UPDATE: fromgate and the Nukkit.Ru community have agreed to take over Voxelwind. I am wishing them the best of luck.
A week later, I purchased voxelwind.com and redirected it to the project's GitHub page. I later set up a simple website, forums and a Jenkins installation. Things went well, and I got accustomed to the MCPE community. Then it all came crashing down just 5 months later.
Why did Voxelwind die?
In short, disinterest killed Voxelwind.
My original goal
My goal for Voxelwind was simple: a scalable, high-quality MCPE server geared towards developers with an eye towards eventual end users using it, with better infrastructure than PocketMine but without the compatibility restraints of MiNET.
To achieve this goal, Voxelwind would take advantage of features like Netty's native support for Linux epoll, leverage native code for communicating with the Minecraft client, support multi-threaded worlds, and feature a good API.
Why did it go wrong?
In essence, my heart wasn't truly into it after about 4 months of work. I continued steadily, but it just wasn't enough for me to push forward. On top of that, I promised a lot, including a plugin to emulate vanilla behavior.
What could I have done differently?
- I should have researched MCPE much better.
- Voxelwind's inception was largely a side project and spur-of-the-moment. The moment I shared it on Twitter, I had lost.
Voxelwind is effectively dead now. For those who are interested in reviving the project, email me at write (-a-t-) imaginarycode (d.o.t) com and I would be happy to transfer ownership of the domain and GitHub repository to you.