@ChrisStrykesAgain wrote:
I'm seeing this same behavior (two identical sets not being equal), with one weird caveat. A force recompile (or any change that naturally causes a recompile) will suddenly make the equality true. Save and close the VI, then re-load it, and it's once again "broken."
- That sounds suspiciously like LabVIEWs Compiler is trying to save a buffer allocation that is going to allocate at ???time or during load. Funky. Look at the last Compiled with xxx optimizations property please.
Compounding things, this manifests in one particular project. If I put my VI to a different project, it no longer reproduces in that context. Closing that context and opening it in the first, will then revert to the "incorrect" behavior.
- Bookshelf A and Bookshelf A' would both exist until the other context closes and the context switcher is aware of both.? Worse, I have no idea if Bookshelves A and A' are equal since they cannot be compared! Nice and muddy buddy!
I have compiled code separated, but for some reason this exact behavior will reproduce on multiple machines given the same set of VIs which seems very odd.
Again, I'm not convinced that you have incorrect and broken used correctly about the comparisons if, bookshelf theory is a good analogy, we should expect what is seen with that 1 exception about force recompile in one context. That slices nicely against Occam's Razor.
Time for me to learn something... I ll expect the unexpected!
"Should be" isn't "Is" -Jay
