> I assume you do anyways, at least implicitly, to save the state of your GUI’s output. You should bring that into focus. Indeed, I do have a YAML / JSON repre

> I assume you do anyways, at least implicitly, to save the state of your GUI’s output. You should bring that into focus. Indeed, I do have a YAML / JSON representation under the hood, which is validated by JSON Schema. In the app, you can click on "file > yaml editor" to access it. I thought about having a mermaid-like syntax on top of my YAML representation but before adding this, I need to have a proper "auto-layout" feature. Currently, the editor is responsible to detect collisions between systems and to "space them out" until no collisions are found. I need to bring that responsibility back to the core library instead of the editor, as right now, the core library only accept systems without collisions to resolve. TL;DR: it's feasible and I'm open to the idea! > What is your opinion on sequence diagrams in general? I like them in small bites, i.e. with few participants and presenting one simple flow. They do become daunting (for me) to read and maintain with more participants and complex flows. I like to sprinkle them in my documentations. In mermaid, I find myself bouncing from "flow diagrams" to "sequence diagrams" all the time. > over animations you have to wait for when you need to review a detail you missed I'm currently working on a "viewer mode" / "presentation mode" as I don't want this continuous playback loop to become the default way to present a flow in a system. This continuous playback is useful as an overview of a flow but not to present it to an audience or to embed in a technical documentation. I want to be able to present a flow, step by step, as if it was a slideshow.