A few more random thoughts in my journey through trying out Fuego.
My first thought on playing with the process designer was that I really liked it because it shows the process activities in swimlanes. The activities are colour-coded (red for human interaction, blue for system-executed), and the shape of the activity icon tells you something about what it is, or you can define your own activity icons on a per-activity basis, which is kind of cool.
But wait… swimlanes are supposed to run parallel to the flow, not perpendicular to it: if the flow runs horizontally (as it does in Fuego and most other process modelling/design tools), the lanes should be horizontal (which they’re not in Fuego). This has the effect of projecting (in a mathematical sense) the two-dimensional orthogonal nature of a swimlane diagram (activities in time sequence versus role) into one dimension: time, with roles dependent on the time axis. This makes for some weird anomalies such as roles repeating along the time axis, and showing automated steps as part of a human role assignment.
Whether you use UML activity diagrams or BPMN or venerable old LOVEM diagrams, swimlanes are supposed to be, by their very nature, orthogonal to (and therefore independent of) the time-based flow of the process, so that you can see all the activities for a single role in a single band regardless of their position in the flow.