Flowcode, I believe is like Labview where the "icons" execute when ALL the data is available. I had to learn LabView when it was being developed Version 2.2.1 and ported to the PC. I was also managing the code development. I had to develop "Flicker fixers" so the screen wouln't flicker.
It was very easy to write processes that executed in parallel which also means potential "race conditions".
At the time, the "Error Cluster" was not well thought out, but it's the bases of every vi (Virtual Instrument) that you develop. You set up each vi, with an error in and error out and some (management stuff) such that it does not execute if there is an error. The "last" vi inthe sequence handles the error handling and "printing" of the error.
This is a key concept. Without it, you can't get very far.
Another key concept is variable types, type casting and structures. LabView also has special numbers called Inf and nan.
An Inf would be created by dividing by zero. Nan is Not a Number.
In this particular project, I chose to:
1) Develop without the instrumentation connected - i.e. simulaton
2) I worked on conceptual issues and program design
3) I did the instrumentation interfacing.
It worked out really well.
One concept that my manager had was to print results after each test. That turned out to be unworkable, Tests were batch printed at the end of the day. One parameter that was entered was area and it was usually the nominal area. You could load the file back and actually change the area and the data would scale accordingly. There was a structure field, a pre-comment and a post comment.