Reading this makes me very sad I had ro drop Godot. The main reason for it being that everyone I’ve heard of uses Spine by Esoteric Software Ltd for animations. There’s a dead fork of a fork of a repository adding support, but although some things have been easy to maintain, it’s broken on Godot 3.1.
Esoteric has been promising support for something like two years and recently went on record as not having started the work yet(!)
I’m sure Godot’s cut-outs would suffice if animators knew how to work them.
It would also be nice to be able to add proper 3D elements into a game made in 2D mode, but with the focus on figuring out OpenGL and Vulkan for another half-year, I don’t really see this is a fit for a long time.
If you don’t need those things, I recommend Godot. The community is sweet and the editor is joyful.
Godot recently added the ability to transform 2D sprites using meshes & bones - with a bit of work that should be able to do most of what Spine can do.
As for proper 3D elements, there is the option of using a render target texture to show a 3D scene in the 2D viewport. However, I expect you’d be on your own when it comes to interacting with the 3D objects - you’d have to do the coorditante transformation maths yourself.
Outsourcing animations to someone used to dealing with separate Spine files does unfortunately not match very well with anything in the proper engine. This is probably the reason a third-party product is popular :/
There were other small glitches as well, which may have been fixed now, like parallax layers and camera zooms not working together too well.
Regardless, I wish Godot the best!
Oh yeah, from that perspective the Godot animation story is far from ideal - any animator would have to work in the likely unfamiliar Godot editor, instead of being able to specialise in the one tech that works across engines. And, while it is constantly getting better, Godot does come with a certain amount of ‘open source’ clunk that would be especially painful to people used to dealing with polished proprietary tools.
If you don’t mind me asking, what engine are you switching to?
UE4 would be overkill, and its Spine support seems to lack normal maps, though I find it very interesting.
We conducted a feasibility study of porting our scripts to a Unity asset, by making an interp or compiler of sorts. It looks very promising and should save us money in the long run.
The sad part is that I got very familiar with Godot, nearly fell in love with it, and know very little of Unity. Now I’m making enough in billables from Python work that I can outsource a bunch of the compiler work and hope this doesn’t backfire, and after this gig see where we’re at.
OTOH we’re needing a meta-ish script system anyway, so if the asset fails us, all we need is another “backend” for Unity.
If Godot worked, and if I hadn’t spent time and money making stuff on it only to learn a design concept for a Unity-based architecture, it would be more affordable. Unity on Linux was also pretty much a broken hack when we started and I prefer working through Linux; now it’s way better.