Future of our CVI Projects

[Wolfgang]

Hi All,

With the various announcements that no new releases, only critical patches are to be expected for LabWindows/CVI in the future we, the users of CVI, are facing the challenge of how to continue our projects based on CVI.

Right now, I see two steps:

1. Within the next few years we probably will have to migrate the IDE, compiler, linker and debugger to another system (e.g., Visual Studio code, clang, ...) while keeping CVIRTE.dll, allowing to maintain and further develop our software, written in C, in a Windows environment.
2. In a second step we probably will also have to replace the CVI libraries (e.g., for UI…) as they may not function properly anymore in future hardware or Windows environments.

These are tasks I would prefer to not face alone and I guess it would be more efficient if all affected users of CVI would share their knowledge of this transition.

How can we do so, set up a Wiki? Where?

Thoughts are welcome… thanks!

[GabelDaniel]

Hi

My thoughs are as follows:

Fully migrate to visula stodio.I do NOT want to rely on any SW from NI if i have to leave CVI.

I can not trust that in future releases of win the NI modules will continue to work and if i put the hours (MAAAAANY) to migrate i will do it all the way.

In the most part of the projects the source is C , so visual studio supports C.

The parts i have to rewrite are "only" UIR , SQL toolkit ,AactiveX for excell,word etc..

After 27 years with CVI it is sad that NI abandoned (with out an official word) all of its trusty usres

Best of luck for all of us

Gabel Daniel

[gdargaud]

I see several issues:

- the FPs. It's possible to extract the c/lib/dll/obj inside and the documentation. It would be possible to write a script to convert the documentation tags (or the .doc files) to doxygen in order to transfer documentation to different compilers (does MSVC supports doxygen ? I haven't used it in 25 years).

- Code generation from FPs: you'd loose it. And that's not just Ctrl-I to insert a function name with its parameters, but also Ctrl-D to declare variables, [Enter] to see the valid UIR constants, etc...

- Various wizard code generators, such as niDAQmx inProject or inMax.

- The UIR. If loading will stay possible as long as the runtime engine works, how do you create a .uir without CVI ? Creating a .tui manually is _possible_ but quite a drag and requires extensive knowledge, and above all you loose all design helpers. That's a major drag.

- The specialized libraries, for instance the Advanced Analysis Library. They'll keep linking of course; and some have open source replacements (such as the GSL), but they are not as easy to use.

- Many other things I'll miss (including the money from teaching and consulting about CVI !!!)

On a personal note I've been programming 80% on Linux for the past decade. It's great for many things, but I haven't found a suitable replacement to CVI for user interfaces. On a recent project we did the  heavy lifting in C and the user interface with a Java web framework. It's a nightmare of complexity and I could have done the same project in 1/tenth the time in CVI.

[ebalci]

New graduates come with bragging about how great Visual Studio is, as if they wrote it, and yet it takes a few minutes to amaze them with CVI.

And besides, not being familiar with C, they lack the basic understanding of how programs "operate".

.NET Framework encapsulates everything and you no longer feel like you are programming a "machine".

The simplicity and intuitiveness of even the Variable window with built-in graphing and exporting capabilities is wonderful.

Never tried to do it in Linux though.

(not related to the topic, but Mr. Dargaud, what is your setup on Linux? Which distribution do you use? I have a recent project. )

In terms of the lifetime of CVI, I think I'll continue using it with whatever version of Windows it is last supported.

I'll probably be too old to continue coding myself when it is completely useless.

Do you think that NI will stop even the activation support?

[GabelDaniel]

Hi S. Eren BALCI

I completely agree with you CVI is wonderful IDE but...

I am approaching now 60 years young and i am the software manager in the company i am  working in, so i must think of the next phase

for the next generation .

As of today NI did not give us any official announcement about the future of CVI but in between the lines one can understand that CVI is going to DIE . As my position in the company i can NOT ignore it.

All of the company management and infrastructures SW (and of course all of the projects) are written in CVI.

Because of that i can not rely any more on CVI.

Gabel Daniel

[RobertoBozzolo]

It's true that NI is leaving CVI in a limbo of no-news about its future, but it's pretty much evident that present release will be the last one. Maybe I'm a bad person or I'm simply thinking negative but... I wonder if NI is just delaying an official announcement to gain another little year of licence renewals!

I'm thinking this way as we have just received the proposal for renewing our SSP program at a discounted price... which is just 150% of the last year!

If this is NI policy with 30-years suscribers, I suppose the future of our CVI projects will be very short!

[scottrod]

This news is very, very, very bad. I had not heard any of the rumblings about this until I found this thread.

National Instruments: We, your loyal customers deserve some specific information about this issue. At my company, we (and I) have been using NI gear and programming in CVI since before it was in C and running using BASIC in 1992. That's 30 years of data acquisition and analysis code, most of which after refactoring, is still in use daily for FAA certification work. We invested over $100,000 in new NI modules this year, all of which are using CVI codes in multiple facilities.

On a personal note, I was set to retire in less than a year. This may actually cause me to have to rearrange my life in a scramble to find solutions  for my company to a suddenly deprecated programming environment. This is not a trivial situation. We need some planning time. NI corporate needs to make some decisions and pass them along to the user base, asap. We are owed that at the very least.

Scott Rodriguez

[ElectroLund]

oh wow, I'm stunned!  And yet... we could probably all see the writing on the wall.  The market share difference between LabView and LabWindows has been 1000/1 forever now?  Still, wow.

Personally, I've never used FPs.  I prefer to write all my own device drivers if there isn't one developed already by vendor.

So, what's next?  For the past 3 years now, I've ramped up my dev in VS Code.  It's a beautiful, light-weight IDE.  In fact, the reason I'm on the forum this instant was to see if there was any more traction with the idea of using it as a CVI IDE replacement.

National Instruments: open source CVI RTE now!

[DieselPower]

I too went through this discussion with NI and learned that I am one of a very small list of people that are writing real time programs using CVI.  It is a great environment for those that like to code with words, not pictures.  NI has pushed LabVIEW so hard that CVI is the red-headed stepchild.  Of course LabVIEW has many more users.  I'm not sure I've ever seen an advertisement on their website for CVI.

I have been working on this solution for several years and I am just now rolling it out only to find that nothing is supported anymore.  I thought the path forward was Linux, but then I found that there is no Linux RT support for CVI.  I reverted back to Pharlap, CVI 2019, and LW CVI RT Module 2019.  We had many cDAQ-9179s which aren't supported, so we had to swap back to 9178s, and we are stuck on DAQmx 18.6.  

I can live with the current software, but I face the risk of my company saying that, since CVI is no longer supported by NI, I am no longer allowed to have it installed on the network.  I'm not sure how many hackers out there are trying to break into CVI/Pharlap systems anyway, but them's the breaks.

Maybe NI could afford the gesture of documenting how to move forward using VS (which I have) while maintaining access to the CVI libraries.

What a disaster.

[ElectroLund]

---

@DieselPower wrote:

Maybe NI could afford the gesture of documenting how to move forward using VS (which I have) while maintaining access to the CVI libraries.

---

Do tell!  How have you managed to move forward with VS?