Hi,
This is in fact not too inconsistent. It is consistent with e.g. how activex
works. Each request for a reference returns a copy of the object (and thus a
new reference).
It isn't ideal in LabVIEW. I'm still looking for a clean solution too.
Comparing type def, name, label, panel bound, etc does work, but isn't very
nice.
Reagrds,
Wiebe.
"arlans" wrote in message
news:50650000000800000049DE0000-1079395200000@exchange.ni.com...
> The control refnum reported by a front panel created reference node is
> not the same as the refnum listed in the FrontPanel.Controls[] array.
> The enclosed example demonstrates this problem. The Cluster node's
> refnum is 0xA0300002 as reported by the reference node. Meanwhile the
> Controls[] never lists this r
efnum but the Cluster's refnum is
> reported as 0xA0800008. This is the same before and after the 'To
> more specific class'.
> This should not be the case. The refnum to an object should be
> consistent. If it isn't then kludgy methods like looking at the
> object's label's text have to utilized to attempt the identification
> of controls. However, this approach is not a clean solution. If you
> have several identical clusters, which is my situation, the names of
> the labels are the same. While I could name each cluster instance
> uniquely, this is still a clumsy and slow when utilizing consistent
> refnums should be available as a more elegant solution.
