[01:31:33] <CIA-111> mmlr * r41864 /haiku/trunk/src/add-ons/kernel/drivers/input/usb_hid/ (MouseProtocolHandler.cpp MouseProtocolHandler.h):
[01:31:33] <CIA-111> Since we never actually read more than one mouse_movement at a time and we only
[01:31:33] <CIA-111> ever wait for reports on demand, there's no need for buffering at all.
[01:31:33] <CIA-111> This removes some unnecessary copying and ring buffer overhead, the ring buffer
[01:31:33] <CIA-111> will now in fact not be created at all.
[02:54:15] <CIA-111> mmlr * r41865 /haiku/trunk/src/add-ons/kernel/drivers/input/usb_hid/ (JoystickProtocolHandler.cpp JoystickProtocolHandler.h): (log message trimmed)
[02:54:15] <CIA-111> Make adjustments to the JoystickProtocolHandler so that it works in accordance
[02:54:15] <CIA-111> with the BJoystick requirements:
[02:54:15] <CIA-111> * Make the Read() non-blocking. This is required as BJoystick is a polling
[02:54:15] <CIA-111> interface. A single current state is used that is updated by a separate thread
[02:54:15] <CIA-111> on report arrival. The thread is spawned as soon as the ProtocolHandler is
[02:54:15] <CIA-111> opened for the first time (and quit at the first wait return after the
[06:24:11] <CIA-111> leavengood * r41866 /haiku/trunk/build/jam/OptionalPackages:
[06:24:11] <CIA-111> Update WebPositive to a build made on a gcc4 r1a3 RC.
[06:24:11] <CIA-111> +alpha3
[10:10:59] <CIA-111> phoudoin * r41867 /haiku/trunk/src/tests/kits/device/stickit_BJoystick/Jamfile: Fix gcc4 build.
[14:39:48] <CIA-111> mmlr * r41868 /haiku/trunk/ (3 files in 2 dirs): (log message trimmed)
[14:39:48] <CIA-111> Cleanup of the _BJoystickTweaker:
[14:39:48] <CIA-111> * Include sorting, header style, macro naming, variable naming, spacing, ...
[14:39:48] <CIA-111> * Some simplifications, early returns and the like.
[14:39:48] <CIA-111> * The device name list items were allocated with new but deleted with free. The
[14:39:48] <CIA-111> whole private BJoystick list being modified by a friend class is dubious
[14:39:48] <CIA-111> though and should be reworked. Ideally node monitoring would be implemented
[15:01:42] <CIA-111> mmlr * r41869 /haiku/trunk/src/kits/device/Joystick.cpp:
[15:01:42] <CIA-111> Ensure that a GetDeviceName() call triggers a ScanDevices() as well, as it can
[15:01:42] <CIA-111> be called without calling CountDevices().
[17:07:59] <CIA-111> mmlr * r41870 /haiku/trunk/ (headers/os/device/Joystick.h src/kits/device/Joystick.cpp): (log message trimmed)
[17:07:59] <CIA-111> * Add a BJoystick::RescanDevices() public function. It allows the BJoystick
[17:07:59] <CIA-111> to be updated with newly plugged in devices without the need to recreate a
[17:07:59] <CIA-111> BJoystick object.
[17:07:59] <CIA-111> * Do an initial (re-)scan on object creation instead of always scanning in
[17:08:00] <CIA-111> CountDevices() and by extension GetDeviceName().
[17:08:01] <CIA-111> * This makes CountDevices() and GetDeviceName() stable with regards to the
[18:05:44] <CIA-111> scottmc * r41871 /haiku/trunk/build/jam/OptionalPackages:
[18:05:44] <CIA-111> Updated ruby to a build which has support for readline. Commented out the symlink for friss until it gets figured out
[18:05:44] <CIA-111> and fixed.
[18:46:52] *** echelog-2 has joined #haiku-jp





top