Leaderboard

Many moons ago we were using the NI 32 channel digital input and output cards (now no longer available). These boards could sink a couple of amps per channel and we needed at least 1.5 amps per channel. There were no other cards on the market that could sink amps (mostly a few hundred milliamps) and the lead time became 4 months. We needed 128 digital outputs and 64 digital inputs (6 cards per machine). Talking with the Electrical guy on the project he said we needed the amps so that we could activate 24v control signals directly without buying relays to intermediate (we would have needed 128 relays at about $8 a pop adding ~ $1000 unbudgeted per machine as well as the headaches of trying to mount them) and the cards were only about $300. He said he could do better in 3 weeks So he designed a 64 channel card with 32 digital in and 32 digital out that could sink 3 amps per channel. He designed each with a serial port (the NI originals were PCI cards) but I wasn't happy with that and made him put in RS485 (today I would have gone with ethernet). So we ended up with 32 channel input, 32 channel output with 1Mbit RS485 comms, able to sink 3 amps per channel. The BOM for the cards cost $30 and we got them built for $100 as prototypes-no doubt that cost would come down if we had gone into production (the electrical guy said that mass production cost would have been about $90 per card, all in). So now we only needed 4 cards and had oodles of digital inputs spare at 60% the cost of the NI cards, although we did need an NI RS485 card so it broke about even. Need more IO? Just stick another card on the bus . I wrote the firmware for the cards and a LabVIEW driver because, of course, the NI drivers were for PCI, not RS485, but there were drop-in replacement VI's and a couple of VI's that had more features that I'd always wished the NI drivers had had. FPGA's are probably a different kettle of fish, however, due to throughput But there are advantages of doing your own thing. The one thing I did learn from the exercise was, if possible, don't put cards in the computer. From that time forward I always looked for ethernet, RS495, or Profibus IO (preferably SCPI instruments where applicable) so that different manufacturers hardware could be easily swapped out-software changes because of hardware changes don't scare me. It also means you don't need the headaches that come with PCI expanders and large industrial computers as well as, pretty much care-free, mounting and wire routing with all the associated connector stress problems.

Dear Santa NI I am now in my 40s with youngish kids, so despite the fact that all I got for Christmas this year was a Pickle Rick cushion I am not actually complaining. However, I would like to get my order in to the Elves as early as possible. This is my wishlist, in no particular order. I expect this list will not be to everyone's taste, this is ok, this is just my opinion. Make LabVIEW free forever. The war is over, Python has won. If you want to be relevant in 5 to 10 years you need to embrace this. The community edition is a great start but is is probably not enough. Note: I accept it might be necessary to charge for FPGA stuff where I presume you license Xilinx tools. NI is and has always been a hardware company. Make all toolkits free. See the above point. Remove all third party licensing stuff. Nobody makes any money from this anyway. Encourage completely open sharing of code and lead by example. Take all the software engineering knowledge gained during the NXG experiment and start a deep refactor of the current gen IDE. Small changes here please though... we should not have to wait 10 years. Listen to the feedback of your most passionate users during this refactor. NXG failed because you ignored us and just assumed we would consume whatever was placed in front of us. I am talking about the people like those reading this post on Christmas day and their spare time because they are so deeply committed to LabVIEW My eyes are not what they used to be, so please bring in the NXG style vector graphic support so I can adjust the zoom of my block diagram and front panel to suit accordingly As part of the deep refactor, the run-time GUI needs to be modernised. We need proper support for resizable GUIs that react sensible to high DPI environments. Bring the best bits of NXG over to current gen. For example the dockable properties pane. (Sorry not much else comes to mind) Remove support for Linux and Mac and start to prune this cross compatibility from the codebase. I know this is going to get me flamed for eternity from 0.1 % of the users. (You pretty much made this decision for NXG already). Windows 10 is a great OS and has won the war here. Get rid of the 32-bit version and make RT 64-bit compatible. You are a decade overdue here. Add unicode support. I have only needed this a few times, but it is mandatory for a multicultural language in 2021 and going forward Port the Web Module to Current Gen. All the news I have heard is that the Web Module is going to become a standalone product. Please bring this into Current Gen. This has so much potential. Stop adding features for a few years. Spend the engineering effort polishing. Fix the random weirdness we get when deploying to RT Open source as many toolkits as you can. Move the Vision toolkit over to OpenCV and make it open source Sell your hardware a bit cheaper. We love your hardware and the integration with LabVIEW but when you are a big multiple more expensive than a competitor it is very hard to justify the cost. Allow people to source NI hardware through whatever channel makes most sense to them. Currently the rules on hardware purchasing across regions are ridiculous. Bring ni.com into the 21st century. The website is a dinosaur and makes me sad whenever I have to use it Re-engage with universities to inspire the next generation of young engineers and makers. This will be much easier if the price is zero Re-engage with the community of your most passionate supporters. Lately it has felt like there is a black hole when communicating with you Engineer ambitiously? What does this even mean? The people using your products are doing their best, please don't patronise us with this slogan. Take the hard work done in NXG and make VIs into a non-binary format human readable so that we can diff and merge with our choice of SCC tools Remove all hurdles to hand-editing of these files (no more pointless hashes for "protection" of .lvlibs and VIs etc) Openly publish the file formats to allow advanced users to make toolkits. We have some seriously skilled users here who already know how to edit the binary versions! Embrace this, it can only help you. Introduce some kind of virtualenv ala Python. i.e. allow libraries and toolkits to be installed on a per-project basis. (I think this is something JKI are investigating with their new Project Dragon thing) For the love of all that is holy do not integrate Git deeply into LabVIEW. Nobody wants to be locked into someone else's choice of SCC. (That said, I do think everyone should use Git anyway, this is another war that has been won). That is about it for now. All I want is for you guys to succeed so my career of nearly 20 years does not need to be flushed down the toilet like 2020. Love you Neil (Edited: added a few more bullets)

I found this gem! Strange that that isn't the recommended setting in the tutorial. http://digital.ni.com/public.nsf/allkb/4A8B626B55B96C248625796000569FA9