Unexpected coercion in LabVIEW 8.6

[altenbach]

(This is just a quick report so I can reference it in the monthly bug thread.)

Download this VI (from this thread) LabVIEW 8.6.

OK, let/s do a few very simple modifications:

1. move the initializer terminal out by one loop.
2. Wire "numeric" to the initializer terminal and remove the bad wires.
3. delete the "1" diagram constant.
4. Right-click the "+" primitive, select "replace...", pick the "+1" primitive.
5. Bam!!!! We get two coercion dots (one on the initializer terminal and on on the array indicator) everything in-between turns EXT!.
6. Why!??

Message Edited by altenbach on 09-10-2008 03:15 PM

[Kameralina]

Hello altenbach,

Yes I was able to reproduce this with LabVIEW 8.6 from my end too.

Thanks for noticing this and providing the feedback; I will make sure that R&D is aware.

Have a great day! 

[Roy_F]

---

FieldKam wrote:

I will make sure that R&D is aware.

---

This was reported to R&D (CAR # 125834) for further investigation.

[RavensFan]

I came across this issue today with LV 2009.  I was going to post a new message, but this thread turned up in the search.  The bug still seems to exist.

1.  Drop I32 constant on diagram.

2.  Drop Feedback Node on diagram.

3.  Wire constant to initializer terminal.

4.  Loop feedback onto itself.  Wire turns blue (I32).  That's good.

5.  Insert +1 primitive on that wire.  It turns orange and it is extended datatype.  Why?

6.  Put indicator on that wire.  The terminal confirms it is EXT.

There is nothing on the diagram that should be forcing the feedback node and wires to orange.  No way to change it.  All you can do is Undo backwards.  Drop the +1 primitive by itself (rather than inserting into the wire) and wiring up both sides to the FB node individually. 

Message Edited by Ravens Fan on 09-12-2009 06:14 PM

[johnsold]

I seem to recall that way back in LV 5 or 6, maybe 7, maybe earlier, that the increment function had a tendency to convert integers to floats.  Is there something deep down in that beast's ancestral DNA which has come back to life?

Lynn 

[Roy_F]

The above CAR is still active, meaning it is targeted to be fixed in a future release. The problem is one of type propagation specific to the feedback node. The workaround is to delete and rewire either side of the feedback node and it will type prop correctly.

[Peñata]

Hi Ravens Fan,

Thank you for reporting this bug. I was able to reproduce this in LabVIEW 2009 and have filed a Corrective Action Request. The CAR number is 186929.

Thank you for choosing National Instruments.

Aaron Pena

National Instruments

Applications Engineer

http://www.ni.com/support

[DFGray]

Another easy workaround is to place a conversion node in the loop just before the feedback node.

[Roy_F]

For those tracking CAR numbers, please track this issue via CAR 125834, not 186929 (186929 is closed as duplicate of 125834).