{"id":1,"date":"2026-07-14T04:57:40","date_gmt":"2026-07-14T04:57:40","guid":{"rendered":"http:\/\/blog.skelform.org\/?p=1"},"modified":"2026-07-15T09:54:22","modified_gmt":"2026-07-15T09:54:22","slug":"hello-world","status":"publish","type":"post","link":"https:\/\/blog.skelform.org\/?p=1","title":{"rendered":"Re-animating the 2D Skeletal Animation Ecosystem"},"content":{"rendered":"\n

Having played Flash games in the 2010s, I was accustomed to the style and charm of characters being rig-animated rather than the usual frame-by-frame. Being your usual programmer who could just barely draw up something fine, I wanted to make rigs and take advantage of procedural animation (characters looking at any angle, inverse kinematics, etc). This is apparently a big ask, as every opportunity comes with some major caveat.<\/p>\n\n\n\n

Let’s start with a bit of context; I primarily want to make my games using libraries<\/a>. These tend to be very simple frameworks, so much so that they don’t have any visual editor<\/a> to lay things out. This is fine as most visual work like level design<\/a>, textures<\/a>, etc. are done in separate programs. Skeletal animation is no different, but the ecosystem isn’t as bright.<\/p>\n\n\n\n

Illusion of Choice<\/strong><\/h2>\n\n\n\n

The 1st option would be DragonBones<\/a> – it’s the most competent while being fully free. It’s got inverse kinematics and mesh deformation; you’d say it’s a steal if they weren’t outright giving it away! Unfortunately no one’s there to give it out anymore. Scroll down to the bottom of the barely-working site and you’ll see this:<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

Even with other signs of life taken into consideration (eg; the Twitter account made in 2020), it’s clear that DragonBones isn’t being maintained anymore. You know what is<\/em> maintained, however? The subscription-based, web-only, AI-centric animator known as LoongBones<\/a>! How’s that for a juxtaposition?<\/p>\n\n\n\n

I would say ‘the 2nd option would be…’ but the rest of the animators have the same issue with minor deviations, so I’m lumping all of them in.<\/p>\n\n\n\n

Besides DragonBones, the other ‘obvious’ choice is Spine<\/a>. It’s the ultimate animator, and I know it is because it costs almost three times my monthly rent!<\/p>\n\n\n\n

\"\"
That rent comparison was for $379, by the way<\/figcaption><\/figure>\n\n\n\n

The Essential version is still a whopping $80. No, it’s not worth it just because the full version is so prohibitively overpriced. The Essential version doesn’t even include mesh deformation, which you could at least get in game engines like Godot or Unity.<\/p>\n\n\n\n

Speaking of those engines… why not pick them? As mentioned previously, I much prefer the workflow of simpler frameworks. It’s not speculation either – I used to take game dev on Unity semi-seriously (only got 2 games to show for it, but unfinished projects and all that). While I haven’t used Godot as much, I’ve heard that making and animating rigs from within it isn’t all that great, and it could still benefit from a dedicated animator.<\/p>\n\n\n\n

Other options like Creature and Spriter have similar price issues as Spine. It’s on a smaller scale, but I’m still not paying that much for something I don’t even know will be good to use when the honeymoon phase of a game’s development is over and it’s time to start actually making things work.<\/p>\n\n\n\n

With all this in mind, I was willing to ignore DragonBones’ issues and try it out anyway – it’s fine. The UI is confusing and I didn’t get far with my own characters, but that’s because I spent most of my time on the runtime<\/a>. Being acquainted with the runtime development process gave me a very bright idea:<\/p>\n\n\n\n

What if I just… made my own thing?<\/p>\n\n\n\n

The 15th Standard<\/strong><\/h2>\n\n\n\n

Now hold on – before you go ‘oh boy, here’s another for the pile<\/a>,’ hear me out.<\/p>\n\n\n\n

DragonBones has more issues than simple abandonment. The editor itself isn’t open-source, and there’s no runtime documentation or specification (reading source code of existing runtimes doesn’t count). While people like me could make said docs, there’s nothing stopping the editor’s inevitable bit rot<\/a>.<\/p>\n\n\n\n

If it wasn’t clear by now, I’m not doing all of this just for myself – I want to see animations like these beyond just from my own games, and no amount of effort on my end will offset the constant decline of an aging editor. With all this being said, I would like to introduce my solution:<\/p>\n\n\n\n

SkelForm<\/strong> – A free and open-source 2D skeletal animator for games.<\/p>\n\n\n\n

Skeletal Animation for Everyone<\/strong><\/h2>\n\n\n\n

That’s the tagline I’m building SkelForm on – it’s a bit cringy, but it helps to keep in mind that SkelForm should ideally be suited for both newcomers and veterans of animation software.<\/p>\n\n\n\n

Let’s get the easy points out of the way; it’s free, not just as in free beer<\/a> but open-source. It’s on Windows, Mac, Linux, and… web! This is big not just because installation isn’t required, but it also provides a fallback if downloaded versions don’t support any particular platform (eg; FreeBSD, mobile, etc).<\/p>\n\n\n\n

A friendly UI<\/strong><\/h3>\n\n\n\n

I mentioned that DragonBones’ UI is confusing, and that complaint did not go in vain; SkelForm’s UI was based on Godot\/Unity’s editors, which make a clear separation between object hierarchy and properties:<\/p>\n\n\n\n

\n
\"\"<\/figure>\n\n\n\n
\"\"<\/figure>\n<\/figure>\n\n\n\n

left – SkelForm; right – Godotred – hierarchy panel
green – properties panel<\/p>\n\n\n\n

The UI across the program trades some ‘efficiency’ for familiarity; many elements like the edit modes (Move, Rotate, Scale) and Armature\/Animation modes could have been icons and\/or been in a single bar, but laying them out as text and spreading them across the UI ensures that no single part is overwhelming.<\/p>\n\n\n\n

\"\"
An example of worst-case scenario for UI entropy<\/figcaption><\/figure>\n\n\n\n

Of course this will only get harder as more features are added, but I think it’s fine in spite of all the features that SkelForm currently has to offer. After all, it shouldn’t be a lot.<\/p>\n\n\n\n

A Lot to Offer<\/strong><\/h3>\n\n\n\n

A skeletal animator isn’t complete until it has inverse kinematics and mesh deformation. Even if most of your rigs use it, the potential of procedural animation and faster workflows can’t be overstated.<\/p>\n\n\n\n

These features will have their own posts when the time comes, so have this rapid-fire list for now:<\/p>\n\n\n\n