Make labview everywhere a reality

[falkpl]

I have used labview for a long time and avid user.  One issue I have been hitting lately is the "LabVIEW everywhere" slogan never really panned out, it has become LabVIEW everywhere NI allows it to be.  I am getting jealous of the Arduino and Rasberry Pi and hundreds of PICS and ARMs not avaliable to me (Yes I have the pro liscence but not embedded).  I wish Labview pro opened up the toolchain and started porting to many other platforms by default.  I am seeing jobs that labview is loosing ot to where it should be much more competetive like the embedded market. 

Essentially I am looking to see the Labview development environment easily work with toolchains for the most popular processors and also open up a simple standard to add targets to projects. 

Wouldnt it be nice to program a $25 ardunio dirrectly from labview (NO THIS IS NOT WHAT THE TOOLKIT IS DOING).  Add a Ardunio target file (maps the io memory to variables and throw down a loop, boolean shift register, a wait and a digital line variable, download to the micro and the blink led example is done.  Really open up the doors for LabVIEW everywhere.

[Peter_B]

This idea already has support.  See a meaty discussion here.  Also consider voting for this idea here.

I dont have the time right now but if somebody could take an educated guess at the following question then we might be in a better position to justify ideas such as this.

What is the percentage of NI's profits due to sales of LabVIEW software. (as distinct from any of their H/W) ?

My gut feeling is that if NI could increase LabVIEW sales by making it work for many more (non-NI) targets then NI would be reaping plenty of profits way above the amount of profit lost by a drop in their H/W sales (since people would buying more non-NI H/W instead).

Note that while R&D costs are fixed, the mark up for a LV license is much more than the mark up for a piece of H/W since it costs nothing to write to a DVD and package it up, but manufacturing a piece of H/W costs parts and labour.  H/W profit margins are much smaller in comparison.

[falkpl]

Although this does allow for third partys to compete with NI in the hardware arena I dont think this would gut their sales.  NI is the porshce of the daq world.  Where I would use such solutions would be on the lower end solutions not on the PXI highend daq vision or other solutions that require power.  This would however cost some losses in the usb 6008-6009 or some SBRIO solutions tha didnt really require the SBRIO in the first place.  NI could offer similar products at a slightly higher cos and I would still recomend and buy the ni solution.

Example:  Small arm micro development from a thirs party ~$50-100  programmed from Labview dirrectly (no toolkit needed)  plays OK with labview some configuration work needed and some custom library develompment (IE LCD screen library I would develop myself because it was not previously done).

NI offers for $300 same specs and micro at the heart of it,  100% compatable and extensive library -  NI wins except for very high volume solutions.

What this does for me is allow labView to Solve problems that I couldnot use labview/daq in the first place due to cost.  I think the LV for embedded was hurt because of the cost structure and therefor a small user base.

If this solution was avaliable it would lower the embedded learning curve and spawn a new generation of LabVIEW users.  What University would not offer this for embedded/physical computing 101 classes. 

This is just my opinion.

PS thanks for the links to similar discussions.

[G-Money]

I know it doesn't answer your idea exactly but I still wanted to point out that there is a free toolkit to target some Arduino targets from LabVIEW (https://decibel.ni.com/content/groups/labview-interface-for-arduino).

[falkpl]

Thanks I have this toolkit but the toolkit puts firmware onto the Arduino, then labview simply sends commands to the processor.  The labview code never compiles and runs on the processor, this is what I would like to do, Use LV to program the processor.

[vitoi]

falkpl,

Your idea is spot on. This is my greatest wish for LabVIEW, which will make it more popular and improve LabVIEW’s power and my career. When I started with LabVIEW about 8 years ago I was very happy with it. There was much innovation such as Events, cRIO and LabVIEW Embedded for Microcontrollers just recently introduced. The future looked bright.

However, NI has decided to keep its captive market at its mercy and force them onto expensive NI hardware. I see no innovation in the past 8 years. Sure the Boolean constant looks better now, but nothing structural that gives me as a LabVIEW programmer a competitive edge.

It’s not only me that sees NI not innovating – as others have said, look at the NI share price, which has plummeted relative to the general market.

I bought a LabVIEW Embedded for ARM development board with great excitement. At first it looked great – easy to program and fun. However, I soon realised that the available Tier 1 boards are well past their prime and that the OS/LabVIEW overhead is crippling. These are not technical restriction, but NI marketing restrictions.

The greatest shame is that there is no competitor to LabVIEW to drive innovation. And when some like Dasylabs show up, NI buys them out.

However, on further thought, I realised that there are competitors. Powerful and inexpensive competitors. For microcontrollers there are free and powerful development environments in C from the microcontroller manufacturers. Even Microsoft provides a free development environment call .NET Micro Framework. Many powerful boards are being introduced such as the $25 Raspberry Pi. The software development environment has moved enormously in the past five years. Development environments that cost many thousands of dollars are now available for several hundred dollars. Powerful microcontroller development boards are available for well under $100. But NI is caught in a time warp and standing still.

For my embedded project, NI tried to railroad me into sbRIO (not OEM friendly) or the C Code Generator (ridiculously expensive). Both are bad choices. So, what did I do? I moved to C – and the world opened up to me. So, my summary at the moment is “LabVIEW is great on the desktop, but when it comes to the embedded world, NI has nothing to offer”. As you’ve pointed out, this is not a technical issue but rather where “NI allows it to be”.

I hope to one day come back to LabVIEW for the embedded environment, but I’m not making myself dependant on NI and putting all my eggs in the NI basket. My ability to get things done and my career are too important to me.

Regards,

Vito

[falkpl]

Nice Vito,

I would love to use LabVIEW as a primary development platform as a complete embedded and not lose my competetive advantage over C which is that I produce the software for a fraction of the cost and in a fraction of the development time.  I have been very suscessfull using NI and Vision in producing solutions that blow away the the other languages, NI doesnt gain much on the sales (a small depolyment fee- sometimes hardware like smart cameras and CL cards) but LabVIEW gets to slowly establish itself as a power player in the industry. 

I would think NI would benifit by becomming a power player in the embedded development hardware then using this as leverage to offer higher end solutions like cRIO or in a perfect world rtDAQ (no this doesnt exist but I wish it did)- essentially a daq line that allows native RT bootable labview code)- same as SBRIO at a lower cost less features and no FPGA.  I could justify paying 5x for a ni product in the embedded world and dont expect NI to offer a $50 board but 250 for the equivilent power and much higher quantities should yield similar profits.

Inly time will tell if LabVIEW Everywhere will happen,  I am sure all of us hope so.

(I hate typing; and dread using C; as the primary embedded platform;  \ \ this doesnt compile #inclued 'LabVIEW')

[joej]

What if LabVIEW were to branch further out into general purpose programming? I just read an article about how Facebook created code to transform something called PHP (a scripting language) into functionally equivalent C++ code.

Facebook claims "PHP has major benefits in terms of programmer productivity ... Simple to learn, simple to write, simple to read, and simple to debug". Sound familiar? 

Also "We are proud to say that at this point, we are serving over 90% of our Web traffic using HipHop [the name of their transforming engine], (which) allows us to keep the best aspects of PHP while taking advantage of the performance benefits of C++".

I read NI has a C code generator. It's only a matter of time, while LV remains constrained to NI hardware and measurement/automation in general, that some other programming productivity art-form catches up - even if its genesis is from a different angle.

Message to NI - Get On With It. It's been a nice 20+ year ride; from what I heard here, the last 8 have been a coast from LabVIEW's perspective. People WANT "Simple to learn, simple to write, simple to read, and simple to debug" in programming, just like they wanted a "good" OS of their own - and made that happen. If YOU dont do it, it WILL come from somewhere - maybe Facebook shares it, like they do their "open compute" hardware designs.

[ErnieH]

Amen. User since 3.1 and becoming very disillusioned. Where is the LabVIEW everywhere we were promised? We did our part promoting it and singing its praises in spite of the opposition encountered to implement it.  Now it is just stagnating and becoming bloatware. I want to use it everywhere, but I don't see that happening or even being seriously considered. Closed system, limited job opportunities, endless self promotion. Not what we signed up for. When they do come up with something innovative (like SBRIO), they don't want to sell it to you or make you work to try to buy it. Give us the tools and get out of the way.

[Peter_B]

It just occured to me that every other idea on this exchange is superficial in comparison to this one.   Just because you might think this idea is unlikely to be implemented by NI, should NOT stop you voting for it.   The fact that a relatively trivial "Opacity/Semi-Transparency appearance settings for attractive Front Panel objects" idea gets more than twice the votes in less time tells me that you LabVIEW users are content with playing in the sandpit rather than thinking BIG.  

C'mon folks, don't be afraid to let NI hear how their current plans for LabVIEW are simply not good enough for you and NI !!

[vitoi]

Surely the smaller Boolean constant is far more useful to us LabVIEW programmers than better embedded offerings including microcontrollers and more OEM friendly sbRIO 🙂