LV 2016 Different Behavior For Dynamic Dispatch (possibly bug?)

[prettypwnie]

I've run across a difference in how LabVIEW 2016 works with dynamic dispatch VIs compared with 2015. If you put the class wire on a shift register of a loop and then have an event structure inside it and in one of the event cases instead of wiring the shift register through, you wire whatever value was originally put into the shift register through (you might do this if you're making a dialog to configure the class and want to let the user cancel it without affecting any changes) it breaks your VI - it doesn't consider this to be a "good enough" way of wiring your dynamic dispatch input to the output. However, if you have this same code in LabVIEW 2015, it works perfectly fine.

See attached code for an example - if you open it in 2016 the run error is broken, in 2015 it is not. It is the exact same code.

Does anyone know if whether this is a bug or just new, intended behavior?

Originally posted on LAVA here: https://lavag.org/topic/19818-lv-2016-different-behavior-for-dynamic-dispatch-possibly-bug/#comment-...

[Kyle97330]

It seems like it should be a bug.

I don't have 2016 installed yet but can you work around it by putting a "Preserve run-time class" node at the end?

[prettypwnie]

Good idea, I just tried it, and yes putting in preserve run-time class at the end makes it not broken. Thanks for the workaround!

[AristosQueue (NI) · Accepted solution]

I confirmed that is a bug. CAR 606847 filed to track it.

[prettypwnie]

Wow, AQ himself replied to my post!!! I feel so special 🙂

[ChrisStrykesAgain]

CAR 606847 was fixed in LabVIEW 2017. See the 2017 LabVIEW Bug Fix List for details: http://www.ni.com/product-documentation/53584/en/