In the series "Strange New Features in LV 2011", here is an interesting one. I may make a fool of myself as usual, but let's begin with nice pictures...
Here is what I was flabbergasted to see (this, BTW, is the VI I used here to illustrate another LV 2011 feature):
So what is the problem? Hidden controls? Well, yeah, they were some TERMINALS hidden a minute ago until I realized that I had pasted one of the screenshots I used for my other post refered to above onto my diagram. But I just deleted it, and as you can see, the diagram is nice and clean. As for the hidden front panel controls, there are none:
The fact of the matter though, is that the controls we are talking about are references to typedef controls that I am using in another VI. But why should that matter?
All right, why bother, let's just solve the error cluster connection problem mentioned as number 3 in the error list:
Now that is interesting. The two other hidden what-not errors are GONE!
If I Ctrl-Z, they are back!!!
Anybody cares to comment?
When NI did the compiler optimizations they have been talking about, it changed when errors/mystery conditions are detected or reported, and it did so in ways which are inconsistent with previous behavior.
Unfortunately that is an unscientific guess because it has no predictive value.
Darn! I forgot about this universal cure to all LabVIEW ailments!
It doesn't fix the problem, though...
Has a CAR been filed for that? I can post a diagram (although it refers to typedef so I may have to put everything in a Zip file).
Can you upload your VI to this thread so we can keep track of and reproduce this issue? It may or may not be related to your other thread but until we know for sure we will be dealing with each issue separately. I will be working with Luke to see if we can combine these or if one is dependent on the other.
Can you give us some specifics on your LabVIEW 2011 software such as 32 bit or 64 bit? Have you installed the f2 patch? You windows themes suggest Windows XP, is this correct? And again 32 bit or 64 bit?
Did you create this subVI from the edit menu after selecting those controls?
These specifics will help me file a corrective action request and getting R&D to easily recreate the problem and find the cause.
The attached Zip file contains a folder with the incriminated VI (Copy of Manage GUI Button Controls.vi) and the typedef it depends on.
Open the VI, notice that it is broken.
Open the Error List (Ctrl-L) and notice the non-sensical error messages.
Connect the error cluster wire in the case structure, notice how the errors disappear.
I am using LV 2011 (11.0) 32 bits on Windows XP Pro 32 bits.
The "Check for updates" does not return anything, so I did not patch it. Should I? From what I read, there is nothing in it I should be concerned about.
While there is a problem in the error messages, the error message itself is indicating an issue - your references are not tied to the typedefs you provided. They are empty clusters. If I update the references to have the typedef, everything is "normal."
It seems the error message you are eeing is valid, although not stated correctly anymore, and is incorrect when you wire up the error cluster.
I don't understand what you are telling me.
The way I created the reference control was to use a reference to a control in the calling VI (the control was linked to the typedef).
Then created a control from this reference. Then copied and pasted it on my subVI front panel.
You can see that the constant I pass as Value in one of the case is itself linked to the typedef. The reference IS NOT a typedef (and should not). How did you "update the references to have the typedef"?