Commit graph

3327 commits

Author SHA1 Message Date
Stefanos A. ef5aedba6f [Win] More robust WGL extension detection
Affects issue #42 and issue #45
2014-01-10 09:24:59 +01:00
thefiddler a4d2a31386 [Mac] OpenGL 3.x/4.x require core profile flag
SDL will fail to construct an OpenGL 3.x/4.x context on Mac OS X,
unless ContextProfileFlags.CORE is specified.

Fixes issue #44

Upstream enhancement request at
https://bugzilla.libsdl.org/show_bug.cgi?id=2342
2014-01-09 23:36:28 +01:00
Stefanos A. 7f64945079 [OpenTK] Do not raise *Frame events when time = 0
Affects issue #40
2014-01-09 00:51:09 +01:00
Stefanos A. 28ac3cec0b [Win] Fix issue #35 (OpenTK over Remote Desktop)
When running over remote desktop without hardware acceleration, there
are no GraphicsModes that support desktop composition. This patch adds
logic to avoid requesting composition-capable modes when running over
RDP.

Additionally, it changes the mode selection logic to consider modes that
support features partially (e.g. 16bpp color instead of 32bpp), albeit
with a heavy penalty over fully supported modes.
2014-01-09 00:14:25 +01:00
Stefanos A. 1b3b510376 [Win] Check registry keys before accessing
Fixes crashes when using OpenTK over the Remote Desktop Client for Mac
(version 2010).
2014-01-09 00:10:41 +01:00
Stefanos A. 4af9d30ccb [Win] Fix issue #33; fix issue #34
This patch adds a workaround for WM_STYLECHANGED messages that are not
delivered when running on Mono/Windows. If we detect Mono, then we call
HandleStyleChanged() directly in order to update the internal state of
our WinGLNative instance.
2014-01-08 22:29:22 +01:00
Stefanos A. 51ad513dbb [Win] Do not overload internal SetWindowLong
The internal function is now appended with "Internal".
2014-01-08 19:22:03 +01:00
Stefanos A. 7363cfee7b [Win] Do not unregister class twice
Only the parent window would register a class, but both the parent and
the child window would unregister it. This is now fixed.
2014-01-08 19:21:29 +01:00
Stefanos A. 51baed7286 [Win] Remove unnecessary #if clauses 2014-01-08 00:45:42 +01:00
Stefanos A. e260a42954 [Win] Refactor huge wndproc into functions 2014-01-08 00:02:27 +01:00
Stefanos A. b6a806a568 [OpenTK] Improved timing stability
OpenTK now directly calculates the elapsed time between UpdateFrame
(RenderFrame) events and compares that directly to TargetUpdatePeriod
(TargetRenderPeriod). This significantly simplifies the implementation
and improves timing stability.
2014-01-07 22:09:02 +01:00
Stefanos A. 6e03d501ae [OpenTK] Fixed Update/RenderTime calculation
These values should only be re-calculated when an Update/RenderFrame
event is raised. Otherwise, they should retain their previous values.
2014-01-07 15:55:11 +01:00
thefiddler 99df27b635 [OpenTK] Corrected GameWindow.Update/RenderFrequency information 2014-01-07 09:11:55 +01:00
thefiddler 56a3dd91e5 [OpenTK] Implemented GameWindow.UpdateTime and RenderTime properties 2014-01-07 09:09:43 +01:00
thefiddler 3856fcd48e [OpenTK] More robust timing for UpdateFrame and RenderFrame
FrameEventArgs.Time should no longer drift from clock time measured
outside GameWindow.
2014-01-07 08:52:02 +01:00
thefiddler 94c3c24bfb [OpenTK] Platform backends should inherit from PlatformFactoryBase
This reduces code duplication wrt deprecated interfaces and common
support code between backends.
2014-01-06 14:49:05 +01:00
thefiddler d84be0d594 [OpenTK] Added base IPlatformFactory implementation
PlatformFactoryBase provides a common base interface for platform
backends. Platform backends should inherit from PlatformFactoryBase in
order to reduce code duplication.
2014-01-06 14:47:58 +01:00
thefiddler a94a293732 [Platform] Implemented legacy IJoystickDriver
LegacyJoystickDriver implements the legacy IJoystickDriver interface
(GameWindow.Joysticks) in terms of the new IJoystickDriver2 interface
(OpenTK.Input.Joystick).

This removes a large chunk of code from each platform backend, as they
no longer need to implement IJoystickDriver themselves. Additionally,
it adds support for device hot plugging which was previously missing.
2014-01-06 14:25:05 +01:00
Stefanos A. 83f54f70aa [OpenTK] Simplify and improve timing calculations
This patch modifies GameWindow.Run() to use a single stopwatch instead
of two separate stopwatches for timing UpdateFrame and RenderFrame
events.

It improves timing accuracy for issue #20 (FrameEventArgs.Time
Inconsistencies)
2014-01-06 01:57:54 +01:00
Stefanos A. 88c57db5b6 [Math] Added MathHelper.Clamp 2014-01-06 01:52:08 +01:00
thefiddler a9ab3650da [Mac] Do not store *State structs directly
Storing Mouse/KeyboardState structs directly makes updates more
difficult than they should be. It is simpler to create simple classes
instead.
2014-01-05 23:58:05 +01:00
thefiddler 4d660fdeba [Input] Added Keyboard/MouseState.SetIsConnected to mirror JoystickState 2014-01-05 23:50:17 +01:00
thefiddler fda2d32d2d [Mac] Map GamePads, MultiAxisControllers and Wheels to joysticks 2014-01-05 22:44:58 +01:00
thefiddler a1123834a0 [Mac] Call DeviceRemoved() in Dispose() event
The necessary cleanup code already exists in DeviceRemoved(). No need
to duplicate this in the Dispose() event.
2014-01-05 22:38:26 +01:00
thefiddler ab85afd5ba [Mac] Avoid IOHIDDeviceConformsTo call in DeviceRemoved handler
DeviceAdded already checks that devices conform to the desired usage
pages. Checking again in DeviceRemoved is unnecessary - if a device
exists, then it has already passed muster.
2014-01-05 22:36:01 +01:00
thefiddler 9e223486f6 [Mac] Implement Joystick.GetGuid() 2014-01-05 22:15:55 +01:00
thefiddler 5c5afb3ea3 [Mac] Calculate joystick Guid for USB devices 2014-01-05 22:06:19 +01:00
thefiddler b458b40683 [Mac] Use out/ref parameters instead of unsafe pointers in bindings 2014-01-05 22:05:54 +01:00
thefiddler fa386dc991 [Mac] Connect devices to first available index 2014-01-05 21:53:28 +01:00
thefiddler 05af21e61b [Mac] Do not call CFRelease on HID properties
HID properties are callee-owned and should not be released by the
caller. Fixes crash (memory corruption) on device hot plugging.
2014-01-05 21:27:34 +01:00
thefiddler 02c9f471ea [Mac] DllImport must be marked as extern 2014-01-05 21:18:20 +01:00
thefiddler 7388bd935d [Mac] Correctly unschedule HIDManager from run loop 2014-01-05 20:17:03 +01:00
thefiddler 4c7f6a92a7 [Mac] Don't let exceptions escape to unmanaged code 2014-01-05 04:20:51 +01:00
thefiddler 892d129e54 [Mac] Fixed DllEntryPointNotFound crash on device unplugging 2014-01-05 04:15:30 +01:00
thefiddler 2ee24efb2f [Mac] Implemented joystick buttons 2014-01-05 04:11:11 +01:00
thefiddler fad42994e1 [Mac] Implemented joystick axis movement 2014-01-05 03:54:53 +01:00
thefiddler 92635b3f35 [Input] Added JoystickAxis/Button.Last; added JoystickAxis.Axis10.
JoystickAxis/Button.Last is used internally to allocate the correct
amount of storage for joystick axes and buttons. JoystickAxis.Axis10 is
required to support the maximum number of axes available on Mac OS X.
2014-01-05 03:54:34 +01:00
thefiddler 59249c2624 [Mac] Implemented joystick device detection 2014-01-05 03:03:46 +01:00
thefiddler 07bcda0f57 [Mac] Retrieve HID joystick name 2014-01-05 02:42:47 +01:00
thefiddler 6dadbd3570 [Mac] Implemented necessary CFString methods 2014-01-05 02:42:33 +01:00
thefiddler 32653d2c64 [Mac] Construct JoystickDevice when joysticks are connected 2014-01-05 00:47:40 +01:00
thefiddler 3757db9978 [Input] Allow extending JoystickDevice<> 2014-01-05 00:46:59 +01:00
thefiddler bf8efea121 [Mac] Only add callbacks for recognized USB HID devices 2014-01-05 00:00:49 +01:00
thefiddler c44b477388 [Mac] Implement device added/removed events for joysticks 2014-01-04 23:58:59 +01:00
thefiddler 9659a1d786 [Mac] Add skeleton HID input implementation for joysticks 2014-01-04 23:55:38 +01:00
thefiddler 3a63496b6d Merge branch 'gamepad' into develop
Conflicts:
	Source/OpenTK/OpenTK.csproj
2014-01-03 02:23:02 +01:00
thefiddler 514390fc9a [Graphics] Added missing documentation comment 2014-01-03 02:18:28 +01:00
thefiddler 53552b0070 [Input] Corrected malformed documentation comments 2014-01-03 02:17:42 +01:00
thefiddler 3095afa18a [Input] Documented all public members 2014-01-03 01:58:16 +01:00
thefiddler b35aad1503 [Input] Removed JoystickState.GetAxis(int) overload 2014-01-03 01:58:08 +01:00
thefiddler 8649e4a044 [Input] Added SetVibration() API skeleton 2014-01-02 19:52:00 +01:00
thefiddler ec43b9ff85 [Input] GamePad and Joystick classes should be sealed 2014-01-02 19:27:04 +01:00
thefiddler b513e35ea8 [SDL2] Implemented PacketNumber property 2014-01-02 19:24:38 +01:00
thefiddler 1ccf7e5ad2 [Input] Added PacketNumber property 2014-01-02 19:24:15 +01:00
thefiddler b5d42b7992 [Input] Implemented GetState() and simplified GetCapabilities() 2014-01-02 18:37:53 +01:00
thefiddler d49af2787f [Input] Simplified configuration parser and added GetEnumerator() 2014-01-02 18:37:20 +01:00
thefiddler 8f00bbc9f2 [Input] Added internal Joystick.GetAxisRaw() method 2014-01-02 18:36:27 +01:00
thefiddler 0593ea62e6 [Input] Added Source and Target to GamePadConfigurationItem 2014-01-02 18:36:09 +01:00
thefiddler bf5d2a738d [Input] Made configuration database non-static 2014-01-02 02:42:51 +01:00
thefiddler 9e4827bf67 [SDL2] Fixed SDL_JoystickGetGUID capitalization 2014-01-02 02:20:51 +01:00
thefiddler a8fb977017 [Input] Initialize JoystickDevice.Details in constructor 2014-01-02 02:20:25 +01:00
thefiddler fb57a9062d [Input] Renamed GamePadMap to GamePadConfiguration 2014-01-02 01:38:12 +01:00
thefiddler dc1ffab78d [Input] Implemented IGamePadDriver.GetName() 2014-01-02 01:34:21 +01:00
thefiddler c2c76f2ab2 [Input] Made the Name property public
We can use the name property to implement a hypothetical
GamePad.GetName() method in the future.
2014-01-02 01:33:57 +01:00
thefiddler b056a50e73 [Input] Added xml documentation for GamePadCapabilities 2014-01-02 01:33:08 +01:00
thefiddler a0dad7f698 [Input] Implemented MappedGamePadDriver.GetCapabilities() 2014-01-02 01:13:20 +01:00
thefiddler f0fbb26efe [Input] Added ability to parse SDL configuration db strings 2014-01-02 01:07:21 +01:00
thefiddler 5e75fac056 [Input] Added configuration db for GamePads
The database is based on the SDL bindings found at
https://hg.libsdl.org/SDL/file/b744b3f8754b/src/joystick/SDL_gamecontrol
lerdb.h
2014-01-02 01:06:47 +01:00
thefiddler b9a8e365de [Input] Added IJoystickDriver2.GetGuid() API 2013-12-31 14:09:17 +01:00
thefiddler cd143af60a [SDL2] Added JoystickGetGUID method 2013-12-31 14:08:28 +01:00
thefiddler 1687518ef5 [SDL2] Add compile-time check for SDL2 GameController vs MappedGamePadDriver 2013-12-31 11:46:40 +01:00
thefiddler ef7f31099d [Input] Removed unnecessary #region 2013-12-31 01:02:07 +01:00
thefiddler 7d4d116ce9 [Input] Renamed GamePadMapping to GamePadMap 2013-12-31 01:01:52 +01:00
thefiddler d4348c5083 [SDL2] Fixed joystick/controller hotplugging support 2013-12-30 17:22:45 +01:00
thefiddler 9c8a5f5028 [SDL2] Fixed GameControllerButton to Buttons translation 2013-12-30 17:15:49 +01:00
thefiddler 43ef78f222 [SDL2] Fix issue where changes to mutable struct were lost 2013-12-30 17:09:20 +01:00
thefiddler 9beb396c9e [SDL2] Explicitly enable joystick and gamepad events 2013-12-30 17:08:42 +01:00
thefiddler 2d303a6884 [Input] Buttons should have [Flags] attribute 2013-12-30 16:07:06 +01:00
Stefanos A. 6faa58aac3 [SDL2] Fixed joystick hotplugging
SDL2 uses a weird system of device ids and instance ids to report
joystick events, where the ADDED event uses a device id and the rest use
instance ids.

The SDL2 joystick driver is now fixed to correctly distinguish between
the two, which fixes hotplugging support for joystick devices.
2013-12-30 15:24:48 +01:00
thefiddler 52b8762593 [SDL2] Fixed mapping of instance ids to device ids 2013-12-30 14:57:28 +01:00
thefiddler a4366e52f5 [Input] Added missing closing brace in ToString() message 2013-12-30 14:56:22 +01:00
thefiddler 064a45e4a7 [Input] Fixed incorrect variable name 2013-12-30 13:49:26 +01:00
thefiddler 1b0a72472e [SDL] Use TranslateAxis() to decode ControllerAxisEvent
SDL GameControllerAxis and GamePadAxes are not interchangeable. The
driver will now correctly interpret incoming SDL messages and update
the GamePadState for the relevant axis.
2013-12-30 13:42:37 +01:00
thefiddler e997ddf9c6 [Input] Fixed GamePadState.SetAxis() implementation
GamePadState.SetAxis() receives a GamePadAxes enumeration, which is a
bitmask of the axes we wish to set. SetAxis now correctly decodes the
bitmask to apply the values we are interested in.
2013-12-30 13:39:43 +01:00
thefiddler 0b8db7c40b [X11] Removed unused variable 2013-12-30 10:25:17 +01:00
thefiddler 4005710b99 [X11] Fixed current resolution detection. 2013-12-30 09:23:33 +01:00
thefiddler 3dbf575765 [Build] Added new files to the relevant csproj scripts 2013-12-28 01:23:39 +01:00
thefiddler 04e6d9335a [Graphics] Improved support for external OpenGL contexts
When combining OpenTK with a third-party OpenGL toolkit, it is now
possible to implement a suitable GetAddress() and GetCurrentContext()
implementation in terms of the third-party toolkit. If no
implementation is specified, then OpenTK will try to guess the most
suitable implementation within its own platform backends.

If no custom implementation is defined, and if no suitable
implementation can be found, then OpenTK will throw a
PlatformNotSupportedException. If a suitable implementation is found or
defined, then OpenTK will attempt to load OpenGL entry points using
that implementation.

In this case third-party toolkit remains solely responsible for
managing its context via its MakeCurrent(), SwapBuffers(), etc
implementations.
2013-12-28 01:14:59 +01:00
thefiddler 5cebaccfca [Dummy] Improved thread-safety; added entry-point loader
DummyGLContext will now attempt to load OpenGL and OpenGL ES entry
points when a suitable OpenGL context is current on the calling thread.
This allows OpenTK to be used on contexts created through third-party
toolkits.
2013-12-28 01:09:59 +01:00
thefiddler 8bcbb06f8e [Platform] Added internal CreateGetAddress() function
CreateGetAddress() constructs a GraphicsContext.GetAddressDelegate that
is suitable for the current platform. This can be used when combining
OpenTK with an OpenGL context created through a third-party toolkit.
2013-12-28 01:08:24 +01:00
thefiddler cf76aa6085 [Mac] Moved GetAddress implementation to its own class
The functions defined in libdl.dylib are useful in more places than
just AglContext. Moving them to their own class ensures we can access
these from wherever we might need them.
2013-12-28 01:06:33 +01:00
thefiddler 9171977079 [SDL2] Added missing [SuppressUnamangedCodeSecurity] attribute 2013-12-28 01:04:46 +01:00
thefiddler b79d951e62 [Mac] Use Cgl.GetCurrentContext instead of Agl
CGL appears to work with both AGL and NSOpenGL contexts, whereas AGL is
limited to AGL contexts. This allows us to be more flexible in terms of
implementation (i.e. we can use Cgl.GetCurrentContext to retrieve a
handle to a context created through SDL, which uses NSOpenGL
internally.)
2013-12-28 01:04:22 +01:00
thefiddler 0d1df41393 [X11] Improve GraphicsMode fallback (fixes issue #23)
When the user requests a GraphicsMode that is not directly supported
by the GPU/drivers, we should relax the requested parameters until
we find a supported mode. An exception should only be thrown when
there is no usable mode.

This makes the X11 backend match the behavior of Windows. The SDL/X11
backend works a little bit differently, in that it falls back to the
a default mode directly if the requested mode is not available. There
is nothing we can do about that.
2013-12-27 14:01:21 +02:00
thefiddler 88f7cd68f5 [X11] Fixed border size on Gnome 3
The _NET_FRAME_EXTENTS atom is implemented differently by
different window managers, when window decorations are hidden
with Motif. Unity returns a 0 size, while Gnome 3 returns the
previous size.

This patch removes that ambiguity: when decorations are hidden,
border size becomes zero. This should work everywhere, unless
some window manager decides to troll us by decorating the window
when we explicitly request no decorations. Sigh...
2013-12-27 13:31:51 +02:00
thefiddler ab8796c942 [Input] Do not crash with scancodes > 255 2013-12-27 11:10:41 +01:00
thefiddler 6c6e09aae6 [X11] Allow resizing with fixed borders
Windows can now be resized programmatically, even when they have
WindowBorder.Fixed. All resizing logic is now consolidated in the
Bounds property, and ConfigureNotify messages are now handled
correctly depending on their source (StructureNotify or
SubStructureNotify.)
2013-12-27 11:07:38 +01:00
thefiddler 4d27b6ee55 Implemented X11 KeyDown/KeyUp events and reduced code duplication 2013-12-25 11:46:02 +01:00
thefiddler ffdf881cea Do not raise KeyPress for control characters 2013-12-25 11:14:30 +01:00
thefiddler f7fbf38c43 IsButtonBind should check for buttons, not axes 2013-12-25 01:47:27 +01:00