View Full Version : Insane new project for my own amusement

10-28-2005, 01:04 AM
After several attempts trying to get the Sample Implementation to compile with Cygwin, and after vainly trying to modifysmall portions of code to accomplish that task, I've decided to do a complete rewrite using the RI as a reference (partly), basing the work on the 1.0 Specification (mostly). I've already gotten a good portion written in the last few days. The purpose behind this project will to be to get a working version to compile and run (and debug, of course) under Windows using Cygwin and on generic POSIX-style environments (which, of course, includes Linux) using the standard "configure/make" system.

The only thing I know about the legal issues of OpenML is from the "Lincensing & Logos" link off the main page, which states this:

Any developer or company can freely download the OpenML Specification & SDK and implement and ship products using OpenML completely free of charge, royalty or licensing. Use of the OpenML Word or Logo Trademark however has certain requirements.

There is also the lincense header at the top of every source file which makes statements that seem to contradict the above one. Which should I follow? This came up as I was trying to write a lincese header for my own project. Mine was to be a GNU flavour (LGPL), and questions abound whether this will jive with the will of Khronos. Also, if this is to be a complete re-write, then could it be that the API falls under a different license than the SI?

Before venturing any further I thought I'd ask Khronos explicitly whether I'm stepping on any legal issues (is the SI "license free" as in 'public domain', I can do what I want with it?) This is me doing that, while accomplishing the dual purpose of announcing my intentions to the community in the only way I know how: by public forum.

Please answer these questions and offer some advice/criticism; shorty, if at all possible, since I will continue my work on this project without much rest.

01-22-2006, 11:34 AM
Just in case anyone thought I wasn't serious, I thought I'd make a premature release announcement.

I've been working on this not-so-insane new project for a good portion of my free time ever since I first made this post. My implementation of the OpenML API is reaching Beta phase and I've begun working on modules as well as the MLU and MLdc libraries.

I'm predicting completion before the summer with both features and preformance surpassing the current RI provided by Khronos. The limiting factor in the project development so far is that it has only been tested on Windows using Cygwin and MinGW. I will soon make the switch to a Unix system where development will continue with X11 integeration.

The license is a basic standard disclaimer. Forks can be made for commercial use which are closed source, but the open sourced version will continue to be developed. This is my first step in a larger project idea to integrate several media API's that complement each other.

If anyone is interested in the development project they are encouraged to contact me jessemaurais@gmail.com

The source code can currently be found on my website http://www.geocities.com/fragtheplanet/Projects/OML-SI-0.3.3.zip Sourceforge account pending.

02-04-2006, 12:59 PM
I've been debugging like mad over the last week and have made quite a bit of progress in that area. I need to test that the queue and fifo works, but everything else seems to do what it should. The C++ wrappers for the DIDD layer are also stable.

I have a second PC setup with Fedora Core 4 installed so that I can work on the Posix port (which has been largely neglected so far).

Everything is going on hold until I construct a proper build system for the library. I considered using SCons but in the end it's evident that the Autotools+libtool route is the way to go. I'm still keeping the Dev-C++ project files around for the Windows build. I attempted creating KDevelop and Anjuta project files but soon found out how bunk those IDE's are for projects such as this.

I'll be writting Doxygen style comments into the source code, but that is low priority right now.

Long range goals include integrating this project with other similar tools, such as OpenAL, and possibly porting the API to the Mono .Net platform.

I have a registered Sourceforge account but haven't posted any work to it yet. I won't until I'm certain of a reasonably stable release build.

There is alot of work ahead of me and not enough time in the day. So I won't even dream of offering a timetable, but as a loose guess I think that the first stable release will be ready around March.

p.s. I decided to move away from the GNU Public License and have instead opted for an MIT style license allowing for closed source forks of this work.