* Move compositor to core, but letting the configuration there
* Rename all files and functions from e_mod_comp_* to e_comp_*
* Move the config dialogs to a new module named conf_comp. It still
uses a domain config, otherwise it would not pick the current
user's configuration. Maybe it would be wise to later on move these
options to e_config
* Fixup the wizard mess linking the header in the build tree in order
to be able to create the config. Since now it's in core, we don't
need to play linking games in the build system
I'm not sure if the wayland part works. It was not even building
previously so I'll let for who cares about this to actually test and
report bugs.
SVN revision: 82454
In order to move composite inside the core we need to kill the
use_composite option. However in some places it is being used to switch
between ARGB and shaped windows. It doesn't make much sense to keep the
advanced/engine dialog to let the user toggle this options if we have
composite always enabledm, but lets allow the user to shoot
himself on the foot (for now).
Next step will be to move the comp module files to core so user can't
unload it anymore.
SVN revision: 82433
as false. this breaks expected functionality from apps - that they
request activation and eg - window is minimized... it gets shown at
least. app here wants attention/activation... this is better than
ignoring ti entirely. theres activation policy config specifically for
turning this off separately. :/
SVN revision: 82402
as an aside, elm_win is SEVERELY (cripplingly) broken with regard to WM hint applying in the sense that it never applies them
at the least, however, this will prevent windows from being randomly resized smaller than minimum sizes
SVN revision: 82394
dropshadow module conflicts with composite, which will be always enabled
by future commits. Remove the module to allow turning composite as
always-enabled.
SVN revision: 82224
- Change module name in whitelist. Its name is bluez4 since it only
supports bluez 4 API. The day it (automatically) also supports bluez 5,
we rename it back.
- Use eina_stringshare_replace when possible
- Change huge if (strcmp(icon, "...") return "....." with a table.
- Don't use stringshare when not needed. There's no reason to
stringshare the icon, just use the static const ones
- Sanitize edbus method table
- Remove bogus NONE entry after NULL in services table
- Fix coding style
- Generalize _unset_dev() to work with any list
SVN revision: 82208
It would be very weird and error-prone on clients to have code like
this:
E_Notification_Notify n;
memset(n, 0, sizeof(*n));
n.app_name = "bla";
e_notification_client_send(...);
eina_stringshare_del(n.icon.icon);
eina_stringshare_del(...);
...
Instead of doing this, replace NULL by "" in the only place it's not
allowed: while sending the D-Bus message.
SVN revision: 82120
typos:
toogle -> toggle
itens -> items
API consistency:
MENU_ITEM_TOGGLE -> E_DBUSMENU_ITEM_TOGGLE
E_DBus_Menu_Item_Event -> E_DBusMenu_Item_Event
(I don't want to bikeshed on this, but we need to choose one of the
possible names, and keep consistent with the rest of the API)
A few var renames: It's E_DBusMenu_Ctx, so let's name the var ctx
instead of menu_data. internal_ctx is no more meaninful than ctx, so
rename it as well.
SVN revision: 82092
Mike Blumenkrantz
Carsten Haitzler
Cedric BAIL
Rafael Antognolli
Lucas De Marchi
Tom Hacohen
Miculcy Brian
Igor Murzov
Lucas Joia
José Roberto de Souza