Kernel 2.6.31 under Mandriva 2009.0

[mhoogend]

Did you make it clear to the sales engineer that the reason they didn't get the sale was due to lack of Linux support?  That is most important to get the message across to them.  Your particular sales engineer may be less versed in Linux, but by making it clear that Linux is what is preventing them to make the sale you are giving them motivation to learn.

[anshuljain]

Yes I did tell them that it was the lack of Linux support that forced me to take the decision. To be fair to the support guys, there was only one engineer (among the 4-5 that I talked to), who was interested in getting me the right answers...which, unfortunately, weren't encouraging. Heck, I'd do a better job on Linux support than the NI guys who had little idea of what I was talking about! No offense meant to anyone...just making my point clear.

[mhoogend]

No offense taken, including by them I'm sure.  I think you also point out something from our end, our engineers get proportional training on topics based on how frequent it will come up.  If support or sales are versed in it, then either you have contact with an expert in the topic or we hear about it often.

I don't envy the job of sales engineers with the large portfolio of products NI has.

[rolfk]

Well, it is not difficult to do a better job than a support person in an area where you are an expert. Support people at NI have to support a range of products, that is simply impossible to be anything near to an expert in everything. Human mind capacity and time to learn is limited and so is their ability to know anything near to expert level on all topics. Considering the market share of Linux in industrial automation, which is the business NI is still mostly operating in, I would find your findings of 1 out of 5 NI engineers to be able to really help you with useful information about Linux topics to be fairly excellent.

Of course is Linux your topic and your expertise and do you want the first person to get in contact with to be able to give you all the information you think you need, and preferably all the other information you don't yet know you'll need, but that is not possible and is called reality. An important part about engineering is to be able and learn to deal with reality in every aspect of work we do. This can be lobbying for extra support for something, but often just means to find the right tools that do the job and the right person that has the right answers and if you can't get them, try to create those tools. There is nothing that would prevent someone from adjusting comedi somewhat and then creating a wrapper around comedi that is at least 95% compatible to the LabVIEW DAQmx interface, other than lack of interest for such a solution. And the fact that it hasn't been done yet points probably into one direction: The effort to do it does not seem to outweight the gain for the limited interested audience.

Lets face it, I would love to work on such a project if I had the time and resources, but I have a job and need at the end of the month a paycheck to pay my bills. And I would assume this is the case for most others into this. If you are in industrial business you need to create things that can be sold. A comedi interface can't be sold, it is more like a service to the community which is mostly for university and hobby use. But this community also seems to lack the interest or ability to create such a solution themselves, possibly because of the admittingly enormous task it involves and the rather limited audience that would use it.