SubVIs and Hidden Controls

[uc_greg]

I have a large main VI with several levels of sub VIs therein. All of these subVIs are hard wired and at no time does the user need any of their front panels. Can I make invisible (i.e., hide) these controls and still get the sub VIs to work? That is easy enough to test, but is it worth the effort? Will hiding controls on the subVI front panels (which never appear) thru a property node (or a "Hide Control" right click when editing the diagram) reduce the computational overhead associated with front panel items of these subVIs? Since the front panels of the subVIs never actually appear, do they have any computational overhead?
Thanks in advance.

[Greg_McKaskle2]

> I have a large main VI with several levels of sub VIs therein. All of
> these subVIs are hard wired and at no time does the user need any of
> their front panels. Can I make invisible (i.e., hide) these controls
> and still get the sub VIs to work? That is easy enough to test, but
> is it worth the effort? Will hiding controls on the subVI front
> panels (which never appear) thru a property node (or a "Hide Control"
> right click when editing the diagram) reduce the computational
> overhead associated with front panel items of these subVIs? Since the
> front panels of the subVIs never actually appear, do they have any
> computational overhead?
> Thanks in advance.
>

Answer is -- hiding the controls will not affect things if the panel is
already closed.
LV is pretty good about avoiding work that doesn't show
anyway.

Greg McKaskle

[mikeporter]

Come to think of it, wouldn't attempting to reduce work by hiding controls and indicators via property nodes make things worse? As I recall a couple seminars at NIWeeks the presence of property nodes forces LV to load the front panel into memory and update it. Yes?

Mike...

[Greg_McKaskle2]

> Come to think of it, wouldn't attempting to reduce work by hiding
> controls and indicators via property nodes make things worse? As I
> recall a couple seminars at NIWeeks the presence of property nodes
> forces LV to load the front panel into memory and update it. Yes?
>


Yes. Every diagram that has property nodes or reference nodes to
controls requires that the panel be kept in memory so that the side
effects can take place -- in case the panel is later opened or the
properties are read back from. Having the panel in memory will not
cause the objects to be drawn as long as the window is closed, but it
does take up memory and it causes LV to make copies of the control data
for when the panel opens.

Greg McKaskle

[uc_greg]

Following from that, would it then be less computationally expensive to use the "Enabled"
property in a property node (set to disabled) to disable a control instead of making it invisible with a property node? Just a regular control (with a Key assignment), not a subVI with its own diagram. Would it make a difference?