Thank you for the kind words. I'm very happy you are enjoying Spine! 🤙
As simon.h mentioned the bot is just to get people answers immediately and handle the easiest questions. It doesn't replace our support. Spinebot did touch on an important factor though: the glTF runtime doesn't have all the features Spine needs.
In the old days people had to use tools like Flash, 3ds Max, Softimage, and similar to make 2D animations, export the data somehow, then write the runtime support themselves. The runtimes are a significant piece but even if you had that done for free, using tools that aren't purpose-built for 2D requires avoiding many features that aren't supported at runtime. That is a huge downside, much larger than it might seem at first. A lot of time can be wasted by accidental usage of unsupported features.
Using Spine for a glTF (or non-Spine runtime) would give a similar experience: you'd need to avoid some Spine features, else your glTF exports would fail or break. Having all runtimes support all Spine features has been a goal of ours from the start, and we've mostly achieved it except for a small number of game toolkit limitations. Given that, I'm not super excited about people needing to walk a tightrope to stay within glTF supported Spine features, though I do understand it would have some nice benefits.
From the beginning of Spine we thought we might need to allow choosing a game toolkit, then disable features that toolkit can't support. Thankfully we haven't needed to do that so far, but something like that might make sense for glTF.
Alternatively, we are planning a "warnings" view, where we collect potential issues (so you don't have to hunt down tree ⚠️ icons). It could make sense for that to flag use of features not supported by a particular game toolkit.
And the end goal is not a game engine, but Blender itself.
In that case a Blender Spine runtime would enable ALL Spine features. That is much more interesting for Blender users, though it loses being able to leverage glTF in the many other places it's supported. We have explored a Blender runtime at some point in the past, though I don't remember how far we got.
While it's certainly useful to be able to render a rig via glTF in the many apps and game toolkits that support it, that support is lacking a fair amount of functionality needed to do 2D well (and comfortably). Spine is purpose-built and has specific features for 2D animation, so will always have a significant advantage over a generic solution.
The biggest factor is what Spine features are impossible in glTF? For example, losing constraints would be quite a big disadvantage and make it a less interesting export target. There could be other deal breaking incompatibilities, like if Spine's key curves can't be mapped to glTF.
I haven't looked at glTF very closely yet, but apparently it has an extension system. It would be interesting to see 1) what features do we get with glTF alone, 2) how much farther can we get writing an extension, and 3) how portable is an extension to other apps that support glTF.
