01-12-2012 05:40 PM
Created a new cluster containing a pass/fail boolean, an integer for "counts", and a short calibration string. Saved this into a typedef'd cluster, dropped it on my block diagram, and it shows up as the mustard-colored error cluster!
Any way to make this not look like an error cluster on the block diagram?
Adding another object in the typedef and making it invisible is kind of an ugly hack, but it does turn the object string-pink.
Solved! Go to Solution.
01-12-2012 05:47 PM
Reorder the elements.
Right click on the cluster in the type def, select "Reorder Controls in Cluster", and click on the elements in a any order other than, bool, int, string.
01-12-2012 06:32 PM
Making your counts U32 instead of I32 might work.
01-12-2012 06:41 PM - edited 01-12-2012 06:42 PM
If a cluster has three elements, and those element data types are (in order) boolean, I32, and string, then it will get drawn like an error cluster. Change anything about the cluster to *not* fit those criteria and it will go back to being drawn like a regular cluster.
Funny thing about it being specific to data type is that if your cluster has a boolean, an I32, and a tree control, it will get drawn like an error cluster. 😉
01-12-2012 07:14 PM
@Darren wrote:
If a cluster has three elements, and those element data types are (in order) boolean, I32, and string, then it will get drawn like an error cluster. Change anything about the cluster to *not* fit those criteria and it will go back to being drawn like a regular cluster.
Funny thing about it being specific to data type is that if your cluster has a boolean, an I32, and a tree control, it will get drawn like an error cluster. 😉
... so the tree is just a string as far as LV is concerned.
Ben
01-12-2012 07:41 PM
01-13-2012 11:17 AM
@matt W wrote:
Reorder the elements.
Right click on the cluster in the type def, select "Reorder Controls in Cluster", and click on the elements in a any order other than, bool, int, string.
That's the reason I asked here, knew there'd be some simple solution to the problem.
@Darin.K wrote:
Making your counts U32 instead of I32 might work.
That worked too. From the string-number conversion in the root VI that actually gets that data, I'd have to cast that to an I32 instead of leaving it as the output's U32. Not a big deal to do, but since this VI is used elsewhere in other projects I'll just leave it the way it is and change the control order.
01-13-2012 11:20 AM
The error cluster is sooo 1990s. It needs to be a class.