Up and Coming

Simple, one page, OpenGL example (Oculus SDK 0.3.2 Preview)

51 replies
11916 views

Hi,

Didn't take too long to modify the earlier example (thread) to work with the new 0.3.2 release.
I think the OGL.Window field in ovrGLConfig is just a leftover because you can pass NULL without any consequences. Luckily, at least, the OGL.WglContext and OGL.GdiDC fields are gone making it easier to do the OSX/Linux port. I'll will try the OSX version this afternoon.

EDIT: Still a bit of OpenGL state restoration to do after ovrHmd_EndFrame() if you want your own vertex arrays to work: GL_ELEMENT_ARRAY_BUFFER and GL_ARRAY_BUFFER.
EDIT: Modified so it should work on Linux (thanks yombo and rjoyce). I don't have a Linux machine so please check. Next up OSX!
EDIT: OSX "support" (big word, just some preprocessor defines). GLFW will create a 2.X context, at least on my machine, so the deprecated functions will work. If not let me know.

Thanks,
Daniel


// GLFWOculusRiftTest
// (c) cThrough 2014 (Daniel Dekkers)
// Version 2014052305 Based on OculusSDK 3.0.2 Preview
// Windows, Linux and OSX

#include <GL/glew.h>
#if defined(_WIN32)
	#include <Windows.h>
	#define GLFW_EXPOSE_NATIVE_WIN32
	#define GLFW_EXPOSE_NATIVE_WGL
#elif defined(__linux__)
	#include <X11/X.h>
	#include <X11/extensions/Xrandr.h>
	#define GLFW_EXPOSE_NATIVE_X11
	#define GLFW_EXPOSE_NATIVE_GLX
#endif
#include <GLFW/glfw3.h>
#if !defined(__APPLE__)
    #include <GLFW/glfw3native.h>
#endif
#include <OVR.h>
#include <OVR_CAPI.h>
#include <OVR_CAPI_GL.h>

const bool l_FullScreen = false;
const bool l_MultiSampling = false;
const bool l_Spin = true;

ovrHmd l_Hmd;
ovrHmdDesc l_HmdDesc;
ovrFovPort l_EyeFov[2];
ovrGLConfig l_Cfg;
ovrEyeRenderDesc l_EyeRenderDesc[2];

GLfloat l_VAPoints[] =
{ 
	 0.5f, 0.5f, 0.5f,
	-0.5f, 0.5f, 0.5f,
	-0.5f,-0.5f, 0.5f,
	 0.5f,-0.5f, 0.5f,
	-0.5f,-0.5f,-0.5f,
	-0.5f, 0.5f,-0.5f,
	 0.5f, 0.5f,-0.5f,
	 0.5f,-0.5f,-0.5f,
	 0.5f, 0.5f, 0.5f,
	 0.5f, 0.5f,-0.5f,
	-0.5f, 0.5f,-0.5f,
	-0.5f, 0.5f, 0.5f,
	-0.5f,-0.5f,-0.5f,
	 0.5f,-0.5f,-0.5f,
	 0.5f,-0.5f, 0.5f,
	-0.5f,-0.5f, 0.5f,
	 0.5f, 0.5f, 0.5f,
	 0.5f,-0.5f, 0.5f,
	 0.5f,-0.5f,-0.5f,
	 0.5f, 0.5f,-0.5f,
	-0.5f,-0.5f,-0.5f,
	-0.5f,-0.5f, 0.5f,
	-0.5f, 0.5f, 0.5f,
	-0.5f, 0.5f,-0.5f
};

GLfloat l_VANormals[] =
{
	 0.0f, 0.0f, 1.0f,
	 0.0f, 0.0f, 1.0f,
	 0.0f, 0.0f, 1.0f,
	 0.0f, 0.0f, 1.0f,
	 0.0f, 0.0f,-1.0f,
	 0.0f, 0.0f,-1.0f,
	 0.0f, 0.0f,-1.0f,
	 0.0f, 0.0f,-1.0f,
	 0.0f, 1.0f, 0.0f,
	 0.0f, 1.0f, 0.0f,
	 0.0f, 1.0f, 0.0f,
	 0.0f, 1.0f, 0.0f,
	 0.0f,-1.0f, 0.0f,
	 0.0f,-1.0f, 0.0f,
	 0.0f,-1.0f, 0.0f,
	 0.0f,-1.0f, 0.0f,
	 1.0f, 0.0f, 0.0f,
	 1.0f, 0.0f, 0.0f,
	 1.0f, 0.0f, 0.0f,
	 1.0f, 0.0f, 0.0f,
	-1.0f, 0.0f, 0.0f,
	-1.0f, 0.0f, 0.0f,
	-1.0f, 0.0f, 0.0f,
	-1.0f, 0.0f, 0.0f
};

GLuint l_VAIndici[] =
{
	0, 1, 2, 3,
	4, 5, 6, 7,
	8, 9, 10, 11,
	12, 13, 14, 15,
	16, 17, 18, 19,
	20, 21, 22, 23
};

// =============================================================================

static void ErrorCallback(int p_Error, const char* p_Description)
{
	fputs(p_Description, stderr);
}

// =============================================================================

static void KeyCallback(GLFWwindow* p_Window, int p_Key, int p_Scancode, int p_Action, int p_Mods)
{
	if (p_Key == GLFW_KEY_ESCAPE && p_Action == GLFW_PRESS) 
		glfwSetWindowShouldClose(p_Window, GL_TRUE);
}

// =============================================================================

static void WindowSizeCallback(GLFWwindow* p_Window, int p_Width, int p_Height)
{
	l_Cfg.OGL.Header.RTSize.w = p_Width; 
	l_Cfg.OGL.Header.RTSize.h = p_Height;

	int l_DistortionCaps = ovrDistortionCap_Chromatic | ovrDistortionCap_TimeWarp;
	ovrHmd_ConfigureRendering(l_Hmd, &l_Cfg.Config, l_DistortionCaps, l_EyeFov, l_EyeRenderDesc);
	glBindBuffer(GL_ELEMENT_ARRAY_BUFFER, 0);
	glBindBuffer(GL_ARRAY_BUFFER, 0);
	glUseProgram(0);
}

// ============================================================================

void RenderCubeVertexArrays(void)
{
	glEnableClientState(GL_VERTEX_ARRAY);
	glVertexPointer(3, GL_FLOAT, 0, l_VAPoints);

	glEnableClientState(GL_NORMAL_ARRAY);
	glNormalPointer(GL_FLOAT, 0, l_VANormals);

	glDrawElements(GL_QUADS, 6*4, GL_UNSIGNED_INT, l_VAIndici);

	glDisableClientState(GL_NORMAL_ARRAY);
	glDisableClientState(GL_VERTEX_ARRAY);
}

// ============================================================================

void RenderCubeFixedFunction(void)
{
	// Obsolete, remains as a fall back for the vertex arrays version...
	glBegin(GL_QUADS);
	glNormal3f( 0.0f, 0.0f, 1.0f);
	glVertex3f( 0.5f, 0.5f, 0.5f);
	glVertex3f(-0.5f, 0.5f, 0.5f);
	glVertex3f(-0.5f,-0.5f, 0.5f);
	glVertex3f( 0.5f,-0.5f, 0.5f);

	glNormal3f( 0.0f, 0.0f,-1.0f);
	glVertex3f(-0.5f,-0.5f,-0.5f);
	glVertex3f(-0.5f, 0.5f,-0.5f);
	glVertex3f( 0.5f, 0.5f,-0.5f);
	glVertex3f( 0.5f,-0.5f,-0.5f);

	glNormal3f( 0.0f, 1.0f, 0.0f);
	glVertex3f( 0.5f, 0.5f, 0.5f);
	glVertex3f( 0.5f, 0.5f,-0.5f);
	glVertex3f(-0.5f, 0.5f,-0.5f);
	glVertex3f(-0.5f, 0.5f, 0.5f);

	glNormal3f( 0.0f,-1.0f, 0.0f);
	glVertex3f(-0.5f,-0.5f,-0.5f);
	glVertex3f( 0.5f,-0.5f,-0.5f);
	glVertex3f( 0.5f,-0.5f, 0.5f);
	glVertex3f(-0.5f,-0.5f, 0.5f);

	glNormal3f( 1.0f, 0.0f, 0.0f);
	glVertex3f( 0.5f, 0.5f, 0.5f);
	glVertex3f( 0.5f,-0.5f, 0.5f);
	glVertex3f( 0.5f,-0.5f,-0.5f);
	glVertex3f( 0.5f, 0.5f,-0.5f);

	glNormal3f(-1.0f, 0.0f, 0.0f);
	glVertex3f(-0.5f,-0.5f,-0.5f);
	glVertex3f(-0.5f,-0.5f, 0.5f);
	glVertex3f(-0.5f, 0.5f, 0.5f);
	glVertex3f(-0.5f, 0.5f,-0.5f);
	glEnd();
}

// ============================================================================

static void SetOpenGLState(void)
{
	// Some state...
	glEnable(GL_CULL_FACE);
	glEnable(GL_LIGHTING);
	glDisable(GL_TEXTURE_2D);
	glEnable(GL_DEPTH_TEST);
	glShadeModel(GL_SMOOTH);
	glEnable(GL_BLEND);
	glBlendFunc(GL_SRC_ALPHA, GL_ONE_MINUS_SRC_ALPHA);
	if (l_MultiSampling) glEnable(GL_MULTISAMPLE);

	// Some (stationary) lights...
	GLfloat l_Light0Position[] = { 5.0f, 6.0f, 3.0f, 0.0f };
	GLfloat l_Light0Diffuse[] = { 1.0f, 0.8f, 0.6f, 1.0f };
	glLightfv(GL_LIGHT0, GL_POSITION, l_Light0Position);
	glLightfv(GL_LIGHT0, GL_DIFFUSE, l_Light0Diffuse);
	glEnable(GL_LIGHT0);

	GLfloat l_Light1Position[] = { -5.0f, -6.0f, 5.0f, 0.0f };
	GLfloat l_Light1Diffuse[] = { 0.6f, 0.8f, 1.0f, 1.0f };
	glLightfv(GL_LIGHT1, GL_POSITION, l_Light1Position);
	glLightfv(GL_LIGHT1, GL_DIFFUSE, l_Light1Diffuse);
	glEnable(GL_LIGHT1);

	// Material...
	GLfloat l_MaterialSpecular[] = { 0.3f, 0.3f, 0.3f, 1.0f };
	GLfloat l_MaterialShininess[] = { 10.0f };
	glMaterialfv(GL_FRONT, GL_SPECULAR, l_MaterialSpecular);
	glMaterialfv(GL_FRONT, GL_SHININESS, l_MaterialShininess);
}

// =============================================================================

int main(void)
{
	// Initialize LibOVR...
	ovr_Initialize();

	l_Hmd = ovrHmd_Create(0);
	if (!l_Hmd) l_Hmd = ovrHmd_CreateDebug(ovrHmd_DK1);

	ovrHmd_GetDesc(l_Hmd, &l_HmdDesc);

	// Start the sensor which provides the Rift’s pose and motion.
	ovrHmd_StartSensor(l_Hmd, ovrSensorCap_Orientation | ovrSensorCap_YawCorrection | ovrSensorCap_Position, ovrSensorCap_Orientation);

	GLFWwindow* l_Window;

	glfwSetErrorCallback(ErrorCallback);

	if (!glfwInit()) exit(EXIT_FAILURE);

	if (l_MultiSampling) glfwWindowHint(GLFW_SAMPLES, 4); else glfwWindowHint(GLFW_SAMPLES, 0);

	ovrSizei l_ClientSize;
	if (l_FullScreen)
	{
		l_ClientSize.w = l_HmdDesc.Resolution.w; // 1280 for DK1...
		l_ClientSize.h = l_HmdDesc.Resolution.h; // 800 for DK1...
		// Create a fullscreen window with the Oculus Rift resolution...
		l_Window = glfwCreateWindow(l_ClientSize.w, l_ClientSize.h, "GLFW Oculus Rift Test", glfwGetPrimaryMonitor(), NULL);
	}
	else
	{
		l_ClientSize.w = 640;
		l_ClientSize.h = 480;
		l_Window = glfwCreateWindow(l_ClientSize.w, l_ClientSize.h, "GLFW Oculus Rift Test", NULL, NULL);
	}

	if (!l_Window)
	{
		glfwTerminate();
		exit(EXIT_FAILURE);
	}

	// Make the context current for this window...
	glfwMakeContextCurrent(l_Window);

	// Don't forget to initialize Glew, turn glewExperimental on to avoid problem fetching function pointers...
    glewExperimental = GL_TRUE;
	GLenum l_Result = glewInit();
	if (l_Result!=GLEW_OK)   
	{
		printf("glewInit() error.\n");
		exit(EXIT_FAILURE);
	}

	// Print some info about the OpenGL context...
	int l_Major = glfwGetWindowAttrib(l_Window, GLFW_CONTEXT_VERSION_MAJOR);
	int l_Minor = glfwGetWindowAttrib(l_Window, GLFW_CONTEXT_VERSION_MINOR);
	int l_Profile = glfwGetWindowAttrib(l_Window, GLFW_OPENGL_PROFILE);
	printf("OpenGL: %d.%d ", l_Major, l_Minor);
	if (l_Major>=3) // Profiles introduced in OpenGL 3.0...
	{
		if (l_Profile==GLFW_OPENGL_COMPAT_PROFILE) printf("GLFW_OPENGL_COMPAT_PROFILE\n"); else printf("GLFW_OPENGL_CORE_PROFILE\n");
	}
	printf("Vendor: %s\n", (char*)glGetString(GL_VENDOR));
	printf("Renderer: %s\n", (char*)glGetString(GL_RENDERER));

	// Create some lights, materials, etc...
	SetOpenGLState();

	// We will do some offscreen rendering, setup FBO...
	ovrSizei l_TextureSizeLeft = ovrHmd_GetFovTextureSize(l_Hmd, ovrEye_Left, l_HmdDesc.DefaultEyeFov[0], 1.0f);
	ovrSizei l_TextureSizeRight = ovrHmd_GetFovTextureSize(l_Hmd, ovrEye_Right, l_HmdDesc.DefaultEyeFov[1], 1.0f);
	ovrSizei l_TextureSize;
	l_TextureSize.w = l_TextureSizeLeft.w + l_TextureSizeRight.w;
	l_TextureSize.h = (l_TextureSizeLeft.h>l_TextureSizeRight.h ? l_TextureSizeLeft.h : l_TextureSizeRight.h);

	// Create FBO...
	GLuint l_FBOId;
	glGenFramebuffers(1, &l_FBOId);
	glBindFramebuffer(GL_FRAMEBUFFER, l_FBOId);

	// The texture we're going to render to...
	GLuint l_TextureId;
	glGenTextures(1, &l_TextureId);
	// "Bind" the newly created texture : all future texture functions will modify this texture...
	glBindTexture(GL_TEXTURE_2D, l_TextureId);
	// Give an empty image to OpenGL (the last "0")
	glTexImage2D(GL_TEXTURE_2D, 0, GL_RGBA, l_TextureSize.w, l_TextureSize.h, 0, GL_RGBA, GL_UNSIGNED_BYTE, 0);
	// Linear filtering...
	glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_LINEAR);
	glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_LINEAR);

	// Create Depth Buffer...
	GLuint l_DepthBufferId;
	glGenRenderbuffers(1, &l_DepthBufferId);
	glBindRenderbuffer(GL_RENDERBUFFER, l_DepthBufferId);
	glRenderbufferStorage(GL_RENDERBUFFER, GL_DEPTH_COMPONENT, l_TextureSize.w, l_TextureSize.h);
	glFramebufferRenderbuffer(GL_FRAMEBUFFER, GL_DEPTH_ATTACHMENT, GL_RENDERBUFFER, l_DepthBufferId);

	// Set the texture as our colour attachment #0...
	glFramebufferTexture(GL_FRAMEBUFFER, GL_COLOR_ATTACHMENT0, l_TextureId, 0);

	// Set the list of draw buffers...
	GLenum l_GLDrawBuffers[1] = { GL_COLOR_ATTACHMENT0 };
	glDrawBuffers(1, l_GLDrawBuffers); // "1" is the size of DrawBuffers

	// Check if everything is OK...
	GLenum l_Check = glCheckFramebufferStatus(GL_DRAW_FRAMEBUFFER);
	if (l_Check!=GL_FRAMEBUFFER_COMPLETE)
	{
		printf("There is a problem with the FBO.\n");
		exit(EXIT_FAILURE);
	}

	// Unbind...
	glBindRenderbuffer(GL_RENDERBUFFER, 0);
	glBindTexture(GL_TEXTURE_2D, 0);
	glBindFramebuffer(GL_FRAMEBUFFER, 0);

	// Oculus Rift eye configurations...
	l_EyeFov[0] = l_HmdDesc.DefaultEyeFov[0];
	l_EyeFov[1] = l_HmdDesc.DefaultEyeFov[1];

	l_Cfg.OGL.Header.API = ovrRenderAPI_OpenGL;
	l_Cfg.OGL.Header.Multisample = (l_MultiSampling ? 1 : 0);
	l_Cfg.OGL.Header.RTSize.w = l_ClientSize.w;
	l_Cfg.OGL.Header.RTSize.h = l_ClientSize.h;
	#if defined(_WIN32)
		l_Cfg.OGL.Window = glfwGetWin32Window(l_Window);
	#elif defined(__linux__)
		l_Cfg.OGL.Win = glfwGetX11Window(l_Window);
		l_Cfg.OGL.Disp = glfwGetX11Display();
	#endif

	int l_DistortionCaps = ovrDistortionCap_Chromatic | ovrDistortionCap_TimeWarp;
	ovrHmd_ConfigureRendering(l_Hmd, &l_Cfg.Config, l_DistortionCaps, l_EyeFov, l_EyeRenderDesc);
	glBindBuffer(GL_ELEMENT_ARRAY_BUFFER, 0);
	glBindBuffer(GL_ARRAY_BUFFER, 0);
	glUseProgram(0);

	ovrGLTexture l_EyeTexture[2];
	l_EyeTexture[0].OGL.Header.API = ovrRenderAPI_OpenGL;
	l_EyeTexture[0].OGL.Header.TextureSize.w = l_TextureSize.w;
	l_EyeTexture[0].OGL.Header.TextureSize.h = l_TextureSize.h;
	l_EyeTexture[0].OGL.Header.RenderViewport.Pos.x = 0;
	l_EyeTexture[0].OGL.Header.RenderViewport.Pos.y = 0;
	l_EyeTexture[0].OGL.Header.RenderViewport.Size.w = l_TextureSize.w/2;
	l_EyeTexture[0].OGL.Header.RenderViewport.Size.h = l_TextureSize.h;
	l_EyeTexture[0].OGL.TexId = l_TextureId;

	// Right eye the same, except for the x-position in the texture...
	l_EyeTexture[1] = l_EyeTexture[0];
	l_EyeTexture[1].OGL.Header.RenderViewport.Pos.x = (l_TextureSize.w+1)/2;

	glfwSetKeyCallback(l_Window, KeyCallback);
	glfwSetWindowSizeCallback(l_Window, WindowSizeCallback);

	GLfloat l_SpinX;
	GLfloat l_SpinY;

	while (!glfwWindowShouldClose(l_Window))
	{
		if (l_Spin)
		{
			l_SpinX = (GLfloat) fmod(glfwGetTime()*17.0, 360.0);
			l_SpinY = (GLfloat) fmod(glfwGetTime()*23.0, 360.0);
		}
		else
		{
			l_SpinX = 30.0f;
			l_SpinY = 40.0f;
		}

		ovrFrameTiming m_HmdFrameTiming = ovrHmd_BeginFrame(l_Hmd, 0);

		// Bind the FBO...
		glBindFramebuffer(GL_FRAMEBUFFER, l_FBOId);
		// Clear...
		glClearColor(0.2f, 0.3f, 0.4f, 1.0f);
		glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT);

		for (int l_EyeIndex=0; l_EyeIndex<ovrEye_Count; l_EyeIndex++)
		{
			ovrEyeType l_Eye = l_HmdDesc.EyeRenderOrder[l_EyeIndex];
			ovrPosef l_EyePose = ovrHmd_BeginEyeRender(l_Hmd, l_Eye);

			glViewport(l_EyeTexture[l_Eye].OGL.Header.RenderViewport.Pos.x,
				l_EyeTexture[l_Eye].OGL.Header.RenderViewport.Pos.y,
				l_EyeTexture[l_Eye].OGL.Header.RenderViewport.Size.w,
				l_EyeTexture[l_Eye].OGL.Header.RenderViewport.Size.h
				);

			// Get Projection and ModelView matrici from the device...
			OVR::Matrix4f l_ProjectionMatrix = ovrMatrix4f_Projection(
				l_EyeRenderDesc[l_Eye].Fov, 0.3f, 100.0f, true);
			OVR::Quatf l_Orientation = OVR::Quatf(l_EyePose.Orientation);
			OVR::Matrix4f l_ModelViewMatrix = OVR::Matrix4f(l_Orientation.Inverted());

			// Pass matrici on to OpenGL...
			glMatrixMode(GL_PROJECTION);
			glLoadIdentity();
			glMultMatrixf(&(l_ProjectionMatrix.Transposed().M[0][0]));
			glMatrixMode(GL_MODELVIEW);
			glLoadIdentity();
			// Translate for specific eye based on IPD...
			glTranslatef(l_EyeRenderDesc[l_Eye].ViewAdjust.x,
				l_EyeRenderDesc[l_Eye].ViewAdjust.y,
				l_EyeRenderDesc[l_Eye].ViewAdjust.z);
			// Multiply with orientation retrieved from sensor...
			glMultMatrixf(&(l_ModelViewMatrix.Transposed().M[0][0]));
			// Move back a bit to show scene in front of us...
			glTranslatef(0.0f, 0.0f, -2.0f);
			// Make the cube spin...
			glRotatef(l_SpinX, 1.0f, 0.0f, 0.0f);
			glRotatef(l_SpinY, 0.0f, 1.0f, 0.0f);

			// Render...
			// RenderCubeFixedFunction();
			RenderCubeVertexArrays();

			ovrHmd_EndEyeRender(l_Hmd, l_Eye, l_EyePose, &l_EyeTexture[l_Eye].Texture);
		}

		// Unbind the FBO, back to normal drawing...
		glBindFramebuffer(GL_FRAMEBUFFER, 0);

		ovrHmd_EndFrame(l_Hmd);

		glBindBuffer(GL_ELEMENT_ARRAY_BUFFER, 0);
		glBindBuffer(GL_ARRAY_BUFFER, 0);
		glUseProgram(0);

		glfwPollEvents();
	}

	glfwDestroyWindow(l_Window);

	glfwTerminate();

	ovrHmd_Destroy(l_Hmd);
	ovr_Shutdown();

	exit(EXIT_SUCCESS);
}

Show previous replies

jherico

Inspiring

"rjoyce" wrote:
"DoZo1971" wrote:
I'm a bit stuck in porting this example to OSX.

The distortion renderer still doesn't work for core profiles so (forced move):

Oh wow I just read this again. My suggestion won't work as its the first thing you eliminate :roll:

Looks like OS X is going to be a pain to support for this...until we can run a core profile with the SDK I guess.

The SDK works with the core profile. On 0.3.1 I had to make a number of changes in the SDK to get it working, but 0.3.2 properly (if inexpertly) uses vertex array objects, so it's at the very least, functional.

Brad Davis - Developer for High Fidelity Co-author of Oculus Rift in Action

DoZo1971Author

Up and Coming

Hm.. ok. A working core profile. I tried yesterday and got a crash. But that, in retrospect, might be related to GLEW:
http://www.opengl.org/wiki/OpenGL_Loading_Library
Drop GLEW?
Furthermore. A core profile would result in extra code to setup the shaders and we can't use the gl[Transformation] calls. Not that I'm lazy, but the example would lose a lot of its simplicity.
Choices, choices...
I don't see hundreds of people who are developing for OSX on this forum anyway.

DoZo1971Author

Up and Coming

@jherico:
I forgot to mention that I find your multi-sampling approach very interesting. It should be in a separate thread. Can you tell us more about efficiency? There is loss at (at least) two levels I guess. The multi-sampled primary window, the increased texture size and some extra work in the distortion shader as well?
It shouldn't be too hard to profile. The multi-sampled primary window versus no multi-sampling would be a constant factor. And then efficiency decrease depending on the multiplication factor of the texture size. How big do you have to go to lose jaggies in the center to a "reasonable degree"? I'm really not into OpenGL enough to give other suggestions. Is a mipmapped texture an option to lose the high frequency noise at the edges?

DoZo1971Author

Up and Coming

@anarkarvre
Yes these bugs, eh features, are maddening.
I spent another hour because I forgot to unbind buffers *after* configureRendering() which was called early on in a OnSize() callback.

rjoyce

Participant

"DoZo1971" wrote:
Hm.. ok. A working core profile. I tried yesterday and got a crash. But that, in retrospect, might be related to GLEW:
http://www.opengl.org/wiki/OpenGL_Loading_Library
Drop GLEW?
Furthermore. A core profile would result in extra code to setup the shaders and we can't use the gl[Transformation] calls. Not that I'm lazy, but the example would lose a lot of its simplicity.
Choices, choices...
I don't see hundreds of people who are developing for OSX on this forum anyway.

Just tested on OS X, if you add glewExperimental = GL_TRUE to the code before the glewInit, it does not segfault anymore... but no cube shows up, and I imagine this is due to the glTranslate* calls being ignored in the Core profile, so the cube is not visible. But the Oculus SDK shader works just fine.

(You also have to add include guards around the glfw3native header as it is not needed on OS X, so I just added #ifndef __APPLE__ around it)

DoZo1971Author

Up and Coming

Hm... I tried again in Xcode with glewExperimental on. And it just... works. I don't understand why because there is a core profile active and I use a lot of deprecated functions.

OpenGL: 2.1 GLFW_OPENGL_CORE_PROFILE
Vendor: NVIDIA Corporation
Renderer: NVIDIA GeForce GT 750M OpenGL Engine

rjoyce

Participant

Hmm it let you create a 2.1 Core profile though, mine was 3.2. I'm new to OpenGL but wasn't the fixed pipeline part of the 2.1 "Core"? So that could be why it worked. I'm also testing on an old laptop, nvidia 320m I believe, so that might have something to do with it.
Either way looks like glewExperimental is the key, and if someone is serious about developing for OS X they should be able to figure it out now :D

jherico

Inspiring

"DoZo1971" wrote:
I don't understand why because there is a core profile active and I use a lot of deprecated functions.

OpenGL: 2.1 GLFW_OPENGL_CORE_PROFILE

There's no such thing as a 2.1 core profile (because there is no such thing as profiles prior to 3.0). Asking what profile you have before 3.0 is meaningless and you can't rely on the results. So what you have is an OpenGL 2.1 context, full stop.

Brad Davis - Developer for High Fidelity Co-author of Oculus Rift in Action

DoZo1971Author

Up and Coming

When you're right, you're right.
Modified the code so the profile is examined only when the major is larger/equal 3.
So,... OSX turns out to be the easiest one. Very good.

mosesr

Participant

Hi everyone,

When I try to build this example in Visual Studio 2012 I get this error:

C:\Users\Moshe\OculusRift\ovr_sdk_win_0.3.2\OculusSDK\LibOVR\Src\OVR_CAPI_GL.h(26): fatal error C1083: Cannot open include file: 'GL/glext.h': No such file or directory

Any idea what could be causing this?

Thanks a lot!

Show more replies

Get Help

Sign up

Login with SSO

Login to the community

Login with SSO