Wednesday, April 8, 2015

The long journey to choosing an engine

Today I have uploaded the project to github. We are going to use Unity3D 5, which is kinda weird choice for an open source game. Why do we took this decision? And I say we because the team also took part in the process.

When I started (or restarted the project), I wanted to use a complete game engine. After 4 years of developing with Ogre, I was totally sure that I had broke with rendering engines. The obvious choices, in the end of 2014, were Urho3D and Godot, I already had discarded Torque3D.

Urho3D seemed good after a few weeks of testing. It had an editor, not integrated into the game, but quite good. Development was active, renderer, even when based on Dx 9/OpenGL 2.1, was good. It lacked important stuff, like terrain editor, but I was sure that it would come at some point. But a couple of months later, well into the development process, I noticed that Urho lacked some sort of integration. A few things didnt work so tight as I expected.

It was time to look at Godot. But the missing feature were here obvious: Godot lacks terrain support and its 3D renderer is obsolete. A nice editor and the one click ddeployment feature cant compensate that. Also, the main effort now seems to be oriented towards 2D rendering.

Unreal Engine was discarded, as it explicitly forbids using it for GPLed titles.

Then I made a quick test with Unity3D 5 and my modellers were very insistent that it was the engine we needed. Also, considering our long term goal (create our own commercial game), we thought that investing time in Unity would pay better result in the future.

Thats why we decided to use Unity.

No comments:

Post a Comment