Recently, I came across several LabView attempts to penetrate the DSP sector (commercial and educational). LabView is the de facto in data acquisition, however at first glance you feel it is kind of alien to signal processing, until you meet an enthusiast (or a marketing engineer) who fiercely tries to convince you otherwise.

In my experience, I bought CompactRio digital signal controller, which has a real-time OS on a dedicated processor in addition to Xilinx FPGA. You can use both or either of them for your application. So I ventured into the realm of LabView programming for my signal processing application. I admit I was lecture at first, since I consider myself a MATLAB veteran, however untapped potentials can be discovered. Especially, when using data-flow programming with FPGAs. 

So what do you think fellow signal processing engineers?

It has been awhile since I have had a serious look at LabView but my take is that it is useful for some tasks but when you start doing serious development it gets unwieldy.

The main problem is it is difficult to use many of the tools that exist for software development.  How do you collaborate, how does version control work (diffs, history, etc).

Also, if there isn't a version of the language that students and working engineers can work with on their own it is bound to fail.  Too many engineers work on their own projects and and/or develop their skills outside of work.  They will be most efficient with the tools they know they can use anywhere.

Thank you for good points raised. I strongly agree with your last point. A student version should be available to play around with. I believe, though, that there is a free student version of LabView for six months only. On an another note, NI took over Ettus and BeeCube software defined radios (SDRs), which are popular SDR vendor in addition to Nutaq. This, in some sense, forces engineers to use LabView for signal processing. So do you think such move would be effective?