Wed, 29 Sep 2010 21:24:50 -0700Documentation clarification
Sam Lantinga [Wed, 29 Sep 2010 21:24:50 -0700] rev 4903
Documentation clarification

Wed, 29 Sep 2010 21:13:52 -0700Make the union nameless to reduce the complexity of the API.
Sam Lantinga [Wed, 29 Sep 2010 21:13:52 -0700] rev 4902
Make the union nameless to reduce the complexity of the API.
(Are there any compilers still in use that don't support this?)

Mon, 27 Sep 2010 01:30:42 -0700Minor cleanup in patches from Coursoud
Sam Lantinga [Mon, 27 Sep 2010 01:30:42 -0700] rev 4901
Minor cleanup in patches from Coursoud

Mon, 27 Sep 2010 01:24:05 -0700Merged the Windows custom window system hooks into the union used by X11.
Sam Lantinga [Mon, 27 Sep 2010 01:24:05 -0700] rev 4900
Merged the Windows custom window system hooks into the union used by X11.
Added Cocoa custom window system hooks

Mon, 27 Sep 2010 00:50:46 -0700attached is a patch to support DirectFB in include/SDL_syswm.h. It
Sam Lantinga [Mon, 27 Sep 2010 00:50:46 -0700] rev 4899
attached is a patch to support DirectFB in include/SDL_syswm.h. It
defines SDL_SYSWM_DIRECTFB as a subsystem.
This allows developers to e.g. access DirectFB's video and picture
providers.

Kind regards,

André

Mon, 27 Sep 2010 00:49:56 -0700XDamageDestroy is called before use_damage is tested. This crashes on my
Sam Lantinga [Mon, 27 Sep 2010 00:49:56 -0700] rev 4898
XDamageDestroy is called before use_damage is tested. This crashes on my
setup. The patch calls XDamageDestroy only when use_damage is true.

Kind regards,

André

Mon, 27 Sep 2010 00:49:34 -0700SDL_DestroyWindow is setting window->magic to NULL too early. -
Sam Lantinga [Mon, 27 Sep 2010 00:49:34 -0700] rev 4897
SDL_DestroyWindow is setting window->magic to NULL too early. -
window->magic is also checked in SDL_DestroyRenderer. All cleanup code
like freeing the renderer and textures is not executed.
The patch moves window_magic = NULL behind the SDL_DestroyRenderer call.

Kind regards,

André

Mon, 27 Sep 2010 00:48:56 -0700Code will never reach SDL_SetTexturePalette since SDL_SetDisplayPalette
Sam Lantinga [Mon, 27 Sep 2010 00:48:56 -0700] rev 4896
Code will never reach SDL_SetTexturePalette since SDL_SetDisplayPalette
will fail if the display e.g. is RGB32 or RGB24. The patch will ignore
errors in SDL_SetDislayPalette.
This should also affect the opengl renderer on systems supporting
paletted textures - mine does not (NVIDIA).
With this patch the directfb drivers delivers the same results for
test_palette like the software renderer.

Kind regards,

André

Sat, 25 Sep 2010 17:06:41 -0700Fixed documentation typo
Sam Lantinga [Sat, 25 Sep 2010 17:06:41 -0700] rev 4895
Fixed documentation typo

Sun, 19 Sep 2010 16:46:56 -0700Fixed compiling with mingw32/64
Sam Lantinga [Sun, 19 Sep 2010 16:46:56 -0700] rev 4894
Fixed compiling with mingw32/64