Well Node-red does not really have an execution thread concept, meaning that you can't really define when nodes are executed. Luna Park does have execution thread wires, which are separated from variable wires.

So I'd say that Node-red is more of a workflow editor, not a visual scripting editor. You can create almost anything you would in code with Luna Park nodes. This is not the case with Node-red which is a bit higher level, and you do need to code the complex things you want to run.

Other than that, there are a lot of convenient features in Luna Park like:

- can be integrated withing any webapp (feels like Node-red needs a dedicated server to run)

- generate standalone JS code from the logic graph (to be run headless, with native performance)

- merge nodes into function that can be used as custom nodes

- type and value inspection at runtime

- a lot more nodes (generated from standard JS functions)

- and a few others features (integrated documentation, node customization, inspector, console, etc...)

Basically, Luna Park and Node-red might seem similar when looking at their UI, but they have a pretty different purpose. Node-red would be useful to automate workflows between systems. Luna Park would be useful to build those systems.