irrlicht/include/IVideoDriver.h

// Copyright (C) 2002-2007 Nikolaus Gebhardt
// This file is part of the "Irrlicht Engine".
// For conditions of distribution and use, see copyright notice in irrlicht.h

#ifndef __IRR_I_VIDEO_DRIVER_H_INCLUDED__
#define __IRR_I_VIDEO_DRIVER_H_INCLUDED__

#include "rect.h"
#include "SColor.h"
#include "ITexture.h"
#include "irrArray.h"
#include "matrix4.h"
#include "dimension2d.h"
#include "position2d.h"
#include "SMaterial.h"
#include "IMeshBuffer.h"
#include "triangle3d.h"
#include "EDriverTypes.h"
#include "EDriverFeatures.h"

namespace irr
{
namespace io
{
	class IAttributes;
	class IReadFile;
} // end namespace io
namespace scene
{
	class IMeshBuffer;
} // end namespace scene

namespace video
{
	struct S3DVertex;
	struct S3DVertex2TCoords;
	struct S3DVertexTangents;
	struct SLight;
	struct SExposedVideoData;
	class IImageLoader;
	class IImageWriter;
	class IMaterialRenderer;
	class IGPUProgrammingServices;

	//! enumeration for geometry transformation states
	enum E_TRANSFORMATION_STATE
	{
		//! View transformation
		ETS_VIEW = 0,
		//! World transformation
		ETS_WORLD,
		//! Projection transformation
		ETS_PROJECTION,
		//! Texture transformation
		ETS_TEXTURE_0,
		//! Texture transformation
		ETS_TEXTURE_1,
		//! Texture transformation
		ETS_TEXTURE_2,
		//! Texture transformation
		ETS_TEXTURE_3,
		//! Not used
		ETS_COUNT
	};


	//! Interface to driver which is able to perform 2d and 3d gfx functions.
	/** The IVideoDriver interface is one of the most important interfaces of
	the Irrlicht Engine: All rendering and texture manipulating is done with
	this interface. You are able to use the Irrlicht Engine by only invoking methods
	of this interface if you would like to, although the irr::scene::ISceneManager interface
	provides a lot of powerful classes and methods to make the programmers life
	easier.
	*/
	class IVideoDriver : public virtual IUnknown
	{
	public:

		//! destructor
		virtual ~IVideoDriver() {}

		//! Applications must call this method before performing any rendering.
		/** \param backBuffer: Specifies if the back buffer should be cleared, which
		means that the screen is filled with the color specified.
		If this parameter is false, the back buffer will not be cleared and the color
		parameter is ignored.
		\param zBuffer: Specifies if the depth or z buffer should be cleared. It is
		not nesesarry to do so if only 2d drawing is used.
		\param color: The color used for back buffer clearing
		\return Returns false if failed. Begin Scene can clear the back- and the z-buffer. */
		virtual bool beginScene(bool backBuffer, bool zBuffer, SColor color) = 0;

		//! Presents the rendered image on the screen.
		/** Applications must call this method after performing any rendering.
		\param  windowId: Handle of another window, if you want the bitmap to be displayed
		on another window. If this is null, everything will be displayed in the default window.
		Note: This does not work in fullscreen mode and is not implemented for all devices (only for
		D3D8, D3D9, Software1 and Software2, and only for Windows).
		\param sourceRect: Pointer to a rectangle defining the source rectangle of the area
		to be presented. Set to null to present everything. Note: not implemented in all devices.
		\return Returns false if failed and true if succeeded. */
		virtual bool endScene( s32 windowId = 0, core::rect<s32>* sourceRect=0 ) = 0;

		//! Queries the features of the driver.
		/** Returns true if a feature is available
		\param feature: A feature to query.
		\return Returns true if the feature is available, false if not. */
		virtual bool queryFeature(E_VIDEO_DRIVER_FEATURE feature) = 0;

		//! Sets the view, world or projection transformation.
		/* \param state: Transformation type to be set. Can be view, world or projection.
		\param mat: Matrix describing the transformation. */
		virtual void setTransform(E_TRANSFORMATION_STATE state, const core::matrix4& mat) = 0;

		//! Returns the transformation set by setTransform
		virtual const core::matrix4& getTransform(E_TRANSFORMATION_STATE state) = 0;

		//! Sets a material.
		/** All 3d drawing functions draw geometry now
		using this material.
		\param material: Material to be used from now on. */
		virtual void setMaterial(const SMaterial& material) = 0;

		//! Returns a pointer to a texture.
		/** Loads the texture if it is not
		already loaded, and generates mipmap levels if desired.
		You can influence how the texture is loaded using the setTextureCreationFlag()
		method.
		The texture can be in BMP, JPG, TGA, PCX, PNG, and PSD format.
		For loading BMP, TGA, PCX, and PSD files the engine uses its own methods.
		PCX loading is based on some code by Dean P. Macri, PNG loading
		is done using a loader by rt.
		For loading JPG-Files the JPEG LIB 6b, written by
		The Independent JPEG Group is used. For PNG loading,
		libPNG is used. Thanx for such great libraries!
		\param filename: Filename of the texture to be loaded.
		\return Returns a pointer to the texture and 0 if the texture
		could not be loaded.
		This pointer should not be dropped. See IUnknown::drop() for more information.*/
		virtual ITexture* getTexture(const c8* filename) = 0;

		//! Returns a pointer to a texture.
		/** Loads the texture if it is not
		already loaded, and generates mipmap levels if desired.
		You can influence how the texture is loaded using the setTextureCreationFlag()
		method.
		The texture can be in BMP, JPG, TGA, PCX, PNG, and PSD format.
		For loading BMP, TGA, PCX, and PSD files the engine uses its own methods.
		PCX loading is based on some code by Dean P. Macri, PNG loading
		is done using a loader by rt.
		For loading JPG-Files the JPEG LIB 6b, written by
		The Independent JPEG Group is used. For PNG loading,
		libPNG is used. Thanx for such great libraries!
		\param file: Pointer to an already opened file.
		\return Returns a pointer to the texture and 0 if the texture
		could not be loaded.
		This pointer should not be dropped. See IUnknown::drop() for more information.*/
		virtual ITexture* getTexture(io::IReadFile* file) = 0;

		//! Returns a texture by index
		/** \param index: Index of the texture, must be smaller than getTextureCount()
		Please note that this index might change when adding or removing textures
		*/
		virtual ITexture* getTextureByIndex(u32 index) = 0;

		//! Returns amount of textures currently loaded
		virtual s32 getTextureCount() = 0;

		//! Renames a texture
		virtual void renameTexture(ITexture* texture, const c8* newName) = 0;

		//! Creates an empty Texture of specified size.
		/** \param size: Size of the texture.
		\param name: A name for the texture. Later calls to getTexture() with this name
		will return this texture
		\param format: Desired color format of the texture. Please note that
		the driver may choose to create the texture in another color format.
		\return Returns a pointer to the new created Texture.
		This pointer should not be dropped. See IUnknown::drop() for more information. */
		virtual ITexture* addTexture(const core::dimension2d<s32>& size,
			const c8* name, ECOLOR_FORMAT format = ECF_A8R8G8B8) = 0;

		//! Creates a texture from a loaded IImage.
		/** \param name: A name for the texture. Later calls of getTexture() with this name
		 will return this texture
		\param image: Image the texture is created from.
		\return Returns a pointer to the newly created Texture.
		This pointer should not be dropped. See IUnknown::drop() for more information. */
		virtual ITexture* addTexture(const c8* name, IImage* image) = 0;

		//! Creates a render target texture.
		/** \param size: Size of the texture, in pixels. Width and height should
		be a power of two (for example 64, 128, 256, 512, ...) and it should not
		be bigger than the backbuffer, because it shares the zbuffer with the
		screen buffer.
		\return Returns a pointer to the created texture or 0 if the texture could not
		be created. If you no longer need the image, you should call ITexture::drop().
		See IUnknown::drop() for more information. */
		virtual ITexture* createRenderTargetTexture(const core::dimension2d<s32>& size) = 0;

		//! Removes a texture from the texture cache and deletes it, freeing lot of memory.
		/** Please note that after calling this, the pointer to the ITexture
		may not be longer valid, if it was not grabbed before by other parts of
		the engine for storing it longer. So it would be a good idea to set all
		materials which are using this texture to null or another texture first.
		\param texture: Texture to delete from the engines cache. */
		virtual void removeTexture(ITexture* texture) = 0;

		//! Removes all texture from the texture cache and deletes them, freeing lot of	memory.
		/** Please note that after calling this, the pointer to all ITextures
		may not be longer valid, if they were not grabbed before by other parts of
		the engine for storing them longer. So it would be a good idea to set all
		materials which are using textures to null first. */
		virtual void removeAllTextures() = 0;

		//! Creates an 1bit alpha channel of the texture based of an color key.
		/** This makes the texture transparent at the regions where this color
		key can be found when using for example draw2DImage with useAlphachannel
		= true.
		\param texture: Texture of which its alpha channel is modified.
		\param color: Color key color. Every pixel with this color will get transparent
		like described above. Please note that the colors of a texture may get
		converted when loading it, so the color values may not be exactly the same
		in the engine and for example in picture edit programs. To avoid this
		problem, you could use the makeColorKeyTexture method, which takes the position
		of a pixel instead a color value. */
		virtual void makeColorKeyTexture(video::ITexture* texture, video::SColor color) = 0;

		//! Creates an 1bit alpha channel of the texture based of an color key position.
		/** This makes the texture transparent at the regions where this color
		key can be found when using for example draw2DImage with useAlphachannel
		= true.
		\param texture: Texture of which its alpha channel is modified.
		\param colorKeyPixelPos: Position of a pixel with the color key color.
		Every pixel with this color will get transparent
		like described above. */
		virtual void makeColorKeyTexture(video::ITexture* texture,
			core::position2d<s32> colorKeyPixelPos) = 0;

		//! Creates a normal map from a height map texture.
		/** If the target texture
		has 32 bit, the height value is stored in the alpha component of the texture as
		addition. This value will be used by the video::EMT_PARALLAX_MAP_SOLID
		material and similar  materials.
		\param texture: Texture of which its alpha channel is modified.
		\param amplitude: Constant value by which the height information is multiplied.*/
		virtual void makeNormalMapTexture(video::ITexture* texture, f32 amplitude=1.0f) = 0;

		//! Sets a new render target.
		/** This will only work if the driver
		supports the EVDF_RENDER_TO_TARGET feature, which can be
		queried with queryFeature(). Usually, rendering to textures is done in this
		way:
		\code
		// create render target
		ITexture* target = driver->createRenderTargetTexture(core::dimension2d<s32>(128,128));

		// ...

		driver->setRenderTarget(target); // set render target
		// .. draw stuff here
		driver->setRenderTarget(0);     // set previous render target
		Please note that you cannot render 3D or 2D geometry with a render target as texture
		on it when you are rendering the scene into this render target at the same time. It is
		usually only possible to render into a texture between the IVideoDriver::beginScene() and endScene()
		method calls. And please also note that the scene will be rendered upside down into the texture
		in some devices (e.g. OpenGL vs. D3D). A simple workaround for this is to flip the
		texture coordinates of the geometry where the render target texture is displayed on.
		\endcode
		\param texture: New render target. Must be a texture created with
		IVideoDriver::createRenderTargetTexture(). If set to 0, it sets the previous render
		target which was set before the last setRenderTarget() call.
		\param clearBackBuffer: Clears the backbuffer of the render target with the color parameter
		\param clearZBuffer: Clears the zBuffer of the rendertarget. Note that, because the frame
		buffer shares the zbuffer with the rendertarget, its zbuffer will be partially cleared
		too with this.
		\param color: The background color for the render target.
		\return Returns true if sucessful and false if not. */
		virtual bool setRenderTarget(video::ITexture* texture,
			bool clearBackBuffer=true, bool clearZBuffer=true,
			SColor color=video::SColor(0,0,0,0)) = 0;

		//! Sets a new viewport.
		/** Every rendering operation is done into this	new area.
		\param area: Rectangle defining the new area of rendering operations. */
		virtual void setViewPort(const core::rect<s32>& area) = 0;

		//! Gets the area of the current viewport.
		/** \return Returns rectangle of the current vieport. */
		virtual const core::rect<s32>& getViewPort() const = 0;

		//! draws a vertex primitive list
		/** Note that there may be at maximum 65536 vertices, because the
		index list is an array of 16 bit values each with a maximum value
		of 65536. If there are more than 65536 vertices in the list,
		results of this operation are not defined.
		\param vertices: Pointer to array of vertices.
		\param vertexCount: Amount of vertices in the array.
		\param indexList: Pointer to array of indizes.
		\param triangleCount: amount of Triangles.
		\param vType: Vertex type, e.g. EVT_STANDARD for S3DVertex.
		\param pType: Primitive type, e.g. EPT_TRIANGLE_FAN for a triangle fan. */
		virtual void drawVertexPrimitiveList(const void* vertices, u32 vertexCount, const u16* indexList, u32 triangleCount, E_VERTEX_TYPE vType, scene::E_PRIMITIVE_TYPE pType) = 0;

		//! Draws an indexed triangle list.
		/** Note that there may be at maximum 65536 vertices, because the
		index list is a array of 16 bit values each with a maximum value
		of 65536. If there are more than 65536 vertices in the list,
		results of this operation are not defined.
		\param vertices: Pointer to array of vertices.
		\param vertexCount: Amount of vertices in the array.
		\param indexList: Pointer to array of indizes.
		\param triangleCount: amount of Triangles. Usually amount of indizes / 3. */
		virtual void drawIndexedTriangleList(const S3DVertex* vertices,
			u32 vertexCount, const u16* indexList, u32 triangleCount) = 0;

		//! Draws an indexed triangle list.
		/** Note that there may be at maximum 65536 vertices, because the
		index list is a array of 16 bit values each with a maximum value
		of 65536. If there are more than 65536 vertices in the list,
		results of this operation are not defined.
		\param vertices: Pointer to array of vertices.
		\param vertexCount: Amount of vertices in the array.
		\param indexList: Pointer to array of indizes.
		\param triangleCount: amount of Triangles. Usually amount of indizes / 3.*/
		virtual void drawIndexedTriangleList(const S3DVertex2TCoords* vertices,
			u32 vertexCount, const u16* indexList, u32 triangleCount) = 0;

		//! Draws an indexed triangle list.
		/** Note that there may be at maximum 65536 vertices, because the
		index list is a array of 16 bit values each with a maximum value
		of 65536. If there are more than 65536 vertices in the list,
		results of this operation are not defined.
		\param vertices: Pointer to array of vertices.
		\param vertexCount: Amount of vertices in the array.
		\param indexList: Pointer to array of indizes.
		\param triangleCount: amount of Triangles. Usually amount of indizes / 3. */
		virtual void drawIndexedTriangleList(const S3DVertexTangents* vertices,
			u32 vertexCount, const u16* indexList, u32 triangleCount) = 0;

		//! Draws an indexed triangle fan.
		/** Note that there may be at maximum 65536 vertices, because the
		index list is a array of 16 bit values each with a maximum value
		of 65536. If there are more than 65536 vertices in the list,
		results of this operation are not defined.
		Please note that some of the implementation code for this method is based on
		free code sent in by Mario Gruber, lots of thanks go to him!
		\param vertices: Pointer to array of vertices.
		\param vertexCount: Amount of vertices in the array.
		\param indexList: Pointer to array of indizes.
		\param triangleCount: amount of Triangles. Usually amount of indizes - 2. */
		virtual void drawIndexedTriangleFan(const S3DVertex* vertices,
			u32 vertexCount, const u16* indexList, u32 triangleCount) = 0;

		//! Draws an indexed triangle fan.
		/** Note that there may be at maximum 65536 vertices, because the
		index list is a array of 16 bit values each with a maximum value
		of 65536. If there are more than 65536 vertices in the list,
		results of this operation are not defined.
		Please note that some of the implementation code for this method is based on
		free code sent in by Mario Gruber, lots of thanks go to him!
		\param vertices: Pointer to array of vertices.
		\param vertexCount: Amount of vertices in the array.
		\param indexList: Pointer to array of indizes.
		\param triangleCount: amount of Triangles. Usually amount of indizes - 2. */
		virtual void drawIndexedTriangleFan(const S3DVertex2TCoords* vertices,
			u32 vertexCount, const u16* indexList, u32 triangleCount) = 0;

		//! Draws a 3d line.
		/** For some implementations, this method simply calls drawIndexedTriangles with some
		triangles. Note that the line is drawn using the current transformation
		matrix and material. So if you need to draw the 3D line independently of the
		current transformation, use
		\code
		driver->setTransform(video::ETS_WORLD, core::matrix4());
		\endcode
		before drawing the line.
		This method was created for making culling debugging easier.
		\param start: Start of the 3d line.
		\param end: End of the 3d line.
		\param color: Color of the line.  */
		virtual void draw3DLine(const core::vector3df& start,
			const core::vector3df& end, SColor color = SColor(255,255,255,255)) = 0;

		//! Draws a 3d triangle.
		/** This method usually simply calls drawIndexedTriangles with some
		triangles. Note that the line is drawn using the current transformation
		matrix and material.
		This method was created for making collision debugging easier. It works with
		all drivers because it does simply a call to drawIndexedTriangleList and
		hence is not very fast but it might be useful for further development.
		\param triangle: The triangle to draw.
		\param color: Color of the line. */
		virtual void draw3DTriangle(const core::triangle3df& triangle,
			SColor color = SColor(255,255,255,255)) = 0;

		//! Draws a 3d axis aligned box.
		/** This method usually simply calls drawIndexedTriangles with some
		triangles. Note that the line is drawn using the current transformation
		matrix and material.
		This method was created for making culling debugging easier. It works with
		all drivers because it does simply a call to drawIndexedTriangleList and
		hence is not very fast but it might be useful for further development.
		\param box: The axis aligned box to draw
		\param color: Color to use while drawing the box. */
		virtual void draw3DBox(const core::aabbox3d<f32>& box,
			SColor color = SColor(255,255,255,255)) = 0;

		//! Simply draws a 2d image without any special effects
		/** \param texture: Pointer to texture to use.
		\param destPos: upper left 2d destination position where the image will be drawn. */
		virtual void draw2DImage(video::ITexture* texture,
			const core::position2d<s32>& destPos) = 0;

		//! Draws a 2d image using a color
		/** (if color is other than
		Color(255,255,255,255)) and the alpha channel of the texture if wanted.
		\param texture: Texture to be drawn.
		\param destPos: Upper left 2d destination position where the image will be drawn.
		\param sourceRect: Source rectangle in the image.
		\param clipRect: Pointer to rectangle on the screen where the image is clipped to.
		This pointer can be NULL. Then the image is not clipped.
		\param color: Color with which the image is colored. If the color equals
		Color(255,255,255,255), the color is ignored. Note that the alpha component
		is used: If alpha is other than 255, the image will be transparent.
		\param useAlphaChannelOfTexture: If true, the alpha channel of the texture is
		used to draw the image.*/
		virtual void draw2DImage(video::ITexture* texture, const core::position2d<s32>& destPos,
			const core::rect<s32>& sourceRect, const core::rect<s32>* clipRect = 0,
			SColor color=SColor(255,255,255,255), bool useAlphaChannelOfTexture=false) = 0;

		//! draws a set of 2d images, using a color and the alpha
		/** channel of the texture if desired. The images are drawn
		beginning at pos and concatenated in one line. All drawings
		are clipped against clipRect (if != 0).
		The subtextures are defined by the array of sourceRects
		and are chosen by the indices given.
		\param texture: Texture to be drawn.
		\param pos: Upper left 2d destination position where the image will be drawn.
		\param sourceRects: Source rectangles of the image.
		\param indices: List of indices which choose the actual rectangle used each time.
		\param kerningWidth: Offset to Position on X
		\param clipRect: Pointer to rectangle on the screen where the image is clipped to.
		This pointer can be 0. Then the image is not clipped.
		\param color: Color with which the image is colored.
		Note that the alpha component is used: If alpha is other than 255, the image will be transparent.
		\param useAlphaChannelOfTexture: If true, the alpha channel of the texture is
		used to draw the image. */
		virtual void draw2DImage(video::ITexture* texture,
				const core::position2d<s32>& pos,
				const core::array<core::rect<s32> >& sourceRects,
				const core::array<s32>& indices,
				s32 kerningWidth=0,
				const core::rect<s32>* clipRect=0,
				SColor color=SColor(255,255,255,255),
				bool useAlphaChannelOfTexture=false) = 0;

		//! Draws a part of the texture into the rectangle.
		/** Suggested and first implemented by zola.
		\param texture: the texture to draw from
		\param destRect: the rectangle to draw into
		\param sourceRect: the rectangle denoting a part of the texture
		\param clipRect: clips the destination rectangle (may be 0)
		\param colors: array of 4 colors denoting the color values of the corners of the destRect
		\param useAlphaChannelOfTexture: true if alpha channel will be blended. */
		virtual void draw2DImage(video::ITexture* texture, const core::rect<s32>& destRect,
			const core::rect<s32>& sourceRect, const core::rect<s32>* clipRect = 0,
			video::SColor* colors=0, bool useAlphaChannelOfTexture=false) = 0;

		//!Draws an 2d rectangle.
		/** \param color: Color of the rectangle to draw. The alpha component will not
		be ignored and specifies how transparent the rectangle will be.
		\param pos: Position of the rectangle.
		\param clip: Pointer to rectangle against which the rectangle will be clipped.
		If the pointer is null, no clipping will be performed. */
		virtual void draw2DRectangle(SColor color, const core::rect<s32>& pos,
			const core::rect<s32>* clip = 0) = 0;

		//!Draws an 2d rectangle with a gradient.
		/** \param colorLeftUp: Color of the left upper corner to draw.
		The alpha component will not be ignored and specifies how transparent the rectangle will be.
		\param colorRightUp: Color of the right upper corner to draw.
		The alpha component will not be ignored and specifies how transparent the rectangle will be.
		\param colorLeftDown: Color of the left lower corner to draw.
		The alpha component will not be ignored and specifies how transparent the rectangle will be.
		\param colorRightDown: Color of the right lower corner to draw.
		The alpha component will not be ignored and specifies how transparent the rectangle will be.
		\param pos: Position of the rectangle.
		\param clip: Pointer to rectangle against which the rectangle will be clipped.
		If the pointer is null, no clipping will be performed. */
		virtual void draw2DRectangle(const core::rect<s32>& pos,
			SColor colorLeftUp, SColor colorRightUp, SColor colorLeftDown, SColor colorRightDown,
			const core::rect<s32>* clip = 0) = 0;

		//! Draws a 2d line.
		/** \param start: Screen coordinates of the start of the line in pixels.
		\param end: Screen coordinates of the start of the line in pixels.
		\param color: Color of the line to draw. */
		virtual void draw2DLine(const core::position2d<s32>& start,
					const core::position2d<s32>& end,
					SColor color=SColor(255,255,255,255)) = 0;

		//! Draws a non filled concyclic regular 2d polyon.
		/** This method can be used to draw
		circles, but also triangles, tetragons, pentagons, hexagons, heptagons, octagons,
		enneagons, decagons, hendecagons, dodecagon, triskaidecagons, etc. I think you'll
		got it now. And all this by simply specifying the vertex count. Welcome to the
		wonders of geometry.
		\param center: Position of center of circle (pixels).
		\param radius: Radius of circle in pixels.
		\param color: Color of the circle.
		\param vertexCount: Amount of vertices of the polygon. Specify 2 to draw a line,
		3 to draw a triangle, 4 for a tetragons and a lot (>10) for nearly a circle. */
		virtual void draw2DPolygon(core::position2d<s32> center,
			f32 radius, video::SColor color=SColor(100,255,255,255), s32 vertexCount=10) = 0;

		//! Draws a shadow volume into the stencil buffer.
		/** To draw a stencil shadow, do
		this: Frist, draw all geometry. Then use this method, to draw the shadow
		volume. Then, use IVideoDriver::drawStencilShadow() to visualize the shadow.
		Please note that the code for the opengl version of the method is based on
		free code sent in by Philipp Dortmann, lots of thanks go to him!
		\param triangles: Pointer to array of 3d vectors, specifing the shadow volume.
		\param count: Amount of triangles in the array.
		\param zfail: If set to true, zfail method is used, otherwise zpass. */
		virtual void drawStencilShadowVolume(const core::vector3df* triangles, s32 count, bool zfail=true) = 0;

		//! Fills the stencil shadow with color.
		/** After the shadow volume has been drawn
		into the stencil buffer using IVideoDriver::drawStencilShadowVolume(), use this
		to draw the color of the shadow.
		Please note that the code for the opengl version of the method is based on
		free code sent in by Philipp Dortmann, lots of thanks go to him!
		\param clearStencilBuffer: Set this to false, if you want to draw every shadow
		with the same color, and only want to call drawStencilShadow() once after all
		shadow volumes have been drawn. Set this to true, if you want to paint every
		shadow with its own color.
		\param leftUpEdge: Color of the shadow in the upper left corner of screen.
		\param rightUpEdge: Color of the shadow in the upper right corner of screen.
		\param leftDownEdge: Color of the shadow in the lower left corner of screen.
		\param rightDownEdge: Color of the shadow in the lower right corner of screen. */
		virtual void drawStencilShadow(bool clearStencilBuffer=false,
			video::SColor leftUpEdge = video::SColor(255,0,0,0),
			video::SColor rightUpEdge = video::SColor(255,0,0,0),
			video::SColor leftDownEdge = video::SColor(255,0,0,0),
			video::SColor rightDownEdge = video::SColor(255,0,0,0)) = 0;

		//! Draws a mesh buffer
		/** \param mb: Buffer to draw; */
		virtual void drawMeshBuffer( const scene::IMeshBuffer* mb) = 0;

		//! Sets the fog mode.
		/** These are global values attached to each 3d object
		rendered, which has the fog flag enabled in its material.
		\param color: Color of the fog
		\param linearFog: Set this to true for linear fog, otherwise exponential fog is applied.
		\param start: Only used in linear fog mode (linearFog=true). Specifies where fog starts.
		\param end: Only used in linear fog mode (linearFog=true). Specifies where fog ends.
		\param density: Only used in expotential fog mode (linearFog=false). Must be a value between 0 and 1.
		\param pixelFog: Set this to false for vertex fog, and true if you want pixel fog.
		\param rangeFog: Set this to true to enable range-based vertex fog. The distance
		from the viewer is used to compute the fog, not the z-coordinate. This is
		better, but slower. This is only available with D3D and vertex fog. */
		virtual void setFog(SColor color=SColor(0,255,255,255), bool linearFog=true, f32 start=50.0f, f32 end=100.0f,
			f32 density=0.01f, bool pixelFog=false, bool rangeFog=false) = 0;

		//! Returns the size of the screen or render window.
		/** \return Size of screen or render window. */
		virtual core::dimension2d<s32> getScreenSize() = 0;

		//! Returns the size of the current render target, or the screen size if the driver 
		//! doesnt support render to texture
		/** \return Size of render target or screen/window */
		virtual core::dimension2d<s32> getCurrentRenderTargetSize() = 0;


		//! Returns current frames per second value.
		/** \return Returns amount of frames per second drawn. **/
		virtual s32 getFPS() = 0;

		//! Returns amount of primitives (mostly triangles) which were drawn in the last frame.
		/** Together with getFPS() very useful method for statistics.
		\return Amount of primitives drawn in the last frame. */
		virtual u32 getPrimitiveCountDrawn( u32 param = 0 ) = 0;

		//! Deletes all dynamic lights which were previously added with addDynamicLight().
		virtual void deleteAllDynamicLights() = 0;

		//! Adds a dynamic light.
		/** \param light: Data specifing the dynamic light. */
		virtual void addDynamicLight(const SLight& light) = 0;

		//! Sets the dynamic ambient light color.
		/** The default color is
		(0,0,0,0) which means it is dark.
		\param color: New color of the ambient light. */
		virtual void setAmbientLight(const SColorf& color) = 0;

		//! Returns the maximal amount of dynamic lights the device can handle
		/** \return Maximal amount of dynamic lights. */
		virtual u32 getMaximalDynamicLightAmount() = 0;

		//! Returns current amount of dynamic lights set
		/** \return Current amount of dynamic lights set */
		virtual u32 getDynamicLightCount() = 0;

		//! Returns light data which was previously set with IVideDriver::addDynamicLight().
		/** \param idx: Zero based index of the light. Must be greater than 0 and smaller
		than IVideoDriver()::getDynamicLightCount.
		\return Light data. */
		virtual const SLight& getDynamicLight(u32 idx) = 0;

		//! Gets name of this video driver.
		/** \return Returns the name of the video driver. Example: In case of the Direct3D8
		driver, it would return "Direct3D 8.1". */
		virtual const wchar_t* getName() = 0;

		//! Adds an external image loader to the engine.
		/** This is useful if
		the Irrlicht Engine should be able to load textures of currently
		unsupported file formats (e.g .gif). The IImageLoader only needs
		to be implemented for loading this file format. A pointer to
		the implementation can be passed to the engine using this method.
		\param loader: Pointer to the external loader created. */
		virtual void addExternalImageLoader(IImageLoader* loader) = 0;

		//! Adds an external image writer to the engine.
		/** This is useful if
		the Irrlicht Engine should be able to write textures of currently
		unsupported file formats (e.g .gif). The IImageWriter only needs
		to be implemented for writing this file format. A pointer to
		the implementation can be passed to the engine using this method.
		\param writer: Pointer to the external writer created. */
		virtual void addExternalImageWriter(IImageWriter* writer) = 0;

		//! Returns the maximum amount of primitives
		/** (mostly vertices) which
		the device is able to render with one drawIndexedTriangleList
		call. */
		virtual u32 getMaximalPrimitiveCount() = 0;

		//! Enables or disables a texture creation flag.
		/** This flag defines how
		textures should be created. By changing this value, you can influence for example
		the speed of rendering a lot. But please note that the video drivers
		take this value only as recommendation. It could happen that you
		enable the ETCM_ALWAYS_16_BIT mode, but the driver creates 32 bit
		textures.
		\param flag: Texture creation flag.
		\param enabled: Specifies if the given flag should be enabled or disabled.*/
		virtual void setTextureCreationFlag(E_TEXTURE_CREATION_FLAG flag, bool enabled) = 0;

		//! Returns if a texture creation flag is enabled or disabled.
		/** You can change this value using setTextureCreationMode().
		\param flag: Texture creation flag.
		\return Returns the current texture creation mode. */
		virtual bool getTextureCreationFlag(E_TEXTURE_CREATION_FLAG flag) = 0;

		//! Creates a software image from a file.
		/** No hardware texture will
		be created for this image. This method is useful for example if
		you want to read a heightmap for a terrain renderer.
		\param filename: Name of the file from which the image is created.
		\return Returns the created image.
		If you no longer need the image, you should call IImage::drop().
		See IUnknown::drop() for more information. */
		virtual IImage* createImageFromFile(const c8* filename) = 0;

		//! Creates a software image from a file.
		/** No hardware texture will
		be created for this image. This method is useful for example if
		you want to read a heightmap for a terrain renderer.
		\param file: File from which the image is created.
		\return Returns the created image.
		If you no longer need the image, you should call IImage::drop().
		See IUnknown::drop() for more information. */
		virtual IImage* createImageFromFile(io::IReadFile* file) = 0;

		//! Writes the provided image to disk file
		/** Requires that there is a suitable image writer
		registered for writing the image to disk
		\param image: Image to write to disk
		\param filename: name of the file to write
		\param param: control parameter for the backend ( eq. compression level )
		\return Returns true on success */
		virtual bool writeImageToFile(IImage* image, const c8* filename, u32 param = 0) = 0;

		//! Creates a software image from a byte array.
		/** No hardware texture will
		be created for this image. This method is useful for example if
		you want to read a heightmap for a terrain renderer.
		\param format: Desired color format of the texture
		\param size: Desired size of the image
		\param data: A byte array with pixel color information
		\param ownForeignMemory: If true, the image will use the data pointer
		directly and own it afterwards.
		If false the memory will by copied internally.
		\param deleteMemory: Whether the memory is deallocated upon destruction
		\return Returns the created image.
		If you no longer need the image, you should call IImage::drop().
		See IUnknown::drop() for more information. */
		virtual IImage* createImageFromData(ECOLOR_FORMAT format,
			const core::dimension2d<s32>& size, void *data,
			bool ownForeignMemory=false,
			bool deleteMemory = true) = 0;

		//! Only used by the internal engine.
		/** Used to notify the driver that
		the window was resized. Usually, there is no need to call this method. */
		virtual void OnResize(const core::dimension2d<s32>& size) = 0;

		//! Adds a new material renderer to the video device.
		/** Use this method to extend the VideoDriver with new MaterialTypes. To extend the
		engine using this method do the following:
		Derive a class from IMaterialRenderer and override the methods you need. For
		setting the right renderstates, you can try to get a pointer to the real rendering device
		using IVideoDriver::getExposedVideoData(). Add your class with
		IVideoDriver::addMaterialRenderer(). To use an object being displayed
		with your new material set the MaterialType member of the SMaterial struct to the
		value returned by this method.
		If you simply want to create a new material using vertex and/or pixel shaders it would
		be easier to use the video::IGPUProgrammingServices interface which you can get using
		the getGPUProgrammingServices() method.
		\param renderer: A pointer to the new renderer.
		\param name: Optional name for this registered material renderer entry.
		\return Returns the number of the
		material type which can be set in SMaterial::MaterialType to use the renderer.
		-1 is returned if an error occured. (For example if you tried to add
		an material renderer to the software renderer or the null device, which do not accept
		material renderers.) */
		virtual s32 addMaterialRenderer(IMaterialRenderer* renderer, const c8* name = 0) = 0;

		//! Returns pointer to material renderer or null if not existing.
		/** \param idx: Id of the material renderer. Can be a value of the E_MATERIAL_TYPE enum or a
		value which was returned by addMaterialRenderer(). */
		virtual IMaterialRenderer* getMaterialRenderer(u32 idx) = 0;

		//! Returns amount of currently available material renderers.
		virtual u32 getMaterialRendererCount() = 0;

		//! Returns name of the material renderer
		/** This string can for example be used to test if a specific renderer already has
		been registered/created, or use this string to store data about materials: This
		returned name will be also used when serializing Materials.
		\param idx: Id of the material renderer. Can be a value of the E_MATERIAL_TYPE enum or a
		value which was returned by addMaterialRenderer(). */
		virtual const c8* getMaterialRendererName(u32 idx) = 0;

		//! Sets the name of a material renderer.
		/** Will have no effect on built-in material renderers.
		\param idx: Id of the material renderer. Can be a value of the E_MATERIAL_TYPE enum or a
		value which was returned by addMaterialRenderer().
		\param name: New name of the material renderer. */
		virtual void setMaterialRendererName(s32 idx, const c8* name) = 0;

		//! Creates material attributes list from a material, usable for serialization and more.
		/** Please note that the videodriver will use the material renderer names from
		getMaterialRendererName() to write out the material type name, so they should be set before. */
		virtual io::IAttributes* createAttributesFromMaterial(const video::SMaterial& material) = 0;

		//! Fills an SMaterial structure from attributes.
		/** Please note that for setting material types of the material, the video driver
		will need to query the material renderers for their names, so all non built-in materials must have been created before
		calling this method. */
		virtual void fillMaterialStructureFromAttributes(video::SMaterial& outMaterial, io::IAttributes* attributes) = 0;

		//! Returns driver and operating system specific data about the IVideoDriver.
		/** This method should only be used if the engine should be extended without having
		to modify the source of the engine. */
		virtual const SExposedVideoData& getExposedVideoData() = 0;

		//! Returns type of video driver
		virtual E_DRIVER_TYPE getDriverType() = 0;

		//! Returns pointer to the IGPUProgrammingServices interface.
		/** Returns 0 if the
		videodriver does not support this. (For example the Software and the NULL device
		will always return 0) */
		virtual IGPUProgrammingServices* getGPUProgrammingServices() = 0;

		//! Clears the ZBuffer.
		/** Note that you usually need not to call this method,
		This is done automaticly during IVideoDriver::beginScene() or
		IVideoDriver::setRenderTarget() if you specify zBuffer=true.
		But if you have to render some special things, you can clear the zbuffer
		during the rendering process with this method another time. */
		virtual void clearZBuffer() = 0;

		//! Returns an image created from the last rendered frame.
		virtual IImage* createScreenShot() = 0;
	};

} // end namespace video
} // end namespace irr


#endif