Using Error In and Error Out Clusters within VIs

[Hooovahh]

---

@mikeporter wrote:

Which brings us back to the real mystery of NIWeek 2014: Why were there no changes in LV that were of any real substance?

---

Maybe this should be a new topic.

As for the Wait, yeah there is no legitimate reason NI can't do it.  That being said no company should make their own, OpenG has a nice one with the ability to not wait on error, and abort on an occurance.  Admittedly I've never used this feature but I like that it's there.

EDIT:  And wasn't there a wait express VI from some DAQ or Real-Time toolkit that has error in and out?

[mikeporter]

---

---

@mikeporter wrote:
I hadn't seen the discussion, but I can sum it up in two letters:

B and S

---

I can't say I agree with (m)any of the things that aristos posted on the idea exchange, but I've never once gotten the impression that he was BSing people. 

---

To be clear, my comment was not personal -- to tell the truth I didn't notice who posted the response. Although I have known Aristos for many years, his name on the post was immaterial to the basic point that the arguments in the post were and are logically indefensible.

Mike...

[dems]

All he says in that post is that they evaluated it and agree its easy to implement, but believed there would be unintended side effects because adding error wires to the connector pane implies that you should wire them up. I've seen people do some really silly things with error wires, so the argument doesn't sound unreasonable to me.

---

It's a primitive.  There is no wires nor sequence structure in a primitive.

---

Yes, yes, but hopefully you get my point anyway. I don't disagree with the idea of having a second VI on the palette to add these wires/features, but I think the primitives should only have things in them which can't be done in regular LabVIEW code.

[dacad]

The Wait(ms) can be done in regular LV code.  I use to always use the old Wait Function ( Wait+ (ms).vi  attached) and then they stopped including (after 8.6 maybe?) and added the new clumsy, big iconed wait vi that only had seconds input.

I finally made my own that had both ms and seconds (Pause.vi). It can be with or without error wires. Just cause a vi has error wires doesn't mean you have to use them. For functions that clearly would no generate errors and for which you do not need flow control, it's perfectly ok to leave the error wires disconnected.  I didn't add output but it could be added easily enough.  I now use this exclusively.

[Dennis_Knutson]

The old wait+ was part of traditional daq (and still is). The new Delay function that takes seconds is not clumsy, has error in/out, and is only big if you choose it that way.

[Hooovahh]

Time Delay is the function I was thinking of.  It is an express VI (BOOO) but you can right click and view it as an icon and it basically is just a wait with error in and out, and the wait is in seconds not ms.  But it is native.

[Dennis_Knutson]

Yeah, I've been using Delay for years since it is naive and available at all customer sites. I don't see the point of modifying the wait(ms) just to add error connections. Having one that skips the wait if an error exists is something different.

[Hooovahh]

And abort on occurance...not that anyone uses that feature of the OpenG one.

[mikeporter]

The express VI also operates in a fundamentally different way. It does not insert any wait time. Every time it gets called it tells you if the defined time period has elapsed.

Mike...

[crossrulz]

---

@mikeporter wrote:
The express VI also operates in a fundamentally different way. It does not insert any wait time. Every time it gets called it tells you if the defined time period has elapsed.

Mike...

---

You are referring to the Elapsed Time VI.  There is also a Time Delay VI that is literally a Wait (ms) with a multiply by 1000 (since the input is in seconds).