instances revealed when doing a compare

[SeanJ]

Hi All,

I've been comparing two VI hierarchies and have come across an unusual set of returns;

The VI, "Measure OSA.vi", returns a number of instances that are different. 

e.g.

Measure OSA.vi:Instance:2

In each case it appears to be the express VI, 'Display Message to User', that is causing the difference/ instance.

Does anyone know why the express VI's are causing these instances.  They don't appear to be re-entrant, I'm a bit baffled.

Thanks,

Sean

[RER]

Hi Sean,

Thanks for posted about this issue. Firstly I wanted to reassure you that, dispite what the compare list implies, there are not multiple instances of your top-level VI. However, I agree that the naming convention being used is missleading.

To replicate your issue I made a simple vi which called 3 express VIs in sequence. I saved this VI (compare1.vi), and made a copy of it (compare2.vi). Effectively, I now had 2 identical VIs with slight different names. While both VIs were open in memory, I went to tools >> compare. As you can see below the compare Window shows a main instance of the top-level VIs PLUS x3 instances of the top-level VI (one per Express VI).

I them compared the 2 top-level VIs (ie. Compare1.vi and Compare2.vi). As you can see below, the comparision tool (correctly) found 0 differences.

I then compared a top-level VI (ie. Compare1.vi) with an instance of that top-level VI (ie. Compare1.vi:Instance:0). As you will see, the tool found many differences. This is because we are NOT actually comparing 2 instances of Compare1.vi... but we are actually comparing Compare1.vi with one of the ExpressVIs.

I wanted to take this oppertunity to reassure you that, dispite the implication, the additional of each ExpressVI does not spawn multiple instances of the top-level VI, and that everything is behaving as intended. However, I do agree that the current naming convention for each Express VI is rather confusing. I will raise this as a corrective action request with the LabVIEW R&D team to see if we can make this a little more obvious in a future release.

Thanks for bringing this to naming convention oddity to our attention. But, to assure you one last time, from a programmatic/functional perspective, everything is working as expected.

Hope this has been helpful,

Kind regards,