Choice between CVI and MS for C++

[EEGMAN]

I want to know the benefits of using Measurement Studio for C++ over LabWindows CVI. Recently a decision was made in our development team to switch to CVI after having invested quite a bit of time into MS for C++ code. Can anyone tell me why that makes any sense?

[duncan carter]

I've been using CVI to make data acquisition front ends with graphs, user,
controls, etc. after using MSVC 6, Borland Builder, and (yuck) VB for
several years. My opinion is that CVI is a much simpler and more straight
forward environment to work in and has most, if not all, of the various bits
that I need instead of needing to integrate bits from various vendors. I've
been porting some projects previously done in MSVC to CVI. I find that this
is much easier that the original project and that there is much less coding.
I'm very happy that I purchased CVI. I also purchased LABView as a part of
the Professional Developers Suite but I'm not so sure about LABView. There
hasn't been much of a learning curve for CVI as I've been programming in
C/C++ since 1988 a
nd CVI seems mostly the same as what I've been doing all
along.

Duncan

"EEGMAN" wrote in message
news:50650000000800000084840000-1042324653000@exchange.ni.com...
> I want to know the benefits of using Measurement Studio for C++ over
> LabWindows CVI. Recently a decision was made in our development team
> to switch to CVI after having invested quite a bit of time into MS for
> C++ code. Can anyone tell me why that makes any sense?

[mfukuchi]

Hello,
If you can tell me what your application does, and what other features you are planning on adding, then we can see what advantages or disadvantages using CVI has over M Studio for VC++.

[EEGMAN]

Our system is full phsyiological monitoring and feedback system that is programmed in Visual C++ 6. We initially used the NI Math classes and found out there were problems with changing the filter coefficients readily in a realtime system. The lead software engineer found out he could build his fiters with an interface that would allow him to do what he wanted to do very easily in CVI and then put a C++ wrapper around them.

My concern with that method is that we lose some flexibility for development in the future and that it impacts the computational speed of the system in a real-time application. Of course we have not benchmarked it yet, so we don't really know if the C++ classes do any better than a wrapped CVI procedural program. We plan to bench
mark the difference and I can let you know how it goes. My suspicion is that it is just much easier to do things in CVI for the engineer. But, I ask at what cost down the road for the overall robustness, maintenance and extensibility of the system for the future.