cuda/examples/marching/openglada

OpenGLAda - Yet Another OpenGL binding for Ada

Overview

OpenGLAda is a thick OpenGL binding for the Ada 2005 programming language. Unlike other, thin, bindings (see the project's homepage for a list), OpenGLAda enriches the original API with concepts and features provided by Ada, like object orientation, type safety and generics.

Besides standard OpenGL functionality, OpenGLAda optionally provides bindings to the following OpenGL-related libraries:

• GLFW: This is a library for creating windows with an OpenGL context on them. It also provides functionality for capturing user input on keyboard, mouse and joystick. Having a window with an OpenGL context is the prerequisite for using any OpenGL functionality. The GLFW binding comes in two flavors: One for GLFW 2.x and one for GLFW 3+. There are significant differences between these two, the most prominent being that GLFW 3 can handle multiple windows. You can set the desired GLFW version for the binding at compile time.
• SOIL: The Simple OpenGL Image Library. This is a very tiny library for loading image files into OpenGL textures. It is public domain. Because it's so tiny, it is linked directly into OpenGLAda. Its source is included in the OpenGLAda sources.
• FTGL: A library built on top of FreeType that provides an API to load TrueType fonts and render text with OpenGL. The Ada wrapper only provides basic functionality to load fonts and render text. As it does not include a wrapper to FreeType, the more low-level functionality has been excluded.

OpenGLAda supports MacOSX, Windows and X11-based systems.

Prerequisites

In order to build OpenGLAda, you need to have:

• A GNAT compiler¹. The GNAT/GCC binary release from GnuAda is known to work as of November 2016 and is preferable if you do not fancy a GPL'd runtime. A GPL'd version is available on AdaCore's Libre Site, though it is known to have issues at least on OSX 10.11². More information is available on the GCC website.
• GPRBuild (is bundled with AdaCore's GNAT distribution). Minimum supported version is the one that comes with GNAT GPL 2012 (20120509). Do not use gnatmake to build the project files, it won't work.
• An OpenGL implementation (usually comes bundled with your graphics driver)
• Optionally GLFW
• Optionally FTGL

¹: You may also be able to build OpenGLAda with another Ada compiler and/or without using the *.gpr files. You just have to import the sources to your project and whichever build system you are using. I never used other Ada compilers apart from GNAT, so if I accidentally used some GNAT-specific features in the code, please drop me a message.

²: With GNAT GPL 2016, runtime loading of OpenGL function pointers does not seem to work on OSX 10.11, see this bug. It is unknown whether this affects other OSX versions or operation systems. The exact cause for this error is unknown. Therefore, using GNAT/GCC is currently strongly recommended.

Compilation

A Makefile is provided mainly for building the tests:

$ make tests

If you're on Windows and do not have the make utility available, you can build the test by executing

$ gprbuild -P glfw_test.gpr   -XWindowing_System=windows
$ gprbuild -P opengl_test.gpr -XWindowing_System=windows

The tests require GLFW, because they need to create windows. By default, they try to link against GLFW 3+. You can instead build the tests against GLFW 2.x by executing:

$ gprbuild -P opengl_test.gpr -XWindowing_System=windows -XGLFW_Version=2
$ gprbuild -P glfw_test.gpr   -XWindowing_System=windows -XGLFW_Version=2

(Substitute windows with x11 or quartz if needed.)

Detailed Instructions

Windows / GNAT GPL 2017 / GLFW 3

• Download and install GNAT GPL 2017 Windows x86 from AdaCore's Libre Site. We assume you install it to C:\GNAT\2017. Make sure the bin folder is in he path of your shell (whichever you are using).

• Download 32-bit Windows binaries from the GLFW download section (tested with GLFW 3.2.1).

• Open the zip folder and copy the file glfw3.dll from the lib-mingw folder into C:\GNAT\2017\lib.

• Open your shell and navigate to the root folder of OpenGLAda. Execute:

   gprbuild -P opengl_test.gpr -XWindowing_System=windows -XGLFW_Version=3
  

• This should produce executables in the bin folder inside OpenGLAda. Now copy the glfw3.dll to that bin folder.

• You should now be able to start the executables either by double-click using the explorer or from the shell.

• Keep in mind that you need to spread the glfw3.dll alongside your binaries for them to work.

Using OpenGLAda in your project

With GPRBuild

The easiest way to use OpenGLAda in your project is to just copy the sources in some dependency folder within your project folder, e.g.:

YourProject
 |
 |-dependencies
 |  |
 |  +-OpenGLAda
 |
 |-your
 |-project
 |-files
 +-...

If you're using the GPRBuild system, you can then just declare using OpenGLAda in your *.gpr file:

with "dependencies/OpenGLAda/opengl";

Alternatively, you can specify the path to the OpenGL project file in as environment variable:

export ADA_PROJECT_PATH=dependencies/OpenGLAda

... and then specify the dependency without the path:

with "opengl";

If you want to use GLFW, you also need to refer to the project opengl-glfw.gpr instead (it automatically adds a dependency to the opengl project).

The project files opengl.gpr and opengl-glfw.gpr take the following scenario parameters:

• Windowing_System: Sets the backend windowing system. Used for GLFW and also for system-dependent parts of the API (GLX, WGL, CGL):

  • x11: X Windowing System (Linux, BSD, etc)
  • windows: Microsoft Windows
  • quartz: Quartz Compositor (OS X)

• mode: May take one of the following values:

  • debug (default): Compile the project with debugging symbols and without optimization.
  • release: Compile the project for a release environment.

• GLFW_Version (GLFW only): Sets the version of the GLFW library to link against. See here for a detailed comparison of the two API versions.

  • 2: GLFW 2.x. Only one window.
  • 3: GLFW 3+. Multiple windows, multiple monitor support, etc.

• Auto_Exceptions: Configures exception handling:

  • enabled (default): After each call to OpenGL, OpenGLAda checks whether OpenGL has set an error flag and if it had, raises the corresponding exception.
  • disabled: The user has to query the error flag on his own.

With other build systems

You can add the OpenGLAda sources to your code, then use whatever build system you want. Just make sure that you link properly against your OpenGL implementation:

• OS X: -framework OpenGL -framework CoreFoundation
• Windows: -lOpenGL32 -lGdi32
• X11-based (Linux, BSD, etc): -lGL -lX11

If you're using GLFW, add -lglfw for GLFW 2 or -lglfw3 for GLFW 3. If you're on Windows and link against GLFW as dynamic library, you also need to add -lwinmm. If you're using FTGL, add -lftgl.

Installation

OpenGLAda is not designed to be installed as a standalone library. The reasoning behind this is that OpenGLAda is a wrapper library that doesn't provide much functionality on its own. Therefore, maintaining a library installing routine does not seem worth the effort - even less as it would be expected to support multiple platforms.

Tests

As mentioned, OpenGLAda contains some tests. You can also see them as examples that demonstrate the basic usage of the API. After building them as described above, you can execute them in the bin directory. Some tests load shader files from the source directory by using relative paths, so they only work with bin as working directory.

For additional information and documentation, see the project's homepage.

Developer Documentation

OpenGLAda autogenerates its API binding to OpenGL. The autogenerated source files are those in src/gl/generated and they are generated from the *.spec files in src/gl/specs. The syntax of the spec files is similar to Ada.

The reason behind this is that all functionality newer than OpenGL 1.1 is not expected to be provided by the OpenGL implementation. Instead, function pointers to the implementations should be queried at runtime. This makes it possible for the user to provide a fallback in case some OpenGL functionality is not available on the target system.

The generate tool compiled from src/generator will take care of creating both the API imports from OpenGL and the code for loading the function pointers at runtime. The tool is only necessary when adding OpenGL API functions to OpenGLAda and thus not of interest to the general user, since the autogenerated files are checked in to version control.

If you change the *.spec files, running make generate afterwards will update the autogenerated files. Be sure to check them in along with the spec changes.

License

OpenGLAda is distributed under the terms of the ISC License. The Ada 2012 logo that is used in the SOIL tests is distributed under the terms of the CC BY-ND 3.0 license, the original author is AdaCore.