I’m a student from germany (Weimar). Currently i’m writing on a game engine based on OpenGL. The name of the Project is GEAR which means “Game Engine Andy Reinamm”.
For my project I’m currently searching for interested developers which have good practice in writing C++ code and OpenGL code.
The game engine is object oriented and based on a Scene Graph implementation. Axis aligned Bounding Box calculation is already implemented as well as basic shapes like a box, a plane and a SkyDome. More will follow (heightmap, particle systems…). I’m very comitted to write efficient OpenGL code which runs very fast on the GPU, because one of my next projects in the next weeks will be to implement a milion particle system which is already under development and test. This is a project which will be part of my Bachelor in the university.
Requirements for the project:
- [li]good C++ practicegood OpenGL practisewriting tidy and efficient code (+documentation)good ideasinterested
- [li]Platform is currently only windows (could be ported to linux later)Development with Visual StudioSource on SVN[*]OpenGL based on SDL
Hopefully there are any developers here which are interested in OpenGL and learning as well as in C++.
Up to now I already have a very well documented interface as an doxygen documentation online: Doxygen Documentation - GEAR
Note that only the code in the NameSpace GEAR is already part of the engine. The other stuff is still not added to the engine.
In the next few days I will create a wiki and the next milestone will be to create and write a few basic tutorials about how to use the interface to create basic stuff like camera movement and simple Meshes.
Documentation is the part i will take care as much as possible.
So its up to you to join the team
In the background i have a designer which has good practice in creating textures and so on.
If you’re interested let me know
I have to add a few things.
The wiki for the GEAR-Lib is now online GEAR-Wiki.
Now you can find more specific informations about the engine and the already implemented features.
Also an SVN Repository is now setup but currently it isnt ready to set public access so this will need a few hours.
Here some more informations:
- [li]SVN - Repository is now onlineA Logo is designed A tutorial to Download and Compile the GEAR-Library exists in the wiki[*] A Featurelist in the Wiki
Feel free to ask some questions
There are some Updates to this presented Project.
Many things have changed and now, the GEAR-Engine has many more features including support of several 3D-Model Formats, Realtime Intersection tests, a Wiki to stay informed about updates, and even more.
Take a look at the new Project site.
There are some Tutorials on the Site, but since my time isn’t enough to hold them up to date on each API change, there are currently not working.
But with a bit work, you can reprogram them and get them working.
When you do so, please let me know and send me the code to update them.
Feel free to ask some further questions.
Also if you’re interested in joining the GEAR Developer Team, please let me know.
But you should have enough time to do so and you shound also be very good in programing C++ and also OpenGL (advanced).
You can get further informations on the Project Page below or by simply sending an email to andy.reimann( at )uni-weimar.de.
Here is the link to the new GEAR Project Page
what’s the performance of your scenegraph like when culling and drawing lots of nodes?
how has performance affected your design choices?
cynically I expect the answers to be “no idea” and “it hasn’t”.
those design choices are more important than the use of VBO/bindless/instancing etc.
You should forget scenegraphs completely. They over-engineer and over-complicate a simple problem, and destroy performance.
If I were writing an engine these days, I’d base my renderer on something like this design:-
First of all I want to thank you for the critical words. It is very useful to hear about other solutions at all.
The visualization library is very cool and flexible, thats right.
But I also have not figured out any big bottleneck in a SceneGraph structure holding in mind that a Scenegraph isn’t the best solution, but it’s also not the slowest one.
In my opinion and with the experience from the GEAR-Project,with a clear design, they do not over-complicate a problem.
Anyway I will not lose sight of the other solution.
With all respect, this is not a place for game engine blogs. If anyone (and his own engine) will keep a blog as a forum thread here, there will be no room to place actual questions.
Create a blog somewhere and advertise it via news threads, if you need, but return here only for specific questions/answers.