Probe on error wire impacts timing considerably

[BowenM]

So I noticed some odd behavior and am curious as to what is going on "under the hood".  Obviously it doesn't impact my real code or anything, but I found it interesting.

I was playing around with a simple cDAQ output task and allowing it to time my loop (ignore the bad code, it was me just testing stuff).  I wanted to see how stable it was, so I dropped a pt-by-pt histogram in to look at the delta T... it acted as expected

After running for a bit, the histogram looked about as would be expected

However, I later dropped a probe on the error wire and it greatly impacted the timing in the loop.  Note that this doesn't happen if I probe any other wire - but reliably happens on error probe.

[JÞB]

I'm assuming you are using recent LabVIEW version where Error Probes are conditional by default.

Does this occur if you:

1. Delete the error probe
2. Select the Wiring Tool
3. Hover over the error wire
4. Right-Click>>Custom Probe>>Controls>>(Select an error in or error out)

???

Alternately, can you duplicate by 

1. Hover over a numeric wire
2. Right-Click>>Conditional Probe>>Conditional <Datatype> Probe

???

[BowenM]

Ah!  You are correct sir.  The first (Custom probe -> controls) does not duplicate.  The second (conditional) does which means it is the overhead of the conditional probe.

I also will admit... I never noticed that error probes were conditional by default.  That will be useful. Thanks 🙂

[JÞB]

---

@BowenM wrote:

Ah!  You are correct sir.  The first (Custom probe -> controls) does not duplicate.  The second (conditional) does which means it is the overhead of the conditional probe.

I also will admit... I never noticed that error probes were conditional by default.  That will be useful. Thanks 🙂

---

Thanks for doing the legwork! please post a simplified version that demonstrates the behavior and steps to reproduce.

Why? Although I expected the conditional probes to add overhead, I would have thought the overhead would still be consistent ( your example suggests that it is not)  So, you may have uncovered a bug.

Not really a hugely important bug but perhaps one that should be known to increase the understanding of the IDE.

"NOT Important you say", Why?  Well, it only effects the IDE and not any deployed code and then only when debugging.

Therefore <Sticking tongue firmly in cheek>

 It seems silly to ask for R&D time for "Debugging a debugging bug when the bug is automatically disabled by disabling debugging especially when disabling debugging is automatic by default"

[BowenM]

---

@JÞB wrote:

---

 "Debugging a debugging bug when the bug is automatically disabled by disabling debugging especially when disabling debugging is automatic by default"

---

That reminds me of this.

Anyway, you can replicate with just a simple wait and an error probe