Terrain threatet like an unwanted child

In 2013, Unity accepted the > 1000 votes for a new terrain editor.
Now we are at 2016, and the terrain core engine is still - according to the Unity road map- "At research".

This is laughable. I'm still not able to use normals and spec on terrain as I can with UE, and I'd have to use serveral asset store add-ons to accomplish what I have in mind.

Why is there not further information about the current "research" status?

 

Because it will likely drop in Unity 6 in some form or other. That's a bare-faced guess. It's still very much in the experimental phase... There is no point on reporting on what can change wildly.

Oh, and UE4's terrain sucks as well. There aren't any super good terrain solutions in general game engines. I've tried the big 3. All of them have drawbacks. Mostly, terrain is one of those big unsolved things in game dev, by unsolved, I mean it's always going to have to be custom if you want the best for your particular game.

When splatting for instance, you quickly blow through all your gpu's bandwidth and this is why UE4 is so slow, and why Unity's implementation is so limited. So to solve this, a lot of thinking out of the box, and perhaps an entirely new approach is required, and many smarter people than us are working on just that as we speak.

I did around a year's research on terrains, which encompassed megatextures, baked streamed chunks, texture splat, vertex splat, hybrids and tried most of the GDC slides approaches. All of them had drawbacks, and compromises be it for authoring, visuals, memory or performance.

So I really don't want Unity to rush this one.

 

But I would like to have some directions. Should I design my terrains with the current, very limited maps and entirely re-do them when Unity updates the terrain core engine? Or will Unity offer a really sophisticated automatic upgrade that will perhaps automatically load spec and height maps if present in the texture's folder while they currently can't be used?
I don't want any specifics, but it would help me if Unity would explain how they plan upgrades once they redesigned their terrain engine.

 

I think the only thing you'll be guaranteed to be able to transfer is a heightmap. The reason this is guaranteed is because you can pull it out of the existing terrain manually if you wish, and pretty much everything in terrain land will accept a heightmap.

As for splats and so on, I doubt this will be immediately compatible. Wouldn't your game be done in 2 years? would you still be doing it? Would you trust the new terrain to be bug free or even good performance? you should be working only with what you have now.

 

I'm still unsure about how to go on because as far as I know, the Blacksmith demo uses mesh terrain.

 

I won't have my game finished in 2 years, rather 3-4 years.
It consists of lots of terrain.
I would simply like some directions or statements from Unity of what they plan.
A post on their blog would be great.

 

If you have that kind of time and specialised needs you are maybe in a good position to do a Terrain system yourself.

If unity had something to present that they can be sure is not falling over next month they would post it.

 

Why can't you use normal maps and specular on the terrain? They're there for me - I'm assuming you mean something else?

I agree terrain needs an overhaul, something voxel based would be great.