FIRST LOOK

My initial impression of FLAX Engine 1.0 was highly positive, as it comes packed with a solid set of modern and useful features, and also provides the full C++ source code. The showcase video had fired me up quite a bit, which led me to have rather high expectations. As you might know, ACE MADDOX is currently developing RC Maniax: Rock ’n Rumble (RCMRR) using Unreal Engine 4. RCMRR is a classic retro racer using modern technology and a few surprise twists in terms of network play and is primarily aimed for release on Xbox Series S|X and Playstation 5 in 2022.

“After over 8 years of constant development, including one year of closed alpha tests followed by another year of closed beta tests, Flax got mature and stable.” — Wojciech Figat, FLAX Developer

[op_gallery row=”2″]

So, I went ahead and downloaded and installed Flax Engine 1.0 and set out to quickly port over a small portion of RCMRR — assets ported included a pair of vehicles plus a prototype race track, collision boxes, and camera setups.

Flax Engine Evaluation: RCMRR Prototype © ACE MADDOX AB

Flax Engine Evaluation: RCMRR Prototype © ACE MADDOX AB

SPENDING TWO DAYS WITH FLAX ENGINE 1.0:

• Flax Engine 1.0’s map and level editor and its interface and “environment” are shaky, to say the least. For example, a simple asset import caused the editor to crash on me several times.
• After a couple of showstopping imports, I instead started to get familiar with Flax Engine 1.0’s editor flow. So I went ahead and tried to implement a couple of basic C++ routines, which in itself went incredibly smoothly. However, compiling/baking my sample project and its code yielded an undocumented error message no matter how many times I retried. Since Flax Engine 1.0 touts itself as a quick and easy game development alternative, I decided to not waste any more time on the C++ compiler errors — it was time to move on with the real test-project: getting our UE4-powered mini-racer RCMRR up to speed in Flax Engine 1.0 — which was ultimately successful, despite numerous bugs.

FLAX 1.0 FEATURES — TESTED OK

• Importing assets and materials.
• Editing materials.
• Creating physics scene.
• Creation of a static race track.
• Exploring and configuration of game camera and scene clipping.

FLAX 1.0 CONCERNS

• Physics: The complexity of the physics used for vehicle handling is too simplified.
• Using an animation graph to ensure wheel movement seemed challenging; probably works — but requires time to get into.
• Unable to refactor or adjust the C++ scripts that come with the engine (because of an editor issue.)
• Lack of replication will prevent strong online/network play, which is a big “no-no” these days.

CONCLUSION

In comparison to Unreal Engine 4 and Unity, Flax Engine 1.0 is closer to Unity since it is quick to get into, and getting a grasp of the overall editor interface felt intuitive. Also, as with Unity,  the level of abstraction is quite high, saving developers time as investing time in low-level work like filenames and folder structure is not necessary, and in general, there is no need to deep-dive into the C++ engine code in the same way UE4 requires for some game genres (whereas a fully-fledged deep-level UE4 game would require five top-engineers if deciding to work on the underlying engine, Flax Engine 1.0 feels manageable for a small team of say two software/rendering engineers.)

In its current state, I am reasonably convinced Flax Engine 1.0 is solid enough to create commercial-grade single-player games with great graphics for Windows and macOS. However, I doubt the engine is ready for consoles, and the lack of replication makes network games impossible — although I am sure we will see it coming in Flax Engine 2.0 or at some other point in the not too distant future.

Oh, almost forget — make sure to check out the Flax Engine’s developer blog for tech news and great tutorial videos.