Doc++, added some TODOs.

doc/best-practices.dox

@@ -80,6 +80,10 @@ vertex and index buffers in both Native Client and Emscripten.
 
 - [PowerVR Performance Recommendations](https://github.com/burningsun/pecker_framework/blob/master/参考资料/PowerVR.Performance%20Recommendations.pdf?raw=true) [PDF]
 
+@subsection best-practices-amd AMD hardware
+
+- [ATI Radeon HD 2000 programming guide](http://developer.amd.com/wordpress/media/2012/10/ATI_Radeon_HD_2000_programming_guide.pdf)
+
 @subsection best-practices-tegra NVidia Tegra hardware
 
 - [Optimize OpenGL ES 2.0 Performance for Tegra](http://docs.nvidia.com/tegra/index.html#GLES2_Perf_Main.html)

doc/opengl-support.dox

@@ -50,7 +50,7 @@ following:
 @todo Replace @extension{APPLE,vertex_array_object} with ARB version (we use only ARB functions without APPLE fallback)
 @todo @extension{APPLE,flush_buffer_range} doesn't add anything to @extension{ARB,map_buffer_range} (remove it and leave only ARB one)
 @todo @extension{EXT,texture_array} overlaps with @extension{ARB,framebuffer_object}
-@todo Replace @extension{NV,depth_buffer_float} with @extension{ARB,depth_buffer_float} and implement the missing @fn_gl{DepthRange} function
+@todo Add @extension{ARB,depth_buffer_float} and implement the missing @fn_gl{DepthRange} function, but keep (and implement) @extension{NV,depth_buffer_float} for non-linear depth buffer
 @todo @extension{ARB,half_float_pixel}, @extension{ARB,half_float_vertex}?
 
 %Extension                                   | Status
@@ -194,6 +194,7 @@ following:
 @subsection opengl-support-44 OpenGL 4.4
 
 @todo Also fallback to @extension{AMD,query_buffer_object}
+@todo @extension{AMD,pinned_memory} "fallback" for @extension{ARB,buffer_storage}
 
 %Extension                                   | Status
 -------------------------------------------- | ------
@@ -276,6 +277,7 @@ supported.
 
 Only extensions not already listed in above tables are included here.
 
+@todo @es_extension{NV,non_square_matrices}
 @todo Support also IMG_multisampled_render_to_texture? It has different enum
     values (!)
 

doc/troubleshooting.dox

@@ -26,7 +26,7 @@
 namespace Magnum {
 /** @page troubleshooting Troubleshooting
 
-@brief Various tricks to overcome to common building and rendering issues.
+@brief Various tricks to overcome common building and rendering issues.
 
 @section troubleshooting-building Building issues
 

src/Magnum/AbstractShaderProgram.h

@@ -181,6 +181,9 @@ bindFragmentDataLocationIndexed(NormalOutput, 1, "normal");
     ES 2.0, similar functionality is available in extension
     @es_extension{NV,draw_buffers}.
 
+@todo @es_extension{EXT,separate_shader_objects} supports explicit attrib
+    location
+
 @subsection AbstractShaderProgram-uniform-location Uniform locations
 
 The preferred workflow is to specify uniform locations directly in the shader

src/Magnum/SceneGraph/Object.h

@@ -93,6 +93,8 @@ class documentation or @ref compilation-speedup-hpp for more information.
 
 @see @ref Scene, @ref AbstractFeature, @ref AbstractTransformation,
     @ref DebugTools::ObjectRenderer
+@todo Consider using `mutable` for flags to make transformation computation
+    available on const refs
 */
 template<class Transformation> class Object: public AbstractObject<Transformation::Dimensions, typename Transformation::Type>, public Transformation
     #ifndef DOXYGEN_GENERATING_OUTPUT