false if there was an error, so fallback and keep looking 2007-05-14 22:52:40 +00:00
Dana Jansens
2e7359a2fd fix fullscreen windows being stuck in the wrong layer when they come out of 
fullscreen.
add a new layer where fullscreen windows go when are aren't focused, below
all normal windows, so you can alt-tab to a below layer window and it will
show itself.
 2007-05-14 22:20:48 +00:00
Dana Jansens
b906db3e11 remove the dialog option for Desktop action for now. can't change desktop during a grab. the dialog for desktop changing will need to change, and then we can add it back 2007-05-14 20:55:36 +00:00
Dana Jansens
9011eb27e0 remove this check, it didnt guarantee anything anyways 2007-05-14 13:44:49 +00:00
Dana Jansens
26ff0cbaa8 the configurenotify needs to come after adjusting the frame's size. otherwise it gets messed up somehow. 2007-05-14 02:27:43 +00:00
Dana Jansens
1a5139f283 send a configurenotify on mapping after the window is shown 2007-05-14 02:20:59 +00:00
Dana Jansens
9d7608789c another little focus cleanup/rearrangement, for focus going to windows that 
are no longer there.
 2007-05-14 01:06:42 +00:00
Dana Jansens
1eff22b1ef remove some old checking because we're not using window properties to store the premax/fullscreen areas. they are always set when the state is turned on. 
fullscreen to the correct monitor.
 2007-05-14 00:07:38 +00:00
Dana Jansens
99e23015cf rename the openbox_* atoms back to ob_* 2007-05-13 22:31:36 +00:00
Mikael Magnusson
73e7b2925a fix error in client_directional_edge_search that ignored all windows in the same layer 2007-05-13 21:13:29 +00:00
Dana Jansens
94d9d069b5 add a debug line 2007-05-13 17:17:23 +00:00
Dana Jansens
11df262d7e add support for _NET_WM_ACTION_ABOVE and _NET_WM_ACTION_BELOW which aren't in the NETWM spec yet, but will be shortly. 2007-05-13 17:15:15 +00:00
Dana Jansens
0ee1581824 fix drawcontents=no 2007-05-13 16:45:58 +00:00
Dana Jansens
fd9b6a43cf always send configure notify when windows map 2007-05-13 16:04:40 +00:00
Dana Jansens
56d7bc50f0 don't allow focus to move during an interactive grab. if users set up actions 
such that this would happen, then kill the interactive grab before moving
focus.
this is to avoid NotifyWhileGrabbed FocusOut's
 2007-05-13 15:29:49 +00:00
Dana Jansens
acfa2af3c2 use the right list in client_add_hide_notify, rename destructor to destroy_notify which is really what it is, and is more consistant now that there are 2 notifies 2007-05-13 07:45:15 +00:00
Dana Jansens
685dca1c5f likewise if a window is unmanaged and we were trying to give it focus, we want to know about it - also if the window gets hidden for some other reason, we also want to know about it. 
add a notifier for windows being hidden, and use that instead - it handles both cases.
 2007-05-13 07:38:35 +00:00
Dana Jansens
1e957a1688 some changes to focus handling. 
most interesting is the change in focus_fallback, which means that it won't
call xsetinput focus in some cases, potentially reducing flicker and stuff.
also potentially producing bugs? heh.

the screen.c focus fallback code doesn't seem to need special cases anymore,
which is really good, if that is really the case.

move the focus_tried stuff out of event.c into focus.c, where it seems to
belong.
 2007-05-13 07:09:34 +00:00
Dana Jansens
811fa21940 remove the edges_hit_layers_below option. don't use windows in other layers for resistance 2007-05-12 18:56:56 +00:00
Dana Jansens
b2006b8b45 don't show handles on windows that cannot resize 2007-05-11 21:31:19 +00:00
Dana Jansens
b23476ed58 fix for circular transiency 2007-05-11 20:45:26 +00:00
Dana Jansens
2baad3abee oops, debug prints 2007-05-11 05:55:45 +00:00
Dana Jansens
8432416d4e pick the closest icon instead of always a smaller one 2007-05-11 05:54:57 +00:00
Dana Jansens
75372cd4a7 set the default window gravity to NorthWest. 
only put the bare minimum inside the server grab.
adjust the frame decorations in client_manage instead of frame_grab_client
remove the last parameter from client_configure_full and just determine it from the other state available.
send configurenotifies at all the right times.
make windows get restored to their position properly when they have a border
 2007-05-11 02:46:59 +00:00
Dana Jansens
65d50861ad when un-showing the desktop if a window is getting shown, iconify the others so just the one is shown 2007-05-10 23:22:54 +00:00
Dana Jansens
f1f946ea02 stacking fixes for layers 2007-05-10 20:23:27 +00:00
Dana Jansens
4a34c7ddb2 don't get the layer from the group like that. let new windows go in their default layer. they are already kept above their parents. 2007-05-10 17:15:41 +00:00
Dana Jansens
11dd7db876 make the frame before you use the frame.. 2007-05-10 05:25:34 +00:00
Dana Jansens
083a7565c7 set the session stuff before trying to apply per-app settings. 
set the decor/functions after restoring all the stuff from session/per-app too
 2007-05-10 05:21:05 +00:00
Dana Jansens
9830440643 make restacking much better, yay 
no more cludge using actions to raise windows. when a window changes layer it
uses add_nonintrusive now so it won't cover the focused window. this way
fullscreen windows when they drop down, don't cover up the new focus target.

fix add_nonintrusive so that if the window is focused it gets added to the top

add back support for ConfigureRequest restacking, this time properly though,
using all the detail and sibling modes. but when windows use this to raise they
are using some old business and we're going to assume they actually want to
activate instead. this means firefox works nicely. yay.

ubuntu's firefox has been made to just stop raising entirely though. !
 2007-05-10 00:57:35 +00:00
Dana Jansens
66afa1dceb when a window is fully maxed, make clicking on the titlebar past the edge buttons count as clicking on the buttons 2007-05-09 22:21:28 +00:00
Dana Jansens
f03eef2795 mem leak 2007-05-09 21:27:45 +00:00
Dana Jansens
7f262bc2a0 add support for _NET_REQUEST_FRAME_EXTENTS 2007-05-09 20:13:20 +00:00
Dana Jansens
851555348e support for _NET_WM_USER_TIME_WINDOW round 2 ! yay abstraction 2007-05-09 17:58:58 +00:00
Dana Jansens
cdb108c76d add support for _NET_WM_USER_TIME_WINDOW. round 1 ! ding. 2007-05-09 17:01:30 +00:00
Dana Jansens
5aea44adc9 punctuation 2007-05-09 03:41:05 +00:00
Dana Jansens
48d7045208 update the client menu when you do stuff without closing it. 
but this means we can't use actions in it. oh well?
can kill the desktop notifiers now too. yay for more obvious code paths.
 2007-05-08 23:50:13 +00:00
Dana Jansens
709ebbb6a2 let you move splash screens if you are so inclined 2007-05-08 22:56:29 +00:00
Dana Jansens
d074b11bf9 let windows that cant iconify still iconify with their parents, but not non-normal windows. 
let you focus windows that wouldnt normally by clicking on their decor (if they have any)
 2007-05-08 22:53:09 +00:00
Dana Jansens
f0e5332b4a but deiconify 2007-05-08 22:36:52 +00:00
Dana Jansens
370b5ba27d dont iconify if you can't 2007-05-08 22:36:26 +00:00
Dana Jansens
b0f43a1159 don't focus splash screens and toolbars and menus when you click on them. or enter them. 2007-05-08 22:29:39 +00:00
Dana Jansens
9782122339 try again. merge r6177-6182 from 3.4-working 2007-05-08 22:03:56 +00:00
Dana Jansens
1e4215c5c6 merge r6176-6182 from 3.4-working 2007-05-08 21:57:36 +00:00
Dana Jansens
aae4f66f68 recalc the layer when above/below changes 2007-05-08 12:22:28 +00:00
Dana Jansens
cc9ef6f22d dont check if relatives have focus. gtk apps all set application even when its from a user event 2007-05-08 02:33:20 +00:00
Dana Jansens
e4cffb6459 redo how a window checks if it can activate. 
added the bahavior that.. if an application is the source of the request, it will only move focus if something related to it has focus already (and user time checks out)
 2007-05-08 01:59:23 +00:00
Dana Jansens
de82c4a5df i dont like this "application window" concept. swap it with "helper" windows instead which are the inverse. helper windows are utility/menu/toolbar windows. 2007-05-08 01:31:51 +00:00
Dana Jansens
ebabf3943c make helper windows share desktops with all their application top level windows 2007-05-08 01:25:30 +00:00
Dana Jansens
1fa08abd9c um... hide the client menu when the window changes desktops. its not visible anymore (probably..) and no send to menu wont be wrong too. 2007-05-07 23:48:05 +00:00
Dana Jansens
280529221e add a notifier for clients changing desktops. use it to update the send-to menu if it changes. it does this by closing/opening the menu.. thats about the best we can do tho with this menu code without huge changes 2007-05-07 23:26:22 +00:00
Dana Jansens
edf29cfe73 when a window becomes transient for its group, it needs to have group transients removed from itself 2007-05-07 17:07:35 +00:00
Dana Jansens
13df16a681 windows on other desktops are iconic state too 2007-05-07 03:16:45 +00:00
Dana Jansens
f138bd8007 this had changed for the animation stuff but it is no longer needed i think. also in show desktop mode, windows aren't iconic state they are just "behind" the desktop. 2007-05-07 03:10:05 +00:00
Dana Jansens
f506cbceae the session properties are not set on the group leader, they are set on the "client leader" which is in the WM_CLIENT_LEADER property. also they are all set when the window maps, and not to be changed thereafter. 2007-05-07 00:10:15 +00:00
Dana Jansens
b02947c09f client machine and wm_command can be set on the group leader 2007-05-06 23:37:30 +00:00
Dana Jansens
2c507a7207 missed some stuff? 2007-05-06 22:58:51 +00:00
Dana Jansens
6cd5f7ea38 document the per app settings in rc.xml.in. some xsd fixes. add the "default" option to all the per-app settings which is the same as not putting it in there at all. 2007-05-06 22:04:55 +00:00
Dana Jansens
3c4d73c484 save session state for old clients that dont use XSMP. more session improvements. with ksmserver we can save a few things before the clients go messing with stuff during the interactive phase. 2007-05-06 11:14:20 +00:00
Dana Jansens
0fc06816aa don't make the desktop fullscreen 2007-05-06 09:10:09 +00:00
Dana Jansens
277db2822d a bunch of thigns got squashed into this commit.. sorry.. 
1. soem bug fixes for window stacking
2. clarify some functions behavior with their names
3. add (untested) support for legacy fullscreen apps. i have no idea what they do on a multihead xinerama setup though and if this would help there or need some changes.
 2007-05-06 08:32:03 +00:00
Dana Jansens
a7f65a818c add an allDesktops option to Next/PreviousWindow to alt-tab across all desktops. 2007-05-06 05:51:08 +00:00
Dana Jansens
eac3a64917 more restrictive to keep clients on screen in more cases 2007-05-06 05:30:12 +00:00
Dana Jansens
1bf9de3809 refactor all the session stuff. yay, it works properly now. 
make sessions save the desktop being displayed. can we get rid of config_firstdesk now?
refactor startup a bit. focus the window that was focused when we were restarted.
have clients properly restore their session state.
add undecorated to the saved session state for clients.
 2007-05-06 05:19:19 +00:00
Dana Jansens
8a99fedecd prepend is faster 2007-05-05 19:27:31 +00:00
Dana Jansens
3409148047 fixes to the properties and net_supported. a bunch weren't supported. better checking at runtime to make sure everything is in the atom. 2007-05-05 18:19:26 +00:00
Dana Jansens
378adaa94f yay. way way cleaner code for iconify animations. let people show/hide the frame logically and it will do everything except during animations the frame will show what it needs to for visual display. 2007-05-05 16:53:48 +00:00
Dana Jansens
b5e597f1b3 lots of fixes for the iconify animation. i think it should all work perfectly now ? 
one change in action.c may affect bindings... THIS COULD BREAK CURRENT CONFIG FILES. ya.
so, if you used a press binding and it moved the window, no enter event will be generated for that anymore because you are holding down the button when the window moves. if you don't like that then use a click binding instead.

1
 2007-05-05 16:16:26 +00:00
Dana Jansens
08e42acaad animate iconify/reestore. yeah. 2007-05-05 05:58:35 +00:00
Dana Jansens
fde5ca09e9 when restoring from show-desktop mode because a window is going to be focused, then don't restore focus how it was before showing the desktop. 2007-05-05 02:48:39 +00:00
Dana Jansens
fce7d9b21e and some more transient fixes 2007-05-05 01:17:05 +00:00
Dana Jansens
1046c19fda more fixes for the transient tree 2007-05-05 01:12:59 +00:00
Dana Jansens
0c15af505f unused variable 2007-05-05 00:48:51 +00:00
Dana Jansens
2713831371 missing paranthesis 2007-05-04 23:30:13 +00:00
Dana Jansens
431035df66 continuation of r6039 in all respects 2007-05-04 23:15:08 +00:00
Dana Jansens
7ff21ad74f add the maginificent client_update_transient_tree. 
add a whole lot of comments ! i think like half of this commit is comments
speed up mapping windows a little! less re-updating of the transient info by far.
and... yeah fix the crash created from r6035.
 2007-05-04 22:50:53 +00:00
Dana Jansens
15ec358151 group transients werent getting added to their parents 2007-05-04 16:14:16 +00:00
Dana Jansens
b5161f8032 revert r6029, as it didnt fix anything. 
however this patch does fix the aforementioned problem.
actions need some reworking... yeah... later...
 2007-05-04 07:21:22 +00:00
Dana Jansens
a823e4786e dont move windows away from omnpresentness when deiconifying them 2007-05-04 07:19:21 +00:00
Dana Jansens
e84fb1b052 drop the DoNotPropogate thing. metacity doesn't set it at all, and i cant tell any difference in the events generated without it. 
grab mouse buttons on the client window itself for client bindings. this fixes the weird "click and drag doesnt work when the window is focused" behavior ive been seeing with kdesktop. hooray !
 2007-05-04 04:47:43 +00:00
Dana Jansens
07af4a6184 some improvements on how group transients get their transient pointers, needed after r6022. 2007-05-04 04:25:44 +00:00
Dana Jansens
147f83fe3c this is gross... i think it can go away from other changes... if openbox goes into infinite loops.. then it needs to come back ! 2007-05-04 04:08:31 +00:00
Dana Jansens
02dda1ef65 set the usertime on windows if they dont provide one, to something reasonable. 2007-05-04 04:06:43 +00:00
Dana Jansens
b71f1cd6a5 from r6017, we no longer need to fallback in client_unmanage. yay 2007-05-04 02:48:23 +00:00
Dana Jansens
00bf97aecd not much.. some debugs and a nicer if statement 2007-05-04 02:47:10 +00:00
Dana Jansens
5784f11132 add the notion of "application windows" meaning normal or dialog type windows. 
let you alt-tab to non-application windows when they dont have an application window in their group
 2007-05-03 20:57:40 +00:00
Dana Jansens
9c19d7f677 find_on_screen takes client width/height, not frame 2007-05-03 03:18:56 +00:00
Dana Jansens
d7108df02b when deiconifying, deiconify all transients, not just direct ones 2007-05-03 00:50:40 +00:00
Dana Jansens
80117db6c4 turn on demands_attention when urgent gets set, and turn it off when urgent is removed, for those apps that think it is 1917 or something 2007-05-02 23:03:43 +00:00
Dana Jansens
97cbacd9e4 make keeping windows on screen much more clever 2007-05-02 02:03:06 +00:00
Dana Jansens
564b727966 fix drawcontent=no resizing 2007-05-01 12:17:37 +00:00
Dana Jansens
cad10d8b6d some old changes to grav.c test, it wasn't a valid test before. 
when a window gets reconfigured, try keep it on the monitor if it was before.
 2007-05-01 05:02:26 +00:00
Dana Jansens
55d2916c1e a whole lot of changes to the moving/resizing code. it was broken for non-northwest gravities. now it is not. at least, that is the idea. 2007-05-01 04:46:29 +00:00
Dana Jansens
1c4c2b92a0 focus_client is already nulled from the fallback, and use client_focused for niceness in the future 2007-04-26 06:14:09 +00:00
Dana Jansens
b673823d0b have to fallback in client_unmanage, until we figure out a way to hear those focusin events without conflicting screens 2007-04-26 06:10:13 +00:00
Dana Jansens
d9b25d4884 add group transients to other transients who arent group transients. is that english? YEP 2007-04-26 02:32:50 +00:00
Dana Jansens
e896709c90 when windows disappear dont move focus from enter event, ever. 2007-04-26 02:21:27 +00:00
Dana Jansens
f6b049cd38 pointer type changed a long time ago 2007-04-25 00:50:20 +00:00
Mikael Magnusson
fb2493115f only grab keybinds on the root window, not every client window 2007-04-24 21:45:43 +00:00
Dana Jansens
d19c184b2d set the client's user_time when it gets activated 2007-04-24 19:55:48 +00:00
Dana Jansens
0116d82463 don't activate and change desktops if the client's not on the current desktop and its user_time is older than the last time you changed desktops 2007-04-24 19:53:50 +00:00
Dana Jansens
ca8665e309 fix a comment 2007-04-24 19:26:48 +00:00
Dana Jansens
8446e50b7a use client_activate to activate newly mapping windows. this means that if a window maps on another desktop and is going to be activated, you will end up on that desktop 2007-04-24 19:20:10 +00:00
Dana Jansens
ac241a8235 set the default icon onto windows which don't provide one 2007-04-24 15:56:37 +00:00
Dana Jansens
0771964c0f set the max settings properly, was setting max_vert for both. 
make sure the desktop property is set when per-app settings change it
 2007-04-24 14:58:29 +00:00
Mikael Magnusson
e86900bd53 fix resizing when drawContents is off, and I think an ifdef was a bit too high up 2007-04-24 12:18:53 +00:00
Dana Jansens
c6fe4f8d8f focus_order_remove is called in a few places. move the check for unsetting focus_client into client_unmanage. 2007-04-24 06:41:04 +00:00
Dana Jansens
ad852b6b56 add support for _NET_WM_SYNC_REQUEST 2007-04-24 06:30:16 +00:00
Dana Jansens
7ae95e4660 save windows' colormaps rather than querying them from the server every time you focus it 2007-04-24 01:29:53 +00:00
Dana Jansens
3a91527fdf hm, that was a silly error. the propogate mask shouldnt get the Hint thing in it. 2007-04-24 01:08:09 +00:00
Dana Jansens
a488360f48 add the PointerMotionHintMask everywhere, we dont need every mouse event 2007-04-24 01:00:27 +00:00
Dana Jansens
268c87ed9e in shrinking resize the frame before the client 2007-04-24 00:02:51 +00:00
Dana Jansens
5316bc6c3e 1. remove some old event handling that was not used at all. yay less bandwidth. 
2. make the client padding color show during resizes with drawContents off
 2007-04-24 00:00:11 +00:00
Dana Jansens
1a005d7291 place transients without a parent in the middle of the screen 
dont use non-normal windows as parents
 2007-04-23 23:46:13 +00:00
Dana Jansens
b4eecc51b5 focus_client was becoming invalid if the focused client was unmanaged. 2007-04-23 23:11:33 +00:00
Dana Jansens
9f11e2af6b cycling between dock windows now possible with the <panels>yes</panels> option 2007-04-23 22:40:33 +00:00
Mikael Magnusson
9e7a240dfd include the header that has gethostname 2007-04-23 19:02:33 +00:00
Mikael Magnusson
141c249b8c show the hostname of forwarded clients in the titlebar 2007-04-23 18:46:19 +00:00
Dana Jansens
a82c1efeed wasteful syncing 2007-04-23 02:00:33 +00:00
Dana Jansens
7e0d660b8b stop listening to events on clients that are being unmanaged before generating new events 2007-04-22 22:56:53 +00:00
Dana Jansens
037654f3f8 use ignore enters instead of removing them from the queue 2007-04-22 16:48:35 +00:00
Dana Jansens
27773e1cb7 don't grab the pointer during interactive events. this allows you to alt-tab during a dnd. 2007-04-22 15:22:25 +00:00
Dana Jansens
a70633d42a eat enter events after unmanaging a window 
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)
 2007-04-22 13:53:28 +00:00
Dana Jansens
6e7abeb860 engrish 2007-04-22 04:35:23 +00:00
Dana Jansens
8a972b3879 ignore BadMatch errors that can't be avoided 2007-04-21 21:36:02 +00:00
Dana Jansens
0e9bd66b75 fixes for transients 
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
 2007-04-21 21:32:21 +00:00
Dana Jansens
d713bdf462 remove the client_time_heap thing. actually its better to just use the user_time on the currently focused window I think. that will handle panels and things too. 2007-04-20 05:22:23 +00:00
Dana Jansens
90cd9c6219 get rid of global client_last_user_time variable. 
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.
 2007-04-18 19:22:59 +00:00
Dana Jansens
64dda2375e saved session information overrides per-app settings, because that's how you set it up in the session 2007-04-12 14:54:54 +00:00
Dana Jansens
33c469f453 set the desktop from per-app-settings properly, i think. 
also fix a compiler warning.
 2007-04-12 14:47:54 +00:00
Dana Jansens
7a1a6da849 better synthetic configurenotify handling. in particular.. 
http://bugzilla.gnome.org/show_bug.cgi?id=399552
on unmap, when we move the client window, we need to send a synthetic configurenotify.
 2007-04-12 14:44:17 +00:00
Dana Jansens
c03c08415f don't fallback when a window is hidden right away, it's redundant, we'll get a focusout for the window after the unmapnotify anyways. 
show new windows when switching desktops before hiding the old windows.
 2007-04-12 02:26:26 +00:00
Dana Jansens
7059cde13b http://mail.gnome.org/archives/wm-spec-list/2006-May/msg00000.html 
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.
 2007-04-05 03:59:24 +00:00
Dana Jansens
0eaf2d15e7 fully maximize windows all the time. size increments are preferred sizes not required ones. 
-- http://tronche.com/gui/x/icccm/sec-4.html#s-4.1.2.3
 2007-03-30 04:37:53 +00:00
Dana Jansens
00ddaf06bb can't use (unsigned)-1 for a max time, because timestamps wraparound and stuff. user the last_user_time as a default for new windows instead. 2007-03-28 19:57:07 +00:00
Dana Jansens
33cc572402 drop some wasted client_validates. 2007-03-28 02:00:10 +00:00
Dana Jansens
d20f84f5de more handling time wraparounds 2007-03-28 01:53:37 +00:00
Dana Jansens
339d767044 handle time wrapping around. 2007-03-28 01:52:06 +00:00
Dana Jansens
eb2a0feb01 watch for reverttoparent reverting to the root window, which will create a DetailInferior focusin event on root. 
adding some debug messages also which I am using..
 2007-03-24 23:19:45 +00:00
Dana Jansens
c4732aeed2 2007-03-24 17:17:02 +00:00
Dana Jansens
e77353f0cd clean up the title getting function 2007-03-24 17:06:19 +00:00
Dana Jansens
4c2b16e72d kill window title numbering... 
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
 2007-03-24 17:01:40 +00:00
Dana Jansens
fb3d2033ba not passing timestamps sometimes and not others. set event_curtime when client messages provide a timestamp. 2007-03-19 17:59:16 +00:00
Dana Jansens
5fa443efc8 oops, logic error 2007-03-17 20:06:16 +00:00
Dana Jansens
73dac2b9b9 fix for bug introduced in revision 5643. logical size is supposed to be updated only when the window is resized. 2007-03-16 19:14:53 +00:00
Dana Jansens
adc5675823 oh, 2 things in this commit.. 
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.
 2007-03-14 15:51:21 +00:00
Mikael Magnusson
768ce7b456 move stuff around in client_unmanage to maybe make closing windows a bit nicer 2007-03-14 14:17:31 +00:00
Dana Jansens
5f7ffb00ba refix for managing iconic windows, without having the frame map which caused flashing. 
instead, keep track of if the frame has been mapped yet. if it hasn't and it is getting hidden, then make sure to unmap the child.
 2007-03-14 13:58:36 +00:00
Dana Jansens
e2521f5c64 a) fix resizerelative moving windows when they reach their minimum size do this by... 
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)
 2007-03-14 03:45:56 +00:00
Dana Jansens
594648729f less wasted updates of wm_state 2007-03-14 02:07:30 +00:00
Mikael Magnusson
53109b7592 comparing strings to random locations in memory is probably not what we wanted to do 2007-03-13 12:34:58 +00:00
Dana Jansens
a9910e520b always unfocus the window when it disappears, no special cases.. (except unmanage) 2007-03-13 04:15:25 +00:00
Dana Jansens
270a5b25df set IconicState on windows that are not visible always. this means windows on other desktops too. 
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.
 2007-03-13 04:04:13 +00:00
Dana Jansens
7f58bd9eaa save and restore pre-maximized/fullscreen sizes properly 
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..
 2007-03-13 02:29:08 +00:00
Dana Jansens
d47c6a0979 fallback is still getting x errors. instead, fallback when the client iconifies, and don't fallback when nothing has focus, send it to where we can control things though. 2007-03-12 17:31:11 +00:00
Dana Jansens
85a1a0cfb3 don't validate inside client_focus. instead, validate before you call it! 2007-03-12 06:04:00 +00:00
Dana Jansens
6b7cc88c9d validate clients before trying to focus them in fallback 
also validate before trying to set the focus on a window

this should fix those crashes, and should just be better over all
 2007-03-12 05:57:30 +00:00
Dana Jansens
c018e21220 i rewrote handling of focus events. this is pretty much based on blackbox's current form, as well as reading the xlib programming manual at: 
http://tronche.com/gui/x/xlib/events/input-focus/normal-and-grabbed.html

this may break for people. that'd be nice to hear about, so it can be fixed. but hopefully this is more robust. it sure is a lot more simple.
 2007-03-12 05:25:34 +00:00
Dana Jansens
d33f44cd86 Rewrite the stacking code. It's a lot faster now, I should think. It's def a more clever algorithm. It deals with group transients much better. 
On that note, utility and menu and toolbar window types are now treated as group transients in terms of stacking and focus and such.
 2007-03-12 02:24:40 +00:00
Dana Jansens
e2b3a32aed We were grabbing SubstructureNotifyMask on the root window. I don't know why.. It meant that we got unmap notifies for windows we weren't even managing and stuff. It ended up breaking VLC when it mapped-unmapped-mapped a window. We grab StructureNotify for all the clients, so I am not sure why this was ever there. 2007-03-11 22:16:42 +00:00
Dana Jansens
cbfa27759d extra braces.. why does the world hate me so? 2007-03-11 19:27:44 +00:00
Dana Jansens
d9601bc196 .. i am pretending to be mika :D :D ok this time for real 2007-03-11 19:27:16 +00:00
Dana Jansens
d98a6183a3 oh.. wasted verbosity.. *strums guitar* 2007-03-11 19:26:28 +00:00
Dana Jansens
e714fa7351 rewrite the focus steal prevention code to clean it up hopefully a bit. also don't steal focus from globally active focus model clients 2007-03-11 19:25:57 +00:00
Dana Jansens
6714a0e2ca Honor transient_for=Root, even thought this is crap and breaks some KDE stuff. Also include a big comment as to why this is suck. 2007-03-11 19:15:28 +00:00
Dana Jansens
7229bea3c6 client_configure_full is trying to move a window, it shouldn't determine the window's monitor from its current position, but from the new position instead. move client_monitor()'s logic into screen, more generically, and let client_configure_full use that 2007-03-11 17:34:07 +00:00
Dana Jansens
62a39c4c70 stacking has been made more reliable with groups and group transients. 
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.
 2007-03-11 16:58:19 +00:00
Dana Jansens
2c98204dfe make _NET_ACTIVE_WINDOW requests look at the timestamp when the user didn't 
request it
 2007-03-11 14:17:14 +00:00