Yesterday I came across Creature; a software animation tool that beside some new features, has a different approach to things than Spine. The core idea is that animation should be done as much as possible using something called "Motors".

Essentially a Motor is either a function or system applied to a set of bones with variables that can be adjusted. On one hand this could be a simple function or preset application (flap wings, bounce, etc). On the other it could be a procedural system that generates an animation, such as Automated Walk Cycles, or even more important, bones that follow a physics model. (An example being a tail that swishes according to the model format)

This final example for us is pretty much the missing feature in Spine. We have a lot of characters which extra secondary motions (hair, spikes, flames, tails, etc) and the time spent on make these look and feel good is just too much.
The problem with it is that we are not trying to make the bones or animation do anything special, but instead are just trying to mimic a physics model which could easily be simulated.

Some of this functionality could be implemented rather quickly in Spine. (Essentially "macro"s with variables that could be applied and tweaked) The other more complex systems would obviously require more work.

Some other functionality of Creature which we have found to be lacking in Spine:

• Sprite Swapping; easily creating animations. In our current pipeline we have setup things such that upon loading a skeleton is scanned for bones with certain names, and their frames replaced with animations. (Hand setting the frames in Spine for trivial animations was just too tedious)
• Path Motion; a no-brainer: the infamous "Growing Vine" example would be no work at all. https://www.youtube.com/watch?v=RpVMiUC8Ewk

The main problem is that our "fix" to most of these things requires us to recode the tools in LibGDX / our game, and then the artist can only see the results after exporting and importing them into the game.

Creature: http://creature.kestrelmoon.com/more_info_2.html

The issue at hand isn't whether or not a talented artist has or hasn't yet spent hundreds of hours creating a gorgeous animation in the tool.

The problem is that we have about 50-ish characters to animate in various animations, and the secondary motion and motion path functionality already saved hours when I asked the animator to compare the tools with a sample animation.

Another unexpected benefit was that the animation only contained keyframes for the "core" animation, which made maintaining it a lot easier.

Why would you need to see work from an artist to acknowledge shortcomings in Spine or strengths in other software?

Most of the features proposed drastically reduce the time required to make certain types of animation, they don't alter the quality of the resulting animation.

Or do you just not agree that features such as second-hand motion are viable and worthwhile? (I am actually curious)

Like I said before, the second-hand motion/physics and path-following are the features I am particularly interested in.

I haven't heard your opinion on those yet.

The others are mostly useful to create a start setup to work with.

The whole point is that the physics is included into Spine so you can see the effects in real-time in-editor and have runtimes which use these same rules so it looks uniform on all different platforms.

The question isn't on whether we should all switch to Creature. (We shouldn't)
The question is what we can add to Spine to make it better.

As for unfinished; the motors (in my interpretation and workflow) are not used to produce final animations.
They are primarily intended as a way to save time, the exact reason why Spine exists.

The same for path following; just imagine how fast the infamous plant grow animation would have been done if path-following was included into Spine.

Essentially I should've renamed the thread: Creature; or What can we add to Spine?