hooovahh

There are several libraries online for simulating keyboard presses and mouse interactions.

https://forums.ni.com/t5/Example-Code/Programmatically-Simulate-Key-Strokes-in-LabVIEW/ta-p/3521756?profile.language=en

https://forums.ni.com/t5/Example-Code/Using-LabVIEW-to-Simulate-Keyboard-Events/ta-p/3996210?profile.language=en

https://forums.ni.com/t5/Example-Code/Simulate-Mouse-Click-on-a-LabVIEW-control/ta-p/3509840?profile.language=en

9 minutes ago, crossrulz said:

But  the real story:  Harrison Ford was sick and so the director just off-the-cuff said to just shoot the guy.  A big sword fight was originally planned, but instead we got one of the most iconic movie moments.

I think this was mentioned in the video I linked to.  I was just thinking about how much work the stunt coordinator, and stunt men had training and planning for a thing that just never happened.  It probably happens a lot in the film industry but as someone in the digital world I tend to keep even the work that I never actually use.

On 1/23/2021 at 11:14 AM, Rolf Kalbermatter said:

That was very lame of Indy! 😆

It was but does give us some character development.  Showing the exhaustion on his face, and surrounded by dozens of people that might attack.  He kinda is like, nah I'm done I got other things to do. 

Also likely considered murder since it would be difficult to claim self defense.

Love it.  BTW here is a crosspost on the dark side.

14 minutes ago, Gribo said:

Take into account that MXI has its own set of limitations. .... It is a single source (NI only, AFAIK).

I've heard conflicting information on this one.  Here is a quote from another champion:

Quote

The MXI standard isn't something NI invented per-se, it sits on top of high-speed server connection standards that have been around for years.  I've had NI provide MXI cables that were made by a different vendor, and have wanted cable lengths that NI don't sell and have found them online.  My guess is that if you can find out what that standard is (start by googling the part number of the connector on one end of a cable you already have), you'll open up a range of cables and bulkheads.

I know I was able to find stuff a few years ago, although not for all the MXI standards.  Your card looks like it's "MXI-Express x1 Copper" which is a renamed "TDP PCI Express x1" - more info that might help here:

https://www.andovercg.com/datasheets/molex-74546-0813.pdf

Also mentioned in this private conversation was a thunderbolt to PXI chassis from a 3rd party that I don't have experience with

https://www.adlinktech.com/Products/PXI_PXIe_platform/PXIChassis/PXES-2314T?lang=en

Yeah I'd suggest looking into MXI.  In particular I've used the PXI chassis that has MXI integrated into it.  I'm pretty sure it was the PXIe-1073.  This is a 5 slot PXI chassis, that can come with a cable and PCI card included. (Be sure and get the right part number if this is what you want).  With this you can plug the PCI card into a compatible desktop PC, and then it will see the PXI chassis and all the devices in it in MAX.  Great for when you need to expand a PC to have a bunch more channels, or cards that are only in the PXI form factor.  Oh and if you have the chassis already, NI sells a PXI card that goes into the first slot, and then connects to the host PC through the PCI card I mentioned earlier.

Another benefit of this is obviously the PC can be upgraded, and the chassis can be reused.  As apposed to having the controller using the first slot.  These more embedded systems have their place, but obviously have a cost premium associated with it.  If space isn't as big of an issue, you just get any normal desktop and use the chassis.  Then upgrade RAM, CPU, motherboard, etc. without having to buy a whole new controller.

I've seen some discussions about trying to use the thunderbolt standard between PXI chassis, but last I knew there were limitations in the bios or kernel that caused issues.  That was on the host side so running it from a laptop might be possible but I have no experience with this and haven't seen anyone do it.  You might want to contact NI.

14 minutes ago, X___ said:

Yeah, well, when was the last time NI asked users what they were expecting in a IDE UI?

Well at least every NI Week since a couple years before NXG was released up until the last one. Hackathons were a pretty open environment to give feedback.  You'd have the attention of some pretty important people in the design of the UI and UX decision making process.  And the IDE actually had a button in it where you could just send NI a message about something you liked or didn't like in it. Not sure if that stayed for the newest versions.  I can't say what NI did or didn't do with this information but they seemed to have plenty of places for users to give feedback on.

One of the key differences in the IDE with NXG is the single window view.  It took some digging but I found an old LabVIEW 8 beta that had a similar design.

I'm guessing they had usability issues or something as it obviously didn't end up like this.  Maybe similar to the ribbon interface and how that was only really used on the Web UI builder and an alpha.

Last I knew the forums had a minimum search length of 4 characters.  At some point there was a warning if you searched for something with less 4 characters.  I'm guessing the period is somehow ignored or reserved for some kind of special formatting.

Don't worry Brian will pick up the tab.

Yeah deploying to a target that I don't have the option to run source on the same OS sounds like a nightmare.  Where I'm from that "throwing it over the fence" which usually results in dozens of builds trying to figure troubleshoot the issue.  I'll agree this is a bad situation.  But I think it would be worst if NI made the same decision in 2 years.  LabVIEW is dead, long live LabVIEW. (hopefully)

In 2014 I had the privilege of being in on one of the early discussions around what would be NXG, before I was a Champion, and I think I was just a CLAD.  They were interviewing each person one at a time with a set of questions and the last one was "What is your biggest concern with this decision by NI?" and my answer was some thing like "What are NI's plans if this fails?"  Saying that this is my career they are potentially messing with, along with the negative effects internally at NI.  They sorta laughed it off and said how unlikely it was but that they would do their best to go back to supporting the more classic LabVIEW.  It probably is the right decision for them to make, but also a very difficult one.  If it were me I'm not sure I would be able to see past the sunk cost fallacy.

Cloneable VI: Should work just fine.  NI did a pretty decent job with VIMs, and has seemed to fix a few edge cases and bugs with them in recent versions.  Should work just fine.

Examples: Yeah I was lazy.  A few of the older examples were intended to show special features of XNodes.  But VIMs don't have all those special features.  So instead of re-writing examples for VIMs and putting it in the package, I just didn't include any.  I should have but wanted to get it out there and I was losing motivation.

Glad it helped.  Is there a repository somewhere of these types of gifs?  I'd love to have a library of controls or something.  Like I mentioned a QControl can probably handle this well.

I've recently released the Variant Repository code as a package on VIPM.IO as VIMs.  Some of the magic is lost going from a series of XNodes to VIMs, and even more wind is taken out of my sail by the fact that Sets and Maps can do what VRs did in a more native way.  However I have some code I'd like to share that relies on Variant Repositories, so I cleaned it up for a release.  And replacing those functions with Sets and Maps would mean my code could only be used by 2019 and newer.  In the future when 2019 doesn't seem so new I'll probably remove the VR dependency.  Or maybe I'm being lazy I can't tell.

10 hours ago, Michael Aivaliotis said:

Who gave JKI permission to do this?

To be clear I don't know that this is what JKI did, and no one told me they did this.  I just know I didn't upload my stuff to VIPM.IO, and saw it was there which lead me to believe that this is what they did.  The rehosting of BSD stuff is probably fine.  The using LAVAG.org in ways that might go against terms of use might not be.

So I'll admit I was dragging my feet a bit when it came to VIPM.io.  I sorta saw this as another standard leading to this XKCD.  But the process is pretty slick, and pretty easy.  And the fact that it seems to have already crawled LAVAs Code Repository means a few of my tools are already there and looking nice. 

One issue I saw is with the screen shot resolution limitation.  800 x 500 is a decent size, but I found some of my palettes didn't fit and needed to be shrunk to fit.  Is this size limitation just to ensure it fits on the page well?  Here is my array package and I needed to scale that image before uploading.  I'd prefer just upload any size (within reason) and have the page scale it to fit the container, then clicking it can show the full size.  Pretty minor and it works fine as it is.

I did use something a while ago that sounds similar to what you are describing.  First I'd like to know if the OS it self is supposed to be on the USB drive too or not.  If not I'd look into Thinstall or Thinapp which is virtualization software.  Years ago I tried it and it worked for running LabVIEW IDE portable.  It scans your system, you install software, it scans again, and then sets up a sandbox with the difference in files that can run on a similar OS without needing to install anything.  It mostly worked but had issues.  If the OS is supposed to be there too, then maybe run some kind of Windows PE environment?  Or use a VM maybe inside Windows PE or even inside of Linux? Lots of options, all are probably terrible.

Pretty hard to tell. I mean the game came out in 2001, so development was probably 1999 or 2000.  The UI could be Window 98, or Windows 2000, and it could be LabVIEW 5 or 6.  But honestly it could be anything.

Being able to combine your data processing, and GUI processing into a single loop is sometimes useful to reduce complexity.  However, plenty of LabVIEW developers intentionally keep those two things separate.  I've of the opinion that as long as you are very confident that the work being done is short, and won't keep the UI from locking up, then this approach is fine.  Of course rarely can I predict how large an application might grow and how much of an impact this will have on the UX.

Awesome and congrats.  I look forward to any kind of revelation or perspective changes you have as a result.

I do something close to Neil.  My messaging structure uses the user events, and so in my event structure I have a long timeout of say 5000ms.  In that timeout case I check to see if the user event is valid (not a number/reference).  If it isn't valid then go clean up, I probably forgot to send a quit message or something shutdown improperly.  I made an idea on the idea exchange to generate an event, when a reference gets destroyed over here which would make this a bit more graceful.

Oh man that looks like a painful thing to debug and isolate.  And an even bigger pain to convince NI that there is an issue when someone in the support call will just have a goal of getting your code working.  I haven't seen anything in 2020 like this yet.  I have seen code that application builder says is broken, I go and run the VI it isn't broken, and then the build can continue. It is related in that it seems to be a compiler issue.

At what point in the video does a tab get removed?  I just watched it and no tab is removed.  I just downloaded the code and you are right that the Remove Tab doesn't work.  Not sure why, but I probably just forgot to implement that.