timskillman
commented
5 years ago Hi Ninja - you certainly did your homework!
My goals for the library are;
1) To make it fast for Pi (i.e. optimised for the Pi) 2) Portable to Android/iOS/PC/MacOS/Linux etc.. 3) Implement effective scene management (such as portal culling) 4) Add basic physics (such as collision and deflection) 5) Provide a basic, script-based, fast UI system 6) Provide tools that are usable on the Pi (such as a 3D editor for model creation and map generation)
I believe there's enough power in the Pi to write some slick software built on speedy graphics ... most apps (if not all) were built on other systems ... so providing a fast engine for the Pi would be a good basis to write apps that don't chug!
You're very welcome to contribute - esp. if you can add some of the above mods and speed things up!
... as for other libraries and SDK's it would be great to offer support - esp. assets .. but as you point out, it may come down to licensing (although https://gamedev.stackexchange.com/questions/129767/can-free-unity-store-assets-be-used-in-non-unity-games/129922 makes some interesting points)
Perhaps a lightweight Unity editor could be written on the Pi? Hmm.
ghost
Hi this is NinjaFootball from youtube. Thanks for posting the link to source.
First some info that might be interesting. Have you heard of simd exists on pi? http://pkoenig10.github.io/pimd/
i.e. cloth https://software.intel.com/en-us/articles/soa-cloth-simulation-with-256-bit-intel-advanced-vector-extensions-intel-avx
Qs: Preface: I am new to the gles 2 on pi. I come from Unity3D background, XNA too but mostly c#. I am eager to try to use every possible piece of metal on a pi zero. I believe the real value to cost ratio is overcome if a pi zero is used to the max. if we look at pi 2/3 lots of competition. But pi zero, with reduced OpenCL, reduced OpenGL and well matched spec gles 2, and with the latest discovery that simd lives on all pi's too. So much can be done for $5 it is crazy. ref: https://anholt.github.io/twivc4/ https://www.raspberrypi.org/forums/viewtopic.php?t=194952
Question 1: What goals do you have for the library?
Question 2: What platforms do you plan to target?
Question 3: Have you had a chance to play with some other pi oriented game solutions. The major two being kha and doom 3 latest. OpenMW and blender game engine too. Steam half-life sdk and similar, The doom solution is unique as megatextures are a good fit for pi, esp. in cloud streaming scenario.
If tried any of these, would you be open to porting aspects here? Existing work, esp. if only best parts are ported offers fun pros like modding/workshop support on pi. Huge savings on testing advanced techniques that can be tricky on lightweight platforms. Story telling built in with sound, compression, fbx sdk's etc. Lots of stuff already road mapped. Tricky part is if okay with their licensing. At any rate I was curious if you had thoughts on these other libraries.
Lastly, your work seems to be the most complete work I have found for pi. I am very excited to try it out and begin contributing. Thanks again for posting the source. Talk soon.