It is not quite clear what you are doing nor what you want to do. Having a VI (or several) that illustrate some of this would really be helpful (make sure, if you do attach code, that you attach actual VIs or Controls, not static pictures of block diagrams or front panels).
Here are some questions:
- Why have you converted all of your Typedef Clusters to Variants? [I'm not even sure I know what this means -- I'm guessing that everywhere in your code that you Bundle By Name a Cluster that has an associated TypeDef, you immediately run the Cluster wire through To Variant. The rationale for doing this escapes me -- I cannot envision a situation where this makes sense.
- To convert (one of) your Variants back to its Cluster, why do you have to "place all of the Typedef Clusters on the Block Diagram"? Why not just put the Typedef of the Cluster you want to convert? [You need, of course, to know, in advance, what is contained in the Variant Wire, which you've sort of "hidden" by making it a Variant].
- It is certainly true that if you change your program by adding a new TypeDef, you have to change your code whenever you need to deal with this TypeDef. This is true whether or not you use Variants.
I had to deal with something similar recently. I had an analysis routine for data generated by Version 2.0 of a program, which had a specific Data Format. However, Version 2.1, 2.2, and 2.3 had slightly different formats (each had its own TypeDef for the data). I handled this by using a Case Statement with the Version Number (as an Enumerated Type) as the selector -- the individual Cases were "Process 2.0 Data", "Process 2.1 Data", etc., that is, each sub-VI "knew" how to process its own data.
Could something similar work for you?
Bob Schor
