Is there a character limit on enum items?

[JeanPierre]

That is the byte length of the type descriptor array that is limited to 65535 thus an array length of 32768 I16. The type is signed I think for historical purposes but in practice it is internally used as an unsigned value. In LabVIEW 8 the problem does not occur because the length header of type descriptors is now 32 bits. That should leave us another 15 years before hitting the limit again...

Edit: and do not forget that the Front Panel itself is kind of a cluster and have its type descriptor including all the controls of the panel.

Message Edité par Jean-Pierre Drolet le 11-18-2005 05:53 PM

[Ben]

Thanks JPD!

The world again makes sense.

Ben

[CoastalMaineBird]

OK, NI Tech support has confirmed that:

You are correct, that is a known issue
with LabVIEW 7.0 and 7.1. As you pointed out
this problem was corrected for LabVIEW 8.0.

They didn't exactly confirm the 32768 number (I've asked for clarification, but heard nothing back yet), but basically said I was out of luck. So I have sidestepped the problem by replacing the ENUM with a string indicator. For display purposes, it's a bit slower, but it avoids all this. I tested adding some more stuff to the ENUM and there is no problem. Which confirms the idea that it's not a problem with the ENUM itself but the things that use it (and contain 60 copies of it). Anyway thanks for your idea.

[CoastalMaineBird]

That is the byte length of the type descriptor array that is limited to 65535 thus an array length of 32768 I16.

Well, I know that my byte length was 30500+ (having flattened the TD to string and counted string length). So what you're saying means I was barking up the wrong tree again. Nevertheless, the fix I applied (see above) removes the problem.

do not forget that the Front Panel itself is kind of a cluster and have its type descriptor including all the controls of the panel.

Well, I have a front panel which had a tab control with ten (10) pages of this cluster, one in each page. That's 1800 controls on one panel, and 600 of those were of this enum, having ~500 bytes each. 600 * 500 = 300000 bytes just for the TDs, assuming no compression, or duplicate removal. Of course if there was duplicate removal, the single cluster wouldn't have a problem. So I don't see how I got away with that, if what you say is true.