drjdpowell

Also cross-posted here,  where AQ responds.

I got a PM from the user, Venkatesh.C.S,  and we found that the issue was only on LabVIEW 2014.   No problem on 2013.  I'm not working with 2014 yet so I haven’t followed up. Venkatesh consulted NI tech support, I think, but they hadn’t yet responded.

Yes, that was it.  Thanks.  Though it’s still amazing the amount of blind trial and error it takes to use the 3D picture control.  

Anyone an expert in the (complex, poorly-documented) 3D picture control?

I am using Text to label an axis (see image) inside a semitransparent open cylinder (actually a mesh with texture).  The axis (made of boxes) shows up through the cylinder wall, but the Text does not.  See how 7", 8" and 9" show up, but 10" and higher don't.  

I cannot figure out why, or how to fix this.  Any ideas?

Also, when using the explicit variety, does LabVIEW handle cleaning up the control/indicator reference?  It's my understanding that if you create a reference to pass to an implicitly linked property node you need to close it to free up the reference. 

No.  Those are static references that don’t need to be closed (closing them does nothing, so don’t bother closing them).  You can tell if a reference is static by probing it to see if it changes on each call (if it doesn’t, it’s static).

In 8 years, that's the first time I've heard the cluster/array request. It comes in sometimes for DVRs and for VI Server class refnum types, but infrequently enough that NI has never acted on it.

To be clear: it is doable, there just hasn't been pressure to do it.

I made the request for it to work for Variants, which would include the cluster support CharlesB wants, and also support mixed clusters of objects and primitive types.

I like Mercurial much better than SVN.  Or rather, I like “TortoiseHg Workbenchâ€.  But I've not really used either much more than as a backup system.

I’m afraid the "terminal thralling†(as NI calls it) only works on LVOOP objects directly.  Not even arrays of objects, which is what I would particularly like.

BTW, “Preserve Run-Time Type" is usually not necessary, as the compiler can trace the object identity through the subVI.  Only if the output type could change, do you consider forcing it not to with â€œPreserve Run-Time Typeâ€.

Here’s a quick example of the cluster method.   I did disabling of the string, hiding unused clusters, and also added some colour based on string length.  LabVIEW 2013.

Cluster View into Array.zip

placement of a fixed maximal number of replicas of the element cluster on FP, and visibility of those needed (will become inconvenient for large numbers); 

This will work with very large numbers of elements, if you keep the full set of data as an array and only write the visible portion of the array into clusters (where there need only be 50 clusters to fill the screen).  Use a separate vertical slider whose Value Changed event triggers a rewrite of a subarray of data into the clusters, complete with disabling selected controls.  One can also change colours or other attributes this way.  

But libraries don’t need to just be a good capability to have for the very advanced programmer who has learned to use them carefully “when neededâ€.  Libraries could be broadly useful at a range of ability levels and for a wide variety of things.

I'm with paul on this one. I think I said this earlier in the thread but my group ran into some of the same issues on a project and ended up in a pretty terrible state, with things like the excel toolkit being pulled onto the cRIO. Careful management of dependencies is pretty easy to ignore in LabVIEW for a while but it bites you in the end.

But these aren’t real dependancies.  "A calls B which calls C†is a dependency of A on C.   “A calls B which is in the same library as X, which calls Y, which is in a library that also has Z, which calls C†is not.  I have no interest in managing false dependancies introduced by lvlibs.  

Suggestion: can the event structure just compare New Value to Old Value and ignore if they are the same?  That will filter all duplicates.

BTW, if I were doing this from scratch I would have the 3rd party app set a single (hidden) “Message†control, that would be a cluster of desired control name and control value.  The “brat†would pool for new messages and then do Value(signaling) on the named control (which it would lookup in a list created at initialization of all the available controls).

I don’t have 2014 and can’t open your VI, but my first thoughts are to type cast the U64 array to an I16 array then do a FOR loop over the number of lines to do the chopping.   Why is this data not in I16 in the first place, BTW?

Note that a named queue can have multiple independent references to the same queue, while the non-reentrant subVI contains only a single reference.  

Local means the same application instance, so separate projects in the same LabVIEW should be remote.   Though there is probably some sharing going, on given what hooovahh posted.    

Hi monzue,

“Messenger Library†is currently working it’s way through the Tools Network system.  That, and a bunch of other tasks, are distracting me from your pm, but I’ll get back to you.  

I find myself in this situation a lot, where I have an Actor that inserts itself into a subpanel. I want to look at what's going on in the Actor Core block diagram.

 

Typically I'd stick a breakpoint there so that the particular Actor clone's BD will pop up. But in that case, the front panel also opens, which very often leads to buggy behavior or errors when the Actor later tries to stick itself into a subpanel.

 

Anyone have some workaround suggestions for this? I run into the problem a lot and always slug through somehow but I figure someone's got a better procedure.

I (not AF, but similar) set the subpanel option “Allow User to Open Block Diagramâ€.  Then I open the block diagram from the subpanel, once the actor’s FP is loaded into it.   Even when the actor leaves the subpanel the diagram stays open (this is a bit strange, since I have open block diagrams with no front panel open).   Much more convenient that having to use breakpoints.

Hi, Why my software is slow? it need about 1s to insert 10 record. Thanks Boone

Most likely because you are not wrapping your INSERTs in a “BEGIN†and “COMMITâ€, so they are all separate transaction.   Each transaction requires verified writing to disk (to ensure no corruption if the computer losses power), and as hard disks only spin at 50-100 Hz you can only get 10-30 transactions/second.   Execute BEGIN before the INSERTs and COMMIT after, then you have only one transaction.  See the Examples (Search SQLite in the example finder) to see how this is done.

In the latest version of the package (coming out soon) I’ve added “BEGIN†and “COMMIT†subVIs to make this easier.

— James

Note for other readers: this package is now available on the LabVEIW Tools Network as "SQLite Libraryâ€.

I had assumed that if I had a reentrant VI and had created multiple instances of it, that I could insert each instance into a separate subpanel. 

You can.  You must have done something wrong and be attempting to put the same instance in the second subpanel.  

Your Xcontrols instances go idle when you unlock the XControl, thus destroying all their owned references just like any other VI going idle.  So I wouldn’t worry about a leak.

Off topic: Came across the concept of a zero-element queue the other day.   Enqueue blocks until another process calls Dequeue, and then they both execute at the same time, passing the element across.

I’m thinking “Messenger Libraryâ€.

I sort of like Isimud

Inventive, but I don’t want inventive, I want obvious.

I'm using non OO based actors and I don't know the issue you are talking about.  Are you you talking about when one actor will keep the application from closing?  A global shutdown command that is a high priority user event takes care of that for me.  One major reason I dislike the official Actor Framework is the reentrant-ness which forces you to other debugging techniques (setting break points in the VI saving, waiting for a clone to be made and stop, then being unable to remove that breakpoint while running)

I mean the “root loop†needed to do an asynchronous call that is blocked by the User having a menu open.  

I personally have had little problem debugging shared reentrant actors.  I just send them a message to show their front panel.  I can even then add probes and restart the program to observe the startup behavior, as the same clone is always reused.  No breakpoints needed.   But there is a learning curve to this, which I’d rather eliminate by making it possible to have non reentrant actors.

Still I trademark it "Glued Framework", (Global User Event Driven)

Simplify it to GLUE, as in “I built my application using GLUE."