LabVIEW
Capture Events When Writing Block Diagrams
Well there are multiple solutions and they might depend on what kind of thing you want to monitor or how to react. Yes an XNode is an option where the user interacts with a node and you can write code as it is interacted with. If a user wires something up do something, is the kind of events you'll have. Or if the user right clicks, or left clicks, or resizes etc. XNodes are experimental, undocumented, and you lots of fun when they work. I did a few presentations on them here, and here, and here is a tool I made for making XNodes.
Without knowing more I might suggest you go another route which is registering for events and monitoring user interactions. This also involves undocumented, and experimental features of LabVIEW along with private methods. But there are some good examples of this in action which work well. Checkout the the improved JKI State Machine Editor which monitors user interactions and can react to objects selected on the block diagram. I believe the source is available and open, if it isn't this is based on an older tool the CaseSelect. A similar tool I use for helping with panes and splitters found here. This works on front panel objects but most of the principals are the same with finding VI activation and then monitoring for item selection.
If you start this code running in parallel with development, you can poll to see how many objects exist on the block diagram of whatever the top-most VI is and how many objects are selected. By checking each successive polling for changes and then logging those exact changes, you can get a pretty good idea of what is being done on the block diagram.
Adding code to parse the arrays of references to actually check what has changed (and to ensure that when the developer switches which diagram they are editing it doesn't log 1000 changes all at once) is left as an exercise for the reader.
